Characterization of virulence of Corynebacterium ulcerans

(Charakterisierung der Virulenzeigenschaften von Corynebacterium ulcerans)

Der Naturwissenschaftlichen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg

zur Erlangung des Doktorgrades Dr. rer. nat.

vorgelegt von

Elena Hacker

aus Erlangen

Als Dissertation genehmigt von der Naturwissenschaftlichen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung: 07. März 2016

Vorsitzender des Promotionsorgans: Prof. Dr. Jörn Wilms

Gutachter: Prof. Dr. Andreas Burkovski

apl. Prof. Dr. Andreas Tauch

Table of contents

Table of contents

1 Summary/Zusammenfassung ...... 1

1.1 Summary ...... 1

1.2 Zusammenfassung ...... 3

2 Introduction ...... 6

2.1 Corynebacteria ...... 6

2.2 Pathogenicity of Corynebacterium ulcerans...... 8

2.2.1 Epidemiology, occurrence and disease symptoms ...... 8

2.2.2 Next generation sequencing of C. ulcerans genomes ...... 11

2.2.3 Virulence factors ...... 12

2.2.3.1 Diphtheria toxin and lysogenic corynephages ...... 12

2.2.3.2 Shiga-like toxin ...... 15

2.2.3.3 Phospholipase D (PLD) ...... 16

2.2.3.4 Other pathogenicity determinants of C. ulcerans ...... 17

2.2.3.5 Identification and characterization of virulence factors ...... 22

2.3 Interaction of C. ulcerans with host cells and cell components ...... 24

2.4 Caenorhabditis elegans as infection model system for pathogenic corynebacteria .26

3 Aim of the work ...... 29

4 Publications ...... 30

4.1 Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources (Hacker et al., 2015a) ...... 30

4.2 The killing of macrophages by Corynebacterium ulcerans (Hacker et al., 2015b) ...41

4.3 Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae (Antunes et al., 2015a) ...... 69

4.4 Caenorhabditis elegans star formation and negative chemotaxis induced by infection with corynebacteria (Antunes et al., 2015b) ...... 79

5 References ...... 90

6 Curriculum vitae ...... 100

7 Acknowledgment/Danksagung ...... 102

Hacker, E.

Summary/Zusammenfassung

1 Summary/Zusammenfassung

1.1 Summary

The emerging infectious agent Corynebacterium ulcerans has been underestimated as human pathogen for a long time. C. ulcerans was mainly associated with mastitis in cattle, non-human primates, and other animals and is also known to be commensal in various domestic and wild animals. While infections in humans were formerly rarely reported and mainly associated with direct contact to infected cattle, these cases appear to be increasing in different industrialized countries during the last decade. Most often, these infections can be ascribed to zoonotic transmission through pet dogs and cats, although transmission of the bacterium from human to human cannot be excluded. Besides causing a respiratory diphtheria-like illness, C. ulcerans can also cause extrapharyngeal infections in humans, including severe pulmonary infections, skin ulcers, and severe necrotizing fasciitis (Mattos-Guaraldi et al., 2008; Meinel et al., 2014, 2015). Meanwhile, respiratory diphtheria-like illnesses caused by toxigenic C. ulcerans are reported more often in Western Europe than infections caused by the classical causative agent of diphtheria, Corynebacterium diphtheriae. Nevertheless, to date, only little is known about factors besides the diphtheria toxin that contribute to virulence of this pathogen and data elucidating mechanisms it uses for host colonization are missing. In this study, first investigations were made to analyze the interaction of C. ulcerans with host cells, and the contribution of different putative virulence factors in these processes. For this purpose, the two C. ulcerans strains 809, isolated from an 80-year-old woman with fatal pulmonary infection, and BR-AD22, isolated from an asymptomatic dog, were subjected to different cell-based assays. Two genes that have a putative function in virulence of C. ulcerans could be inactivated by insertion mutagenesis in strain BR-AD22. The pld gene codes for phospholipase D (PLD), the major virulence factor of Corynebacterium pseudotuberculosis. In this closely related species, PLD plays a crucial role in dissemination of the bacteria within the host. The resulting mutant in C. ulcerans BR-AD22 was designated ELHA1. In the second mutant strain ELHA3, CULC22_00609, a homolog of C. diphtheriae DIP0733 was inactivated. DIP0733 was found recently to be involved in adhesion to and invasion of epithelial cells and host cell death (Sabbadini et al., 2012; Antunes et al., 2015a). Adhesion and invasion assays with the two epithelial cell lines HeLa (cervix carcinoma cells) and Detroit562 (pharynx carcinoma cells) revealed that C. ulcerans is able to bind to epithelial cells in high amounts and multiply during in vitro infection. Moreover, the bacteria were able to invade the cytoplasm of epithelial cells. Mutant strains of the putative virulence factors phospholipase D and DIP0733 homolog CULC22_00609 showed no influence on colonization under the experimental conditions tested. The observed internalization into

Hacker, E. 1

Summary/Zusammenfassung epithelial cells might be an effective mechanism for immune evasion of C. ulcerans supporting the establishment and progress of infections. Scanning electron microscopy of cells infected with C. ulcerans suggested a clustered and localized adhesion pattern, indicating the presence of a limited number of specific cell surface receptor sites. Typical V-shaped bacteria due to snapping division seen in fluorescence microscopy were a sign for growth both on the surface and inside the epithelial cells. Furthermore, as was shown by measurement of the transepithelial resistance of infected Detroit562 cell monolayers, C. ulcerans has a detrimental effect on eukaryotic cells (Hacker et al., 2015a). Once a pathogen has entered its host, it is immediately confronted with the innate immune system, helping the host to eliminate the foreign material again. To investigate the interaction of C. ulcerans with components of this defense system, the human monocytic cell line THP-1, which can be differentiated into macrophage-like cells, was infected with C. ulcerans wild type strains and corresponding mutants. The survival and intracellular replication of the bacteria in this phagocytic cell line was analyzed by counting intracellular CFU after different points in time. In addition, the reaction of the cells to the bacterial infection was studied by determination of cytokine levels, activation of NF-κB, and release of LDH as a sign of cell death. Through FACS analysis, the influence of C. ulcerans on cell death was investigated. We showed that this bacterium is not immediately inactivated, but rather was able to survive inside THP-1 macrophages for several hours and even replicate during the first hours of infection. Therefore, macrophage function is obviously impaired through C. ulcerans infection. This was confirmed by Lysotracker staining and fluorescence microscopy, which exhibited a delay in phagolysosome fusion. Upon infection, THP-1 cells produced high amounts of the cytokines IL-6 and G-CSF and the NF-κB signaling cascade was activated. Furthermore, LDH activity measured in the supernatants of infected cells indicated cell death. This idea was supported by detachment of adherent cells after infection with C. ulcerans. By the use of 7- AAD staining and FACS analysis, it was verified that this species causes a form of cell death in THP-1 cells resembling necrosis (Hacker et al., 2015b). Moreover, the nematode Caenorhabditis elegans was used as model system to study virulence properties of C. ulcerans in comparison with the closely related pathogen C. diphtheriae and the non-pathogenic species C. glutamicum (Antunes et al., 2015b). All these coryneform bacteria could induce star formation in C. elegans, a symptom which was previously described to occur upon infection with natural C. elegans pathogens. In addition, a severe tail-swelling phenotype, the so called Dar (deformed anal region) formation, was observed for all strains. C. diphtheriae and C. ulcerans could colonize the nematodes, persist there and impair the survival of the host dramatically. Interestingly, C. elegans was able to distinguish between pathogenic and non-pathogenic food sources and showed aversive learning behavior.

Hacker, E. 2

Summary/Zusammenfassung

1.2 Zusammenfassung

Der immer häufiger auftretende Erreger Corynebacterium ulcerans wurde lange Zeit als Humanpathogen außer Acht gelassen. C. ulcerans wurde hauptsächlich mit Mastitis in Vieh, nicht-menschlichen Primaten und anderen Tieren in Verbindung gebracht und ist auch als Kommensale in einer Reihe an Haus- und Wildtieren bekannt. Während Infektionen in Menschen früher selten berichtet und vor allem mit direktem Kontakt zu infiziertem Vieh in Verbindung gebracht wurden, scheinen diese im letzten Jahrzehnt in verschiedenen industrialisierten Nationen zuzunehmen. Die meisten dieser Fälle können auf zoonotische Übertragung durch Haushunde und -katzen zurückgeführt werden. Die Übertragung von Mensch zu Mensch kann allerdings nicht ausgeschlossen werden. Neben respiratorischen Diphtherie-ähnlichen Erkrankungen, kann C. ulcerans auch extrapharyngale Infektionen, wie unter anderem schwere Lungeninfektionen, Hautgeschwüre und schwere nekrotisierende Fasziitis in Menschen auslösen (Mattos-Guaraldi et al., 2008; Meinel et al., 2014, 2015). Mittlerweile wird von respiratorischen Diphtherie-ähnlichen Erkrankungen, die von toxigenen C. ulcerans Stämmen verursacht werden, in Westeuropa öfter berichtet als von durch den klassischen Diphtherie-Erreger Corynebacterium diphtheriae hervorgerufene Infektionen. Nichtsdestotrotz ist bislang nur wenig über Faktoren neben dem Diphtherietoxin bekannt, die zur Virulenz dieses Pathogenes beitragen und experimentelle Daten über Mechanismen welche dieses Pathogen nutzt um den Wirt zu kolonisieren, fehlen. Im Rahmen dieser Arbeit wurden erste Untersuchungen zur Analyse der Interaktion von C. ulcerans mit Wirtszellen und der Beteiligung verschiedener putativer Virulenzfaktoren in diesen Prozessen gemacht. Dazu wurden die beiden C. ulcerans Stämme 809, der aus einer 80-jährigen Frau mit fataler Lungenentzündung isoliert wurde, und BR-AD22, der aus einem asymptomatischen Hund isoliert wurde, verschiedenen Zell-basierten Assays unterzogen. Zwei Gene mit einer putativen Virulenzfunktion in C. ulcerans konnten durch Insertionsmutagenese in BR-AD22 inaktiviert werden. Das Gen pld kodiert für die Phospholipase D (PLD), den Hauptvirulenzfaktor von Corynebacterium pseudotuberculosis. In dieser nah verwandten Spezies spielt PLD eine wichtige Rolle in der Verbreitung des Bakteriums innerhalb des Wirts. Die erhaltene Mutante wurde ELHA1 genannt. In der zweiten Mutante ELHA3 wurde CULC22_00609 inaktiviert, ein Homolog von DIP0733 aus C. diphtheriae, für welches seit kurzem bekannt ist, dass es an der Adhäsion und Invasion von Epithelzellen und an deren Zelltod beteiligt ist (Sabbadini et al., 2012; Antunes et al., 2015a). Adhäsions- und Invasionsanalysen mit den zwei Epithelzelllinien HeLa (Cervixcarcinom- zellen) und Detroit562 (Pharynxcarcinomzellen) zeigten, dass C. ulcerans in der Lage ist, in großer Zahl an Epithelzellen zu binden und sich während der in vitro-Infektion zu vermehren. Darüber hinaus waren die Bakterien fähig in das Zytoplasma der Epithelzellen einzudringen.

Hacker, E. 3

Summary/Zusammenfassung

Die Insertionsmutanten in den putativen Virulenzfaktoren PLD und dem DIP0733-Homolog zeigten unter den getesteten experimentellen Bedingungen keinen Einfluss auf die Wirtskolonisierung. Die beobachtete Internalisierung von C. ulcerans in Epithelzellen könnte ein effektiver Mechanismus von C. ulcerans sein, die Immunantwort zu umgehen, was folglich die Etablierung sowie den weiteren Fortschritt der Infektion fördert. Rasterelektronen- mikroskopie von infizierten Zellen zeigte ein geclustertes und lokalisiertes Adhäsionsmuster von C. ulcerans, welches auf eine limitierte Anzahl an spezifischen Rezeptorbindestellen schließen lässt. Bei fluoreszenzmikroskopischen Analysen wurden charakteristische V- förmige bakterielle Teilungsstadien, die aufgrund der für Corynebakterien typischen Schnappteilung entstehen, beobachtet. Diese sind ein Zeichen für Wachstum der Bakterien sowohl auf der Oberfläche, als auch innerhalb der Epithelzellen. Des Weiteren konnte durch Messung des transepithelialen Widerstands von infizierten Detroit562-Zell-Monolayern gezeigt werden, dass C. ulcerans einen zellschädigenden Effekt hat (Hacker et al., 2015a). Ist ein Pathogen in seinen Wirt eingedrungen, wird es sofort mit dem angeborenen Immunsystem konfrontiert, welches dem Wirt dabei hilft, das fremde Material wieder zu eliminieren. Um die Interaktion von C. ulcerans mit Komponenten dieses Abwehrsystems zu untersuchen, wurde die humane monozytische Zelllinie THP-1, welche in Makrophagen- ähnliche Zellen differenziert werden kann, mit C. ulcerans Wildtyp- und daraus abgeleiteten Mutantenstämmen infiziert. Hier wurde einerseits das Überleben in dieser phagozytischen Zelllinie durch Zählen der intrazellulären CFU nach verschiedenen Zeitpunkten analysiert. Andererseits wurde die Reaktion der Zellen auf die bakterielle Infektion durch Bestimmung der Zytokinlevel, Aktivierung von NF-κB und der Freisetzung von LDH als Zeichen von Zelltod, untersucht. Durch FACS-Analysen wurde genauer überprüft, ob und wie die Zellen durch die C. ulcerans Infektion sterben. Dieses Bakterium war in der Lage mehrere Stunden in THP-1 Makrophagen zu überleben und sich während der ersten Stunden der Infektion sogar zu vermehren. Diese offensichtlich beeinträchtige Funktionalität der Makrophagen wurde durch Lysotracker-Färbung und Fluoreszenzmikroskopie bestätigt indem gezeigt wurde, dass die Phagolysosomenfusion durch C. ulcerans verzögert wird. Als Reaktion auf die Infektion produzierten THP-1 Zellen große Mengen der Zytokine IL-6 und G-CSF und die NF-κB Signalkaskade wurde aktiviert. Außerdem konnte LDH-Aktivität im Überstand infizierter Zellen gemessen werden, was auf Zelltod hindeutet. Diese Vermutung wurde durch das Ablösen adhärenter Zellen nach Infektion mit C. ulcerans unterstützt. Durch 7-AAD-Färbung und FACS- Analysen konnte bestätigt werden, dass diese Spezies einen Nekrose-ähnlichen Zelltod in THP-1 Zellen verursachen kann (Hacker et al., 2015b). Des Weiteren wurde der Nematode Caenorhabditis elegans als Modellsystem verwendet um Virulenzeigenschaften von C. ulcerans verglichen mit dem nah verwandten Pathogen C. diphtheriae und der nicht-pathogenen Spezies Corynebacterium glutamicum zu untersuchen

Hacker, E. 4

Summary/Zusammenfassung

(Antunes et al., 2015b). Es konnte gezeigt werden, dass all diese coryneformen Bakterien star formation in C. elegans induzieren können, ein Symptom, das schon für natürliche C. elegans Pathogene beschrieben wurde. Außerdem wurde ein starkes Anschwellen des Schwanzes, die sogenannte Dar (deformed anal region) -Bildung beobachtet. C. diphtheriae und C. ulcerans können die Nematoden kolonisieren, dort persistieren und das Überleben des Wirts dramatisch beeinträchtigen. Interessanterweise war C. elegans in der Lage, zwischen pathogenen und nicht-pathogenen Futterquellen zu unterscheiden und zeigte ein aversives Lernverhalten.

Hacker, E. 5

Introduction

2 Introduction

2.1 Corynebacteria

According to the latest edition of “Bergey’s Manual of Systematic Bacteriology”, the genus Corynebacterium belongs to the family Corynebacteriaceae, the order Corynebacteriales, and the class Actinobacteria (Busse, 2012; Goodfellow, 2012; Goodfellow & Jones, 2012). The class Actinobacteria includes Gram-positive bacteria with a high G+C content in their DNA base composition and has been subdivided into 15 orders, 43 families, and 203 genera based on 16S rRNA gene sequencing and supragenic relationships (Ventura et al., 2007; Goodfellow, 2012). Members of the order Corynebacteriales are characterized by the cell wall chemotype IV with arabinose and galactose as the major cell wall sugars and cell wall peptidoglycan of the A1γ type with meso-diaminopimelic acid. Another typical feature is the presence of mycolic acids (Goodfellow & Jones, 2012). Together with the monospecific genus Turicella, the genus Corynebacterium forms the family Corynebacteriaceae, which is clearly separated from related families classified in the order Corynebacteriales (Busse, 2012; Tauch & Sandbote, 2014). The genus Corynebacterium was first described by Lehmann and Neumann in 1896 as a taxonomic group of bacteria with morphological similarities with the diphtheroid bacillus (Lehmann & Neumann, 1896). Today, 97 species have been assigned to the genus Corynebacterium, including soil bacteria with biotechnological importance, commensals of humans or animals, as well as bacteria with pathogenic effects towards mammals (Tauch & Sandbote, 2014; Rückert et al., 2015). The most prominent member of the genus Corynebacterium is Corynebacterium diphtheriae, representing the type species of this diverse taxon and known as the causative agent of the disease diphtheria (Barksdale, 1970; Bernard & Funke, 2012; Burkovski, 2014). Classical respiratory diphtheria is characterized by a sore throat, low fever, and malaise. Symptoms range from mild pharyngitis to severe hypoxia with pseudomembrane formation due to the diphtheria toxin. This potent virulence factor is transferred to the bacterial genome through lysogenization by corynephages. Besides C. diphtheriae, two other species of the genus Corynebacterium can be infected by tox gene- carrying corynebacteriophages and are therefore potential DT producers: C. pseudotuberculosis and C. ulcerans. These three pathogens and potential carriers of the phage-borne diphtheria toxin gene form the “C. diphtheriae group” (Riegel et al., 1995; Funke et al., 1997). Interestingly, the three closely related species colonize men and animals to different extents (see Figure 1). As mentioned previously, C. diphtheriae is the classical agent of the respiratory disease diphtheria in humans. The transmission occurs from person to person by close physical contact via respiratory droplets and smear infections. Infections in

Hacker, E. 6

Introduction animals have been reported but are very uncommon. However, C. diphtheriae has been isolated from cows, horses and domestic cats (Corboz et al., 1996; Hall et al., 2010; Leggett et al., 2010). Transmission from animals to humans has not been reported, but the possibility cannot be excluded. C. pseudotuberculosis, in contrast, mainly infects sheep and goats, where it causes caseous lymphadenitis, a disease that leads to economic losses worldwide (Dorella et al., 2006). The bacteria may be transmitted to humans through close physical contact with infected cattle. Although infections in humans through toxigenic C. pseudotuberculosis are rather rare and symptoms of classical diphtheria have not been reported, this species is a potential host of tox-carrying corynephages, for which it may serve as a reservoir. C. ulcerans is a commensal microorganism in domestic and wild animals and was primarily known for causing mastitis in cattle (Hommez et al., 1999). Human infections were rare and have traditionally been reported among rural populations with direct contact to domestic livestock or consumption of raw milk and other unpasteurized dairy products (Bostock et al., 1984; Hart, 1984). However, during the last decade, human infections associated with C. ulcerans appear to be increasing in various countries and can most often be ascribed to zoonotic transmission.

C. ulcerans C. diphtheriae C. diphtheriae

C. ulcerans C. ulcerans C. pseudotuberculosis

Figure 1: Transmission of toxigenic corynebacteria. Transmission of C. diphtheriae from person to person occurs by close physical contact and this species has also been isolated from animals. C. ulcerans infections occur via close physical contact with pets and farm animals or animal bites. However, person-to-person transmission of C. ulcerans cannot be excluded. C. pseudotuberculosis mainly infects cattle, sheep, and goats and can be transferred from infected animals to humans by close physical contact (modified after Burkovski, 2014).

Hacker, E. 7

Introduction

2.2 Pathogenicity of Corynebacterium ulcerans

2.2.1 Epidemiology, occurrence and disease symptoms

Disease symptoms caused by C. ulcerans in humans include, in addition to diphtheria-like illness and classical respiratory diphtheria with pseudomembrane formation, tonsillitis, pharyngitis, pneumonia, peritonitis, lymph node abscesses, subcutanous abscesses, skin ulcers, and severe necrotizing fasciitis (Taylor et al., 2002; Elden et al., 2007; Mattos-Guaraldi et al., 2008; Schuhegger et al., 2009; Berger et al., 2011; Putong et al., 2011; Corti et al., 2012; Berger et al., 2013; Yoshimura et al., 2014; Meinel et al., 2015). Figure 2 gives examples of different disease manifestations.

A B

C D

Figure 2: Different disease manifestations in humans with C. ulcerans infection. A) Bull neck appearance and B) pseudomembrane formation on the tonsils of a 15-year-old male from the Philippines (Putong et al., 2011), C) skin lesion with yellowish membrane on the leg of an 80-year-old woman from Brazil (Mattos-Guaraldi et al., 2008) and D) skin ulcer covered with a gray-brown membrane and erythema on the hand of a 29-year-old male from Switzerland (Corti et al., 2012).

In most cases, patients had close contact to pet animals like cats and dogs, which are likely to be the transmitters of the bacterium to its human host. The range of mammal hosts of C. ulcerans that may serve as a reservoir for human infections is very broad and includes, besides the classically described farm animals like cattle, dogs, cats, goats, pigs, ground squirrels, otters, camels, monkeys, whales, wild boars and others (Olson et al., 1988; Hommez et al.,

Hacker, E. 8

Introduction

1999; Bergin et al., 2000; Tejedor et al., 2000; Foster et al., 2002; Morris et al., 2005; Seto et al., 2008; Hogg et al., 2009; Schuhegger et al., 2009; Sykes et al., 2010; Contzen et al., 2011; Hirai-Yuki et al., 2013). Person-to-person transmission of C. ulcerans has not yet been verified but the possibility cannot be excluded. Wagner et al. described two cases where a patient suffering from C. ulcerans infection had contact with an asymptomatic C. ulcerans carrier. In one case, toxigenic C. ulcerans was isolated from a 20-year-old male with a sore throat and from his 18- year-old sibling. In the other case, a 35-year-old male, who had respiratory diphtheria with a pseudomembrane, harbored toxigenic C. ulcerans. The organism was also isolated from his asymptomatic 11-year-old son. In both cases, no animal source for infection with C. ulcerans was found (Wagner et al., 2010). Another two-person cluster was reported by Konrad et al. only recently. In this case report, toxigenic C. ulcerans was isolated from a 13-year-old girl with tonsillitis and her asymptomatic 81-year-old grandmother, both living in the same household. Although the family lived on a farm with cattle, pigs, a dog, a pet cat, and three farm cats, the girl only had close contact to the pet cat and the dog, and the grandmother had no close animal contact. Unfortunately, testing of the animals was not performed, but all were symptomless. The authors concluded that a person-to-person transmission after an initial zoonotic transmission from the pet cat to the girl might be likely (Konrad et al., 2015). The transmission from animals to humans has been only occasionally verified through molecular biological techniques. In two cases from France, the C. ulcerans strains isolated from the patients were indistinguishable from samples isolated from their dogs (Lartigue et al., 2005; Bonmarin et al., 2009). In another case in the United Kingdom, it was also verified through molecular typing that the isolates from the patient and her dogs correspond to a single C. ulcerans strain (Hogg et al., 2009). Transmission through a pig was confirmed in one case in Germany, where a 56-year-old female farmer showed severe signs of classical respiratory diphtheria. All family members as well as several farm animals were analyzed for the presence of C. ulcerans. In one of the 19 asymptomatic pigs examined, a toxigenic strain of C. ulcerans was found. Sequencing and ribotyping confirmed the identity of both strains (Schuhegger et al., 2009). In 2010, C. ulcerans was isolated from an 86-year-old woman in a German hospital. Here, it was proven by ribotyping and multi locus sequence typing (MLST) that the human isolate was identical to an isolate from her asymptomatic cat (Berger et al., 2011). In another case of C. ulcerans infection in Japan, pulsed-field gel electrophoresis showed that isolates from nasal secretion and pharyngeal content from the patient’s cat was to 100 % identical to the strain isolated from the patient (Yoshimura et al., 2014). Very recently, two non-toxigenic tox gene-bearing C. ulcerans strains were isolated from a traumatic ulcer from a 61-year-old male and his asymptomatic dog. Both share the same MLST type, and only three single

Hacker, E. 9

Introduction nucleotide polymorphisms (SNPs) were found to differentiate the two isolates, supporting a zoonotic transmission to or between the patient and his dog (Fuursted et al., 2015). Despite the previous examples, the sources for infection and modes of transmission are in most cases speculative. Furthermore, the number of infections may be strongly underestimated as many clinicians may not be aware of the possibility of infection with this pathogen. In addition, qualified laboratories for the isolation and identification of C. ulcerans are unavailable in many cases. Here, further investigations about the clinical and microbiological characteristics of the infection with this pathogen are necessary, as well as the introduction or improvement of routine laboratory procedures to diagnose C. ulcerans infection. König and coworkers published a protocol for multilocus sequence typing (MLST) to type C. ulcerans in a rapid and cost-effective way. MLST is one of the most important tools in modern epidemiology to understand transmission pathways. This technique gives insights into zoonotic transmission of C. ulcerans from animals to humans and human-to-human or animal-to-animal transmission. For MLST in C. ulcerans, König and coworkers amplified and sequenced internal fragments of seven house-keeping genes according to the published scheme for C. diphtheriae with minor modifications. This study revealed that all sequence types (STs) derived from animals, which were at least twice present in their analysis, were also found in human isolates. With this method, it was also demonstrated that there are certain C. ulcerans STs which carry the gene for diphtheria toxin more often, suggesting that these STs might be connected with higher virulence and a more severe outcome of the disease (König et al., 2014). Meinel and coworkers used comparative genomic and next generation sequencing (NGS) to understand the transmission pathway of C. ulcerans in detail (Meinel et al., 2014). Nine toxigenic C. ulcerans isolates from human patients from Germany and their domestic animals were analyzed by NGS followed by SNP detection and the obtained genomic information was compared with published genomes from Brazil, Japan, and France. Strains from different groups showed a high variation of SNPs (> 5000), whereas the isolates within a pair (patient and patient’s pet animal) only exhibited single differences, indicating an origin of a common precursor. This result clearly demonstrated a zoonotic transmission of toxigenic C. ulcerans between animals and humans at a molecular level. Through further phylogenetic analysis, it was shown that three of the four German pairs cluster together with a French isolate, whereas the Brazilian and Japanese isolates seem to be less related. This indicates the existence of a European genotype of C. ulcerans that differs from the genotypes described for South America and Asia. Comparing these results with MLST, which is a fast and cost-effective tool for rough phylogenetic analysis, NGS gives a much more detailed analysis and a more robust discrimination between closely related isolates and is therefore favorable when a deep understanding of transmission pathways is needed (Meinel et al., 2014).

Hacker, E. 10

Introduction

2.2.2 Next generation sequencing of C. ulcerans genomes

Besides the analysis of transmission pathways, next generation sequencing is a helpful tool to better characterize the virulence mechanisms of this emerging pathogen, to find and analyze factors involved in host-pathogen interaction, to identify antibiotic or vaccine targets, as well as distinct features of strains from human or animal sources. In 2011, the first DNA sequencing study, which focused on the comparison of the complete genome sequences of two non-toxigenic C. ulcerans strains and the detection of candidate virulence factors, was published (Trost et al., 2011). The strains investigated here were C. ulcerans BR-AD22, isolated from an asymptomatic dog, and C. ulcerans 809, isolated from an 80-year-old woman with fatal pulmonary infection (Mattos-Guaraldi et al., 2008; Dias et al., 2010). The two sequenced genomes revealed a high degree of synteny of orthologous genes over the entire length of the bacterial chromosomes. Remarkable differences in the overall synteny of genes were assigned to the varying number of prophages and to the repertoire of potential virulence factors (Trost et al., 2011). In 2012, the genome of the tox- gene carrying strain C. ulcerans 0102, which was isolated from the pharyngeal pseudomembrane of a 52-year-old woman in Japan, was sequenced (Hatanaka et al., 2003; Sekizuka et al., 2012). The complete genome sequence of C. ulcerans 0102 revealed first data on the gene content and the architecture of a tox-positive prophage in this species. These initial genome sequencing studies already indicated that C. ulcerans is highly susceptible for horizontal gene transfer leading to intra-species genome variations which could cause increased virulence of C. ulcerans strains. Since then, numerous genome sequences became available from toxigenic and non- toxigenic C. ulcerans strains: toxigenic strain RAH1, isolated from a 51-year-old woman with symptoms of respiratory diphtheria (Sangal et al., 2014); non-toxigenic strain FRC58, isolated from the secretions of an 86-year-old hospitalized patient with bronchitis (Silva et al., 2014); strain 210932, a toxigenic isolate from a non-specified human clinical source (Viana et al., 2014); strain FRC11, isolated from a 74-year-old human with a leg ulceration (Benevides et al., 2015); toxigenic strains LSPQ-04227 and LSPQ-04228, isolated from two different patients, one from a blood sample and the other one from a scar exudate following thoracic surgery (Domingo et al., 2015); and finally non-toxigenic strains 1090-14 and 1130-14, isolated from the wound of a 61-year-old countryside gardener and from a nasal specimen from the patient's asymptomatic dog, respectively (Fuursted et al., 2015). In an additional study, nine isolates derived from human patients and their domestic animals were analyzed by applying next-generation sequencing and comparative genomics (Meinel et al., 2014). In a similar approach, toxigenic strains isolated from a wound swab sample of a 53-year-old man in Germany and from the patient‘s asymptomatic pet dog were sequenced and compared to

Hacker, E. 11

Introduction support recent zoonotic transmission of this bacterial pathogen (Meinel et al., 2015). The National Center for Biotechnology Information (NCBI) genome database lists additional C. ulcerans genome sequences, complete and permanent draft, that are not yet published. The current set of complete genome sequences shows that the C. ulcerans genome consists of a circular chromosome within a size range from 2.43 Mb (C. ulcerans 1313002) to 2.61 Mb (C. ulcerans BR-AD22) and a chromosomal G+C content of 53.3 to 53.4 %. The number of protein-coding genes predicted by bioinformatic methods is in the range from 2093 (C. ulcerans 1313002) to 2276 (C. ulcerans BR-AD22) proteins.

2.2.3 Virulence factors

Symptoms evoked due to a C. ulcerans infection may be similar to classical respiratory diphtheria with pseudomembrane formation, but can also include tonsillitis, pneumonia, skin ulcers, and others. These are probably caused by a set of different virulence factors. Bacterial virulence factors in general are compounds (toxins, enzymes, or other molecules) contributing to the infectious potential of a pathogenic microbe and can either confer the ability to attach to and colonize specific host tissues or cause damage to the host (Madigan et al. 2009). At least three different toxins that might be involved in the outcome of disease can be produced by C. ulcerans. One is the diphtheria toxin, transferred to the bacteria via a tox+ corynephage, which can lysogenize C. diphtheriae, C. pseudotuberculosis, and C. ulcerans. Another factor is a Shiga-like toxin that was only recently detected through whole genome sequencing of the human isolate C. ulcerans 809. The last one is phospholipase D, known to be the major virulence factor of the closely related animal pathogen C. pseudotuberculosis. Besides these virulence factors, others are probably contributing to the bacterial pathogenicity. All factors known so far are described in the following.

2.2.3.1 Diphtheria toxin and lysogenic corynephages

The best analyzed virulence factor of C. ulcerans is the diphtheria toxin (DT). This toxin is also the major virulence factor of the closely related species C. diphtheriae and is responsible for the classical symptoms of the respiratory disease diphtheria, characterized by a sore throat, a typical sweetish-putrid smell of the patients’ breath, and the formation of a pseudomembrane in the respiratory tract that can lead to suffocation. In tropical and subtropical regions, cutaneous diphtheria, characterized by skin lesions mainly on hands, feet, and lower legs, is even more common (Höfler, 1991; for recent review, see Burkovski, 2014). DT is a cytotoxic protein inhibiting cellular protein biosynthesis in eukaryotes. Extracellular DT produced by C. diphtheriae is a polypetide consisting of 535 amino acids and

Hacker, E. 12

Introduction has a molecular weight of about 58 kDa (Holmes, 2000). DT is composed of two fragments, the amino-terminal fragment A and the carboxy-terminal fragment B, which are linked by disulfide bonds. Fragment A contains the catalytic domain (C-domain), fragment B corresponds to the translocation and receptor binding domains (T-domain and R domain, respectively). Diphtheria toxin binds to a specific surface receptor of the host, which has been identified as precursor of heparin-binding epidermal growth factor (EGF) like growth factor (proHB-EGF) (Naglich et al., 1992). The amino acid residues Asp106-Pro149, the EGF-like domain, represent the receptor binding site and the interactions with fragment B of DT depend on residue Glu141 (Louie et al., 1997). DT binding results in receptor-driven endocytosis and translocation of the enzymatically active fragment A to the cytoplasm. Due to acidification, the toxin changes conformation, and fragment A is translocated into the cytosol. Fragment A catalyzes the transfer of ADP-ribose from NAD to elongation factor 2 (EF-2). In this way, EF-2 is inactivated and chain elongation during protein synthesis in the host cells is inhibited, resulting in cell death (Pappenheimer, 1977; Holmes, 2000; Collier, 2001; for recent review, see Varol et al., 2014). The tox gene for diphtheria toxin synthesis is located in the genomes of a family of corynephages, which are able to integrate into the chromosomes of C. diphtheriae, C. ulcerans and C. pseudotuberculosis (Sangal & Hoskisson, 2014). Therefore, not all C. ulcerans strains harbor the genomic information for this toxin, but rather only those that are lysogenized by tox- encoding corynephages. The phage attachment sites (att) are conserved between C. ulcerans and C. diphtheriae and located in the tRNA-Arg gene region (Cianciotto et al., 1986; Seto et al., 2008). Sing et al. sequenced the tox gene of two human C. ulcerans isolates associated with extrapharyngeal diphtheria: one, A6361, caused a cutaneous infection and the second, A2911, a lethal necrotizing sinusitis. The nucleotide and deduced amino acid sequences differed about 5 % between the tox genes from the two C. ulcerans samples and a C. diphtheriae reference strain (NCTC_10648). Most of the variations were located in the translocation (T) and the receptor binding (R) domain of the B fragment of the protein (Sing et al., 2003). In another study, Sing et al. analyzed the tox gene of C. ulcerans X959 isolated from a patient with classical pharyngeal diphtheria. Here, C. ulcerans tox differed from the corresponding gene in C. diphtheriae (from the corynephage omega (tox+) genome) in 26 amino acids, resulting in 95 % homology. Again, most of the amino acid exchanges are located in the B fragment of DT. The authors speculated that the amino acids at positions 22 and 183 of the A fragment might be involved in pseudomembrane formation as these are the only two amino acids that differed between the extrapharyngeal C. ulcerans isolates and the C. diphtheriae and C. ulcerans strains from pharyngeal disease (Sing et al., 2005). Meanwhile, more sequences of tox genes or parts of it from different human or animal C. ulcerans isolates are available in the NCBI database. Seto et al. sequenced the tox genes of

Hacker, E. 13

Introduction five more strains isolated in Japan, whereof two were isolated from killer whales (strains Ran and O-9) and three from humans (0509 and 0607 from throat swabs and 0510 from a lung specimen). The two strains isolated from the whales possessed identical tox sites. Two of the human isolates (0509 and 0607) were indistinguishable from a C. ulcerans reference strain (0102), but one (0510) showed slight nucleotide changes resulting in 5 varying amino acid residues compared with the reference strain (99 % homology of nucleotide and deduced amino acid sequence). Compared to a C. diphtheriae reference (tox from corynephage beta), the homology of the nucleotide and amino acid sequence of tox from the C. ulcerans isolates was at only 95 % (Seto et al., 2008). Further 42 tox genes of C. ulcerans isolates obtained from dogs in Japan were sequenced and compared to the published tox sequences of human C. ulcerans isolates. Here, any of the human strains compared with dog isolates showed a similarity of over 98 %. However, similarity was again reduced to around 95 % when these sequences were compared to a C. diphtheriae reference strain (NCTC 13129). When the amino acid sequences of five C. diphtheriae strains were compared among each other, only differences of up to two base pairs, but no variations in the deduced amino acid sequence were found (Katsukawa et al., 2009, 2012). Taken together, the findings of the three different research groups confirmed that all C. ulcerans tox genes form a monocluster distinct from that of C. diphtheriae. Contrary results were only published by Contzen and coworkers, who analyzed a part of the sequence of tox genes from two C. ulcerans isolates from diseased wild boars in Germany and found significant differences between the obtained sequences and nucleotide sequences described for other C. ulcerans strains (95 % similarity), but a higher similarity to C. diphtheriae tox gene (97 % similarity). Although the genomic information for DT was found in these isolates, the Elek test to detect the expression of this toxin gave negative results (non-toxigenic tox-bearing strains) (Contzen et al., 2011). Until recently, conversion of a non-toxigenic to a toxigenic bacterium by prophage integration, as it is well described for C. diphtheriae, was also assumed to be the exclusive mechanism for DT integration in the genome of toxigenic C. ulcerans. However, Meinel and coworkers found a novel, putative diphtheria toxin-encoding pathogenicity island (PAI) that is present in seven of the nine toxigenic strains analyzed in their study (Meinel et al., 2014). This putative PAI is completely different to the known prophage encoding DT, but is located at the same genomic position in all strains. This locus is also known to be targeted by other events of horizontal gene transfer. Integration of toxigenic prophages in the other two strains analyzed and also in the genome of the sequenced strain C. ulcerans 1020 (Sekizuka et al., 2012; Meinel et al., 2014) as well as another putative virulence factor, a Shiga-like toxin of C. ulcerans 809 (see below), was found at the same genomic site. The novel putative PAI has a size of 7571 bp and a low G+C content (48 % compared to an average content of 53 %). Together with a perfect 100 bp direct repeat duplicating parts of the tRNA-Arg at the 3’ end of the PAI, this

Hacker, E. 14

Introduction suggests a horizontal gene transfer. Whether this PAI is specific to C. ulcerans and how exactly it is transferred to the bacterial genome still needs to be elucidated (Meinel et al., 2014). Importantly, diphtheria toxoid vaccine is directed against diphtheria toxin of C. diphtheriae and its efficacy against DT of C. ulcerans still remains unknown because of the amino acid sequence differences between the toxins. Furthermore, as the vaccination is directed against the action of the toxin and not the bacterium itself, the colonization by C. ulcerans (toxigenic or non-toxigenic) is probably not affected.

2.2.3.2 Shiga-like toxin

In 2011, Trost et al. sequenced the whole genomes of two C. ulcerans strains from human clinical (809) and animal (BR-AD22) sources. Bioinformatic analysis revealed the presence of a gene encoding a Shiga-like toxin in strain 809. This gene is annotated as rbp and the gene product as a putative ribosome-binding protein (RBP) containing a domain named “ribosome inactivating protein” (Pfam domain 00161). Although there is only 24 % identity between the deduced amino acid sequence of rbp and the A chain of the Escherichia coli Shiga-like toxins 1 and 2 (SLT-1 and SLT-2, respectively), all highly conserved amino acid residues needed for the catalytic N-glycosidase activity are present. Also, the tertiary in silico structure of RBP showed significant structural similarities to the A chain of SLT-1 from E. coli (Trost et al., 2011) (Figure 3). The protein structure of Shiga-like toxins consists of two domains: the A polypeptide, which confers the activity, and the B polypeptide pentamer, which is responsible for the receptor-specific binding. The A subunit has an N-glycosidase activity that cleaves an adenine from the 28S rRNA of the 60S cytoplasmic ribosome, which renders the 28S rRNA unable to interact with the elongation factors EF-1 and EF-2. In that way protein synthesis in eukaryotic cells is inhibited. The B subunit forms a pentamer that binds to the receptor globotriaosylceramide (Gb3) in eukaryotic cells. Shiga-like toxins then enter the cells by receptor-mediated endocytosis (Endo et al., 1988; Lingwood, 1993; Sandvig & van Deurs, 1994). SLT-1 and SLT-2 are considered to be part of the genome of a toxin-converting lambdoid prophage and to be inserted into bacterial genomes via transduction (Mizutani et al., 1999). In C. ulcerans 809, the rbp gene (CULC809_00177) has a very low G+C content of 45 % compared to 53 % in the complete genome. Furthermore, this gene is located between genes coding for a putative phage integrase (CULC809_00176) and a transposase (CULC809_00178), which suggests a horizontal gene transfer to the genome of C. ulcerans 809 (Trost et al., 2011). The function of this protein in respect to pathogenicity of C. ulcerans still remains to be elucidated by experimental approaches.

Hacker, E. 15

Introduction

Figure 3: 3-D model of the A chain of SLT-1. Similarities with RBP of C. ulcerans 809 are indicated in red (Trost et al., 2011).

2.2.3.3 Phospholipase D (PLD)

Besides the diphtheria toxin and the Shiga-like toxin RBP, phospholipase D (PLD) is thought to be a major virulence factor of this species. In contrast to DT, which is only harbored by C. ulcerans strains lysogenized by tox-carrying bacteriophages, and RBP, which until now was only detected in the genome of one C. ulcerans strain, the gene for the exotoxin PLD is present in the chromosome of all C. ulcerans strains. Phospholipase D is a sphingomyelinase (phosphatidylcholine phosphatidohydrolase, EC 3.1.4.4.) that catalyzes the hydrolysis of phosphatidylcholine to phosphatidate and choline and also cleaves ester bonds in sphingomyelins, which are components of mammalian cell membranes. PLD is the most important virulence determinant of the closely related C. pseudotuberculosis. This pathogen causes primarily caseous lymphadenitis (CLA) in sheep and goats, one of the most prevalent bacterial diseases in sheep, but can also infect other mammals like horses, cattle, and humans (Batey, 1986). Studies with C. pseudotuberculosis PLD-mutant strains showed that PLD is an essential virulence factor that is involved in the dissemination of the bacteria from initial infection sites to secondary sites within the host. Strains with inactivated PLD are unable to cause abscession of the lymph nodes and cannot establish CLA (Hodgson et al., 1992; McNamara et al., 1994). Furthermore, PLD seems to be involved in killing of macrophages, as a PLD-deficient mutant strain caused significantly less cell death in murine J774 macrophages than did the wild type strain (McKean, 2007). C. pseudotuberculosis and C. ulcerans are the only members in the species Corynebacterium that are equipped with this toxin (Barksdale et al., 1981). The PLD genes from C. pseudotuberculosis and C. ulcerans share 80 % homology in the nucleotide sequence and their products 87 % homology in their amino acid sequence (Cuevas & Songer, 1993). For C.

Hacker, E. 16

Introduction pseudotuberculosis, a signal sequence of 26 amino acids was defined with a cleavage site at Pro26 (McNamara et al., 1994) and through sequence homology studies, a signal peptide for C. ulcerans ending at Pro26 could be deduced (McNamara et al., 1995). The secreted proteins are 282 amino acids in length and have a molecular weight of 31.2 kDa for C. pseudotuberculosis and 31.0 kDa for C. ulcerans. While homologs of PLD are produced as exotoxins in Loxosceles spiders (van Meeteren et al., 2004), only one other bacterial species is known to produce a similar phospholipase D to that produced by C. pseudotuberculosis and C. ulcerans: Arcanobacterium haemolyticum. The pld gene from A. haemolyticum has 65 % DNA homology with the two corresponding corynebacterial genes. The homology of the amino acid sequence of the PLD protein is 64 %. The similarity of the PLD from C. pseudotuberculosis, C. ulcerans, and A. haemolyticum is not surprising as A. haemolyticum is similar to pathogenic corynebacteria in different aspects and has formerly also been considered a member of the genus Corynebacterium (Collins et al., 1982). However, no significant similarities with any other phospholipases or toxins from other bacterial species have been found (Cuevas & Songer, 1993). Despite the similarities in nucleotide and amino acid sequence, the genomic organization of pld varies between the two corynebacterial and the arcanobacterial species. While C. ulcerans and C. pseudotuberculosis exhibit a similar genomic environment of the pld gene with a transposase protein encoding gene on the one side and a fagC gene on the other side of pld, in A. haemolyticum, pld is flanked by genes encoding a hypothetical protein and three tRNAs (Hacker et al, 2015a). Studies with an A. haemolyticum PLD-deficient mutant strain revealed reduced adhesion ability to HeLa cells, but enhanced invasion rates. The authors explained this finding by a loss of cell viability upon secretion of PLD by intracellular wild type bacteria, which was proven by a viability assay (Lucas et al., 2010). Experimental data on the involvement of PLD in virulence of C. ulcerans are scarce. Within the scope of this work, the interaction of C. ulcerans wild type strains and a PLD- deficient mutant strain with epithelial cells was investigated. Here, no significant influence of PLD in adhesion to and invasion into epithelial cells was observed (Hacker et al., 2015a). When THP-1 macrophages were infected with this mutant strain, a reduced cytotoxicity compared to the wild type strain could be deduced from lactate dehydrogenase measurements (Hacker et al., 2015b). An effect on host cell viability was already observed for a PLD-deficient C. pseudotuberculosis mutant strain (McKean et al., 2007).

2.2.3.4 Other pathogenicity determinants of C. ulcerans

Corynebacteria are characterized by a complex cell wall architecture with an outer layer of mycolic acids functionally equivalent to the outer membrane of Gram-negative bacteria

Hacker, E. 17

Introduction

(Burkovski, 2013). This permeability barrier is a key feature of the CMNR group of Actinobacteria, comprising the genera Corynebacterium, Mycobacterium, Nocardia and Rhodococcus. The effects of mycolic acids on host-pathogen interaction are best investigated in Mycobacterium tuberculosis. Here, trehalose dimycolate, the cord factor, was shown to inhibit fusion events between phospholipid vesicles inside the host macrophage and to be involved in macrophage activation (Indrigo et al., 2003). The function of mycolic acids as virulence determinant in C. ulcerans has not yet been studied so far but it might be likely that the corynemycolic acid layer plays an important role in stress resistance and pathogenicity also in this species. When the first complete genome sequences of C. ulcerans were published in 2011 for strain 809, a human isolate, and BR-AD22, isolated from an asymptomatic dog, they were screened for the presence of prominent virulence factors (Trost et al., 2011). Both were found to be non-toxigenic in respect to diphtheria toxin but positive for phospholipase D. To find other candidate virulence factors, the annotated proteomes of the two strains were analyzed bioinformatically for protein precursors containing a N-terminal secretion signal and proteins that exhibit a C-terminal LPxTG motif that mediates the anchoring to the bacterial cell wall. In this screen, twelve putative virulence factors present in both strains and two additional proteins which are only encoded in the genome of C. ulcerans 809 were detected: the vsp2 gene, encoding a secreted serine protease, and the rbp gene, coding for a putative ribosome binding protein. In 2012, another complete genome sequence of a C. ulcerans strain isolated from a patient with diphtheria-like illness was published. This strain, named C. ulcerans 0102, carries a tox positive prophage and the sequences for the twelve reported virulence factors the previously sequenced C. ulcerans strains 809 and BR-AD22 share, but not the two unique genes from strain 809 (Sekizuka et al., 2012). In 2014, Sangal and coworkers sequenced a C. ulcerans strain isolated from a 51-year-old woman with classical respiratory diphtheria symptoms. This strain, designated RAH1, was positive for not only the diphtheria toxin and the other twelve putative virulence factors found in the formerly sequenced strains, but also for the vsp2 gene, only known from C. ulcerans 809 so far (Sangal et al., 2014). Very recently, two non-toxigenic tox gene-bearing C. ulcerans strains sharing the same MLST were isolated from a traumatic ulcer from a 61-year-old male (strain 1090-14) and his asymptomatic dog (strain 1130-14) (Fuursted et al., 2015). The isolated strains also harbor the sequence of the only randomly detected vsp2 gene and furthermore, a novel spa variant with only 67 % sequence homology to spaD from the originally sequenced strain C. ulcerans 809. Trost and coworkers found two gene clusters with an LPxTG motif that are considered to encode adhesive pilus structures in both strains analyzed, BR-AD22 and 809. These pili are covalently anchored to the corynebacterial cell wall and are supposed to play a role in the adhesion process of the pathogen to the host cell. The first detected gene cluster is genetically

Hacker, E. 18

Introduction organized like the spaDEF gene region in C. diphtheriae NCTC 13129, which encodes the SpaDEF pilus. In C. diphtheriae, this adhesive pilus consists of SpaD, the major pilin, SpaE, the minor pilin subunit, and SpaF, located at the pilus tip (Gaspar & Ton-That, 2006). The pilus- specific sortases SrtB and SrtC are essential for the assembly of the pilus and are encoded within this spaDEF gene region in C. diphtheriae. In the same way, the sortase genes srtB and srtC are present in the spaDEF region of the two analyzed C. ulcerans strains. Together with the housekeeping sortase gene srtD, they are probably involved in the assembly of the pilus structure on the C. ulcerans cell surface (Trost et al., 2011). A second pilus cluster consists of the minor pilus protein encoding gene spaB, the tip protein encoding gene spaC, and a single sortase gene, srtA. Here, the major pilin subunit, the pilus shaft, is missing when compared to the corresponding gene cluster in C. diphtheriae NCTC 13129 (Ton-That & Schneewind, 2003). In C. diphtheriae, it was shown that the minor pilin subunit SpaB can mediate adherence of the bacteria to human pharyngeal epithelial cells without the pilus shaft (Mandlik et al., 2007; Chang et al., 2011). Therefore, Trost et al. suggested that SpaB/SpaC as homo- or heterodomeric proteins are probably covalently anchored to the cell surface of the two investigated strains and arrange the tight contact between host and bacteria without a pilus shaft (Trost et al., 2011). In the pool of protein precursors with N-terminal secretion signals, homologs of two C. diphtheriae genes coding for enzymes that are probably involved in the interaction of the bacterium with epithelial cells were detected in the sequenced C. ulcerans genomes (Trost et al., 2011). One is the resuscitation-promoting factor interacting protein DIP1281 (RpfI/RpfA) that is involved in both cell surface organization, and adhesion to and internalization in epithelial cells (Hett et al., 2008; Ott et al., 2010). The other is DIP1621, a hydrolase associated to the cell wall (CwlH) and involved in adherence to epithelial cells in C. diphtheriae (Kolodkina et al., 2011). Furthermore, the extracellular neuraminidase NanH is a potential virulence factor of C. ulcerans. In C. diphtheriae, a homolog of this enzyme was characterized and shown to have neuraminidase and trans-sialidase activity (Mattos-Guaraldi et al., 1998; Kim et al., 2010). Sialidases, or neuraminidases, are glycosyl hydrolases catalyzing the removal of terminal sialic acid residues (N-acetylneuraminic acid, Neu5Ac) from different glycoconjugates (Vimr, 1994). Sialic acid is usually the terminal sugar residue on the oligosaccharide chains of cell surface or serum glycoconjugates (Varki, 1993). Diverse organisms including vertebrate animals, viruses, and microorganisms harbor sialidases, but plants do not (Roggentin et al., 1993). Most of the more than 70 known microorganisms with sialidase activity have close contact with mammals, either as commensals or as pathogens (Roggentin et al., 1993; Varki, 1993). Since sialidases can contribute to the recognition of sialic acids exposed on host cell surfaces and consequently support the invasion and spread of the organisms within the host, they are

Hacker, E. 19

Introduction considered as potential virulence factors in many pathogenic bacteria (Vimr et al., 2004). Sialidases also have the capacity to modify the ability of the host to respond to infection by scavenging sialic acids from host cells and decorating their own cell surface with these sialic acids. In that way, bacteria can elude the host immune mechanisms that would otherwise rapidly clear an unsialylated strain (Vimr & Lichtensteiger, 2002). While C. diphtheriae, C. pseudotuberculosis, and C. ulcerans produce sialidases, C. pseudotuberculosis and C. ulcerans were shown to produce the highest levels of the enzyme and induce a rapid development of purulent lesions when introduced into the skin of guinea pigs or rabbits (Arden et al., 1972). Furthermore, Corfield described the direct toxic effect on host tissue or the interference of sialidases with host immunologic and other defense mechanisms, like a decrease of erythrocyte and leukocyte circulation half-life, an increase in the antibody titer, the loss of cell surface receptors, or the creation of bacterial binding sites on epithelial cells leading to invasion (Corfield, 1992). Another gene detected in the C. ulcerans genome sequences is the cpp gene encoding corynebacterial protease 40 (CPP). A homolog of this enzyme of C. pseudotuberculosis has proteolytic activity and was shown to be of the serine protease type (Wilson et al., 1995). It was identified as an antigen that protects sheep against caseous lymphadenitis (Walker et al., 1994). These previous results are doubtful, as an amino acid sequence alignment of the respective enzyme from C. pseudotuberculosis and C. ulcerans with the α-domain of the endoglycosidase EndoE from the human pathogen Enterococcus faecalis revealed that all three proteins harbor the conserved FGH18 motifs (Collin & Fischetti, 2004; Trost et al., 2011). This catalytic motif is a common feature in glycosyl hydrolases of family 18, including members with endo-β-N-acetylglucosaminidase activity (Terwisscha van Scheltinga et al., 1996). E. faecalis EndoE was identified as a secreted enzyme, which by two distinct activities hydrolyzes the glycans on RNase B and IgG. The first activity is mediated by the α-domain and the latter by the β-domain, but both activities are probably important for the molecular pathogenesis and persistence of E. faecalis during human infections (Collin & Fischetti, 2004). In C. ulcerans, the β-domain is missing, but nevertheless, Trost et al. suggested that interaction of the bacteria with mammalian hosts is possible by glycolytic modulation of host glycoproteins (Trost et al., 2011). Moreover, both strains carry the genes tspA and vsp1, coding for extracellular proteins of the serine protease type. In addition, the vsp2 gene detected in C. ulcerans 809 and later also in RAH1, 1090-14, and 1130-14 encodes another secreted serine protease. This enzyme family is known to interact with host cell tissue components or components of the host’s defense system and shows a wide range of pathogenic effects (Dubin, 2002). In 2012, Sabbadini and coworkers studied the involvement of the surface-exposed non- fimbrial 67-72p protein (DIP0733) from C. diphtheriae in the invasion process of the bacteria

Hacker, E. 20

Introduction into epithelial cells (Sabbadini et al., 2012). Previously, this protein was described as adhesin involved in the interaction with human epithelial cells and erythrocytes (Colombo et al., 2001; Hirata et al., 2004). Based on studies using isolated protein, Sabbadini et al. concluded that DIP0733 may play a direct role in bacterial invasion and induction of apoptosis of epithelial cells in the early stages of C. diphtheriae infection (Sabbadini et al., 2012). The mutation of the corresponding gene in C. diphtheriae CDC-E8392 influenced adhesion and invasion efficiency negatively compared to the wild type (Antunes et al., 2015a). The analogous protein in C. ulcerans BR-AD22 is CULC22_00609, annotated as uncharacterized conserved putative membrane protein (NCBI nucleotide database). Comparison of C. diphtheriae NCTC 13129 DIP0733 with C. ulcerans BR-AD22 CULC22_00609 through nucleotide sequences alignment showed that the two genes are 74 % identical. On the amino acid level the sequence identity is 81 %. Similar results were achieved with CULC809_00602, the corresponding sequence of C. ulcerans 809, whereas comparison between the sequences of the two C. ulcerans strains among each other showed high identities of each 99 % on nucleotide and amino acid levels (see Table 1). The involvement of the DIP0733 corresponding protein in C. ulcerans pathogenesis is hardly understood so far and has to be elucidated by experimental studies. First analysis with a BR-AD22 mutant in the CULC22_00609 gene (ELHA3) showed no significant influence in the process of adhesion or invasion (Hacker et al., 2015a). Intracellular survival in THP-1 macrophages of this mutant strain is reduced compared to wild type C. ulcerans (unpublished data) and also the cytotoxic effect towards the eukaryotic cells is diminished (Hacker et al., 2015b).

Table 1: % Homology of nucleotide (bold) and deduced amino acid sequence (italic) of 67-72p protein from C. diphtheriae and C. ulcerans.

C. diphtheriae C. ulcerans C. ulcerans

NCTC 13129 BR-AD22 809 C. diphtheriae 100 / 100 74 / 81 74 / 80 NCTC 13129 C. ulcerans 100 / 100 99 / 99 BR-AD22 C. ulcerans 100 / 100 809

Very recently, a so far unknown protein in corynebacteria was detected in a human C. ulcerans isolate in a putative prophage integration site. This protein shares 42 % identity and 58 % similarity to the RhuM virulence factor of Salmonella enterica (Meinel et al., 2014). For Salmonella, it was shown that inactivation of this protein highly reduces virulence and leads to a decrease in mortality of C. elegans after S. enterica infection (Tenor et al., 2004). However, whether this protein also plays a role in virulence of C. ulcerans remains unknown up to now.

Hacker, E. 21

Introduction

2.2.3.5 Identification and characterization of virulence factors

As shown above, C. ulcerans may harbor a variety of putative virulence factors. However, these findings are based on sequence homologies to known virulence factors of other species only and experimental data concerning the influence of these factors on pathogenicity of C. ulcerans towards humans is scarce. To get deeper insight into the function of specific proteins in host colonization and mediation of disease, these factors have to be analyzed in more detail. One approach is to work with isolated proteins, which can be difficult to get in high amounts and pure form. Another approach is to construct specific mutant strains which either lack the gene of interest or overexpress it. As the bacterial background stays constant in this latter approach, it may be favorable in many cases. One also has to consider the interaction with the host and its reaction to the infection under different circumstances. This often plays a crucial role in the outcome of disease. Casadevall and Pirofski proposed to ask whether it is the interaction between a host and a microbe that damages the host instead of focusing on what microbes do or do not do (Casadevall & Pirofski, 2014). Both microbes and hosts are variable, so it is often not a specific virulence determinant of the pathogen that causes disease but more the interplay of more of these factors together with the host’s immune system. This may vary in different people, at different points in time and in a different context. Often, strategies and structures of pathogenic bacteria obviously involved in causing disease in humans are also found in commensal microbes. Hill therefore suggested to distinguish between these “niche factors” and true “virulence factors”. The latter ones are those which play a significant role in pathogenesis but are not found in commensal bacteria that occupy the same part of the body (Hill, 2012). Tauch and Burkovski transferred this concept to the highly diverse genus Corynebacterium and found that only diphtheria toxin and phospholipase D may be considered as virulence factors sensu strictu (Tauch & Burkovski, 2015). The ribosome binding protein RBP, the endoglycosidase of the EndoE family CPP, and the RhuM-like protein, which are found in single C. ulcerans strains, are regarded as candidate virulence factors. However, all others may better be classified as niche factors (Tauch & Burkovski, 2015). Table 2 shows an overview of all factors discussed here that might contribute to the pathogenicity of C. ulcerans and a proposed classification into real virulence factors, candidate virulence factors or niche factors.

Hacker, E. 22

Introduction

Table 2: Genes and factors involved in pathogenicity of C. ulcerans, their proposed function and classification.

Gene / factor Proposed function

True virulence factors

tox / DT Diphtheria-like toxin

pld / PLD Phospholipase D

Candidate virulence factors

rbp / RBP Shiga-like ribosome-binding protein

Endoglycosidase of the EndoE family cpp / EndoE (or CCP) (previously: Corynebacterial protease CP40)

nanH / NanH Sialidase precursor (neuraminidase H)

rhuM / RhuM RhuM-like protein

Niche factors

Mycolic acids

spaB / SpaB Surface-anchored protein (minor pilus subunit)

spaC / SpaC Surface-anchored protein (pilus tip protein)

spaD / SpaD Surface-anchored protein (major pilus subunit)

spaE / SpaE Surface-anchored protein (minor pilus subunit)

spaF / SpaF Surface-anchored protein (pilus tip protein)

cwlH / CwlH Cell wall-associated hydrolase (DIP1621 from C. diphtheriae)

rpfI / RpfI/RipA Resuscitation-promoting factor interacting (DIP1281 from C. diphtheriae) protein

Multifunctional virulence factor in CULC22_00609 C. diphtheriae, involved in hemagglutination, (DIP0733 from C. diphtheriae) adhesion, invasion and induction of apoptosis

Others (unclassified)

tspA Trypsin-like serine protease

vsp1 Venome serine protease

vsp2 Venome serine protease

Hacker, E. 23

Introduction

2.3 Interaction of C. ulcerans with host cells and cell components

Although C. ulcerans is recognized as an emerging pathogen that is obviously equipped with a set of potent virulence factors, molecular data on elucidating the mechanisms this bacterium uses for host colonization are scarce. A first in vivo experimental infection model was set up by Dias et al. in 2011. Mice were infected intravenously with different doses of non-toxigenic C. ulcerans wild type strains 809 and BR-AD22 and toxigenic CDC-KC279. The mortality rates were highest for the human clinical isolate 809, followed by the animal clinical isolate CDC- KC279, and the lowest mortality rate was observed for BR-AD22, isolated from an asymptomatic dog. They also analyzed the arthritogenicity, with 85 % of mice infected with CDC-KC279 and 71 % of mice infected with 809 showing clinical signs of arthritis, whereas BR-AD22 had no arthritogenic potential. Furthermore, viable bacteria were recoverable out of the blood of mice infected with 809 and CDC-KC279 until day 3, and until day 2 for mice infected with BR-AD22. Viable bacteria were also recovered from the kidney, liver, and spleen of infected mice until day 6 (CDC-KC279) and day 4 (809 and BR-AD22) and in joints until day 20 (all strains), whereas in lung, heart and brain no bacteria were detected. The experiments performed in this study showed high virulence of C. ulcerans independent of diphtheria toxin production (Dias et al., 2011). Due to the lack of a functional DT receptor, mice are insensitive to DT (Mitamura et al., 1995) and therefore interfering and detrimental effects of the toxin can be excluded in this model system. As C. ulcerans is considered to be a primarily respiratory pathogen, Mochizuki and coworkers established an intranasal infection model in mice to analyze pathogenesis of C. ulcerans in vivo (Mochizuki et al., 2015). In this study, the distribution of non-toxigenic (0102) and toxigenic (ATCC51799) C. ulcerans in infected mice, as well as serum titers and the lethality rates of the animals, were investigated. Mice were challenged with the bacteria intranasally and blood samples, nasal, oral, and tracheal swabs as well as lung, liver, kidney, and spleen homogenates, cecal content, and colorectal feces were tested for the presence of viable C. ulcerans for 14 days. Both toxigenic and non-toxigenic C. ulcerans strains showed lethal effects towards the mice within the time period investigated, whereas a non-pathogenic C. glutamicum control strain (ATCC 13032) did not. C. ulcerans 0102 was recovered from trachea throughout the observation period, from nasal and oral cavity until day 6 and 4, from the eyes until day 8 and from blood until day 4 after inoculation. Additionally, C. ulcerans spread in the lung, liver, kidney, and spleen. Viable bacteria were detected in these organs until day 12 after infection and lesions were observed in all organs except in the liver. However, an increase of serum anti-diphtheria toxin titer or anti-C. ulcerans IgG titer was not observed, which might be due to an insufficient time period for observation (Mochizuki et al., 2015). In

Hacker, E. 24

Introduction conclusion, this intranasal experimental infection system in mice provides an effective system for the analysis of C. ulcerans infection in vivo without the interfering effects of diphtheria toxin. Ott and coworkers (2012) evaluated invertebrate infection models for pathogenic corynebacteria. In this study, larvae of the greater wax moth Galleria mellonella and the nematode C. elegans were infected with two C. ulcerans wild type strains (809 and BR-AD22) and a PLD-deficient mutant strain (ELHA1) with a BR-AD22 wild type background. In the C. elegans model system, C. ulcerans was able to colonize the host and strain-specific differences in colonization were observed. While C. ulcerans 809 colonized the hindgut of the worm, C. ulcerans BR-AD22 spread over the whole body. The PLD mutation showed no effect in colonization behavior. In G. mellonella, C. ulcerans exhibited high degrees of melanization, which is linked to pathogen killing and phagocytosis in lepidopteran larvae (Nappi & Christensen, 2005; Senior et al., 2011). Furthermore, larvae became rapidly immotile and died within 48 h, whereas C. diphtheriae strains carried along in this study had a less severe effect. The authors assigned higher virulence of C. ulcerans compared to C. diphtheriae in this model system reflecting the broader host spectrum of C. ulcerans. In G. mellonella, lack of PLD caused a less severe response, indicating the virulence function of this protein in C. ulcerans (Ott et al., 2012). Furthermore, the affinity of three C. ulcerans strains isolated from asymptomatic dogs (BR-AD22, BR-AD41 and BR-AD61), one human C. ulcerans isolate (809), and strain CDC- KC279 to the human ECM/plasma proteins fibrinogen, fibronectin, and Type I collagen was investigated (Simpson-Louredo et al., 2014). Binding to these proteins may serve as a virulence property by mediating adhesion to host cells as a first step of colonization. In addition, the recognition by the innate immune system may be avoided favoring systemic dissemination and invasion of deeper tissues. The authors concluded that strain-dependent differences in the ability to bind to human fibrinogen, fibronectin, and collagen may contribute to variations in the pseudomembrane formation and virulence potential towards the host as it was previously shown for C. diphtheriae (Sabbadini et al., 2010; Simpson-Louredo et al., 2014). In the framework of this study, first data on cell culture infection experiments investigating the adhesion to and invasion into epithelial cells, as well as the interaction with human macrophages as important part of the innate immune system, were published (Hacker et al., 2015a, 2015b). The epithelial cell lines HeLa and Detroit562 were infected with the two non-toxigenic C. ulcerans wild type strains 809, isolated from a bronchoalveolar lavage (BAL) sample of an 80-year-old woman with fatal pulmonary infection and BR-AD22, isolated from a nasal swab of an asymptomatic dog, as well as two mutant strains ELHA1 and ELHA3 lacking the genes for PLD and CULC22_00609, the DIP0733 homolog, respectively. Host-pathogen interaction studies included the analysis by scanning electron microscopy, fluorescence microscopy, and counting of adherent and intracellular colony forming units. Both wild type

Hacker, E. 25

Introduction strains exhibited a high infectious potential as they were able to adhere to these epithelial cells in high amounts and also invade into them. Furthermore, as was shown by measurement of the transepithelial resistance, bacteria had a detrimental effect on epithelial cells. However, the two putatitve virulence factors showed no influence on colonization under the experimental conditions tested. The investigation of the interaction with human THP-1 macrophages revealed that C. ulcerans strains BR-AD22 and 809 are able to survive in macrophages for at least 20 hours, as was shown by count of colony forming units inside THP-1 cells after different time points. Fluorescence microscopy with a lysotracker dye revealed that uptake of C. ulcerans leads to delay of phagolysosome maturation. Upon infection, THP-1 cells produced high amounts of the cytokines IL-6 and G-CSF and the NF-κB signaling cascade was activated. Furthermore, detrimental effects of C. ulcerans on the macrophages were deduced from cytotoxicity measurements and FACS analyses. Here, both mutant strains ELHA1 and ELHA3 showed reduced cytotoxicity towards THP-1 cells, indicating multiple mechanisms of host- pathogen interaction.

2.4 Caenorhabditis elegans as infection model system for pathogenic corynebacteria

Caenorhabditis elegans is a free living-nematode that is widely used to study host-microbe interactions. It offers several advantages as model organism like a small body length (≤ 1.5 mm), a rapid generation time of only three days, a large brood size of about 300 progeny per adult and its transparent body, which permits direct microscopical examination of different biological processes (Sifri et al., 2005). Furthermore, the nematode can be easily genetically manipulated allowing to study the function of different host genes in host-pathogen interaction. C. elegans is a bacterivorous nematode and E. coli OP50 is used as normal laboratory feeding bacterium. By substitution of this food source through other bacteria, their (pathogenic) effect towards the nematode can be easily monitored by analyzing, for example, their motility, the pathogenic burden, or the survival of the worms contact to the bacteria of interest (Marsh & May, 2012). Sifri et al. (2005) summarized five distinctive mechanisms of worm killing: infection with (intestinal) colonization, persistent infection, invasion, biofilm formation, and toxin- mediated killing. Recently, Ott and coworkers (2012) provided a proof of principle study for application of C. elegans as infection model system for pathogenic corynebacteria. Corynebacteria used to infect the worms exhibited strain-specific colonization and killing of C. elegans, indicating differences in virulence between the strains (Ott et al., 2012). Later, C. elegans was applied to investigate the role of pili in pathogenesis of C. diphtheriae. It was shown that the toxigenic and piliated strain NCTC 13129 is able to rapidly

Hacker, E. 26

Introduction kill the worms, whereas non-toxigenic strains and toxigenic strains that lack pili are attenuated in nematode killing (Broadway et al., 2013). In a study of the multifunctional virulence factor DIP0733 of C. diphtheriae, the C. elegans model system was used to analyze its involvement in host interaction. A C. diphtheriae mutant strain lacking the corresponding gene showed significant attenuation in proliferation inside of the nematode, host colonization and killing (Antunes et al., 2015a). Moreover, another factor contributing to the pathogenic potential of C. diphtheriae, the tellurite-resistance protein CDCE8392_0813, was shown to play a role in killing of C. elegans (Santos et al., 2015). The natural nematopathogenic species Microbacterium and Leucobacter, both coryneform bacteria, display remarkable modes of pathogenicity in C. elegans. Microbacterium nematophilum and Leucobacter sp. can cause the Dar (deformed anal region) phenotype, characterized by a distinctive rectal swelling. This swelling is caused by bacteria that adhere to the rectal and post-anal cuticle of susceptible nematodes and leads to constipation and decelerated growth in infected worms. However, infection with Microbacterium is non-lethal. Leucobacter causes a more virulent version of this infection and may kill the worms after rectal invasion (Hodgkin et al., 2000, 2013). Additionally, Leucobacter induces aggregation of worms, designated as worm- or death-stars. The worms are irreversibly sticked together by their tails and adult worms trapped in these aggregates were immobilized and subsequently died. Larval worms in contrast were sometimes able to escape the worm-stars by undergoing autotomy (Hodgkin et al., 2013). Using the human and animal pathogens C. diphtheriae and C. ulcerans, as well as the non-pathogenic species C. glutamicum, Antunes et al. (2015b) showed that these coryneform bacteria can also induce star formation in C. elegans independent of their pathogenicity, although in a slower and less strong mode. Also the severe tail-swelling phenotype (Dar) was observed for all Corynebacterium species tested after two days of infection. However, significant species-specific differences were observed and the non- pathogenic C. glutamicum caused generally less Dar-formation than pathogenic corynebacteria. C. glutamicum also had only a minor influence on survival of C. elegans, whereas nematodes were killed due to infection with C. diphtheriae and C. ulcerans. The authors concluded that their results might indicate that distinct, but general properties of the genus members, like components of the unique corynebacterial cell envelope, are responsible for the symptoms rather than specific virulence factors (Antunes et al., 2015b). C. elegans has no cell-mediated immunity, but comprises complex mechanisms for disease resistance. These include avoidance behaviors, physical barriers, and the secretion and action of antimicrobial molecules like lectins and lysozymes (for review see Marsh and May, 2012). The avoidance behavior was shown to be used by C. elegans as defense mechanism for pathogenic corynebacteria. The nematodes avoided the two pathogenic species C. diphtheriae and C. ulcerans and preferred the non-pathogenic C. glutamicum.

Hacker, E. 27

Introduction

Furthermore, C. elegans showed aversive learning, as worms that were previously in contact with pathogenic corynebacteria avoided these at earlier time points compared to untrained worms (Antunes et al., 2015b).

Hacker, E. 28

Aim of the work

3 Aim of the work

C. ulcerans is an emerging pathogen transmitted by a zoonotic pathway to humans. While formerly infections in humans were rarely reported and mainly associated with direct contact to infected cattle, human infections caused by C. ulcerans appear to be increasing in various industrialized countries during the last decade. Nevertheless, to date, only little is known about factors besides the diphtheria toxin that contribute to virulence of this pathogen, and data elucidating mechanisms it uses for host colonization are lacking up to now. The aim of this study was to investigate the interaction of C. ulcerans with host cells regarding adhesion to and invasion into epithelial cell lines as well as the survival in phagocytic cells. To address this, infection assays were carried out of eukaryotic cell lines with two C. ulcerans wild type strains from human and animal sources. In addition, insertions mutagenesis of putative virulence factors in order to characterize their function in pathogen-host interaction was performed. By immuno-fluorescence microscopy and scanning electron microscopy, the colonization of the host cells was visualized. Furthermore, the reaction of the infected cells to bacterial infection was analyzed by determination of cytokine levels and NF-κB activation. Through FACS analysis, it was investigated whether and how cells die upon infection with C. ulcerans. Moreover, the nematode C. elegans was used as model system to study virulence properties of C. ulcerans in comparison with the closely related pathogen C. diphtheriae and non-pathogenic C. glutamicum.

Hacker, E. 29

Publications

4 Publications

4.1 Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources (Hacker et al., 2015a)

Hacker, E. 30

Publications

Microbiology (2015), 161, 1582–1591 DOI 10.1099/mic.0.000121

Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources Elena Hacker,1 Lisa Ott,1 Kristin Hasselt,23 Ana Luiza Mattos-Guaraldi,3 Andreas Tauch4 and Andreas Burkovski1

Correspondence 1Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg, Erlangen, Germany Andreas Burkovski 2BioCer Entwicklungs-GmbH, Bayreuth, Germany [email protected] 3Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil 4Centrum fu¨r Biotechnologie, Universita¨t Bielefeld, Bielefeld, Germany

Corynebacterium ulcerans is an emerging pathogen transmitted by a zoonotic pathway to humans. Despite rising numbers of infections and potentially fatal outcomes, data on the colonization of the human host are lacking up to now. In this study, adhesion of two C. ulcerans isolates to human epithelial cells, invasion of host cells and the function of two putative virulence factors with respect to these processes were investigated. C. ulcerans strains BR-AD22 and 809 were able to adhere to Detroit562 and HeLa cells, and invade these epithelial cell lines with a rate comparable to other pathogens as shown by scanning electron microscopy, fluorescence microscopy and replication assays. Infection led to detrimental effects on the cells as deduced from measurements of transepithelial resistance. Mutant strains of putative virulence factors phospholipase D and DIP0733 homologue CULC22_00609 generated in this study showed no influence on colonization under the experimental Received 30 April 2015 conditions tested. The data presented here indicate a high infectious potential of this emerging Accepted 9 June 2015 pathogen.

INTRODUCTION C. ulcerans is widely recognized as a commensal micro- organism in domestic and wild animals. The range of Corynebacterium ulcerans is a pathogenic member of the mammal hosts observed for C. ulcerans is extremely broad genus Corynebacterium, which is part of the family Coryne- and includes cattle, goats, pigs, wild boars, dogs, cats, bacteriaceae, the order Actinomycetales and the phylum Actinobacteria (Tauch & Sandbote, 2014). Within the phy- ground squirrels, otters, camels, monkeys, whales, water lum Actinobacteria, the genera Corynebacterium, Nocardia rats and others (Pappenheimer, 1977; Olson et al., 1988; and Mycobacterium form a monophyletic branch, the Hommez et al., 1999; Bergin et al., 2000; Tejedor et al., CMN group, based on their unusual cell envelope compo- 2000; Foster et al., 2002; Morris et al., 2005; Seto et al., sition (Burkovski, 2013). C. ulcerans was first described by 2008; Hogg et al., 2009; Schuhegger et al., 2009; Sykes Gilbert & Stewart (1927), who isolated the bacterium from et al., 2010; Contzen et al., 2011; Hirai-Yuki et al., 2013; the throat of a patient with respiratory diphtheria-like ill- Eisenberg et al., 2015). Human infections are rare and ness. When lysogenized by a tox gene-carrying coryneph- have traditionally been reported amongst rural populations age, C. ulcerans can – like Corynebacterium diphtheriae – with direct contact to domestic livestock or consumption of produce diphtheria toxin (Sangal & Hoskisson, 2014) and raw milk and other unpasteurized dairy products (Bostock during the past decade diphtheria-like infections with toxi- et al., 1984; Hart, 1984). However, during the last decade, genic C. ulcerans have outnumbered those caused by toxi- human infections associated with C. ulcerans, e.g. genic C. diphtheriae in many industrialized countries diphtheria, severe necrotizing fasciitis and skin ulcers, (Zakikhany & Efstratiou, 2012). appear to be increasing in various countries and can most often be ascribed to zoonotic transmission (Mattos-Guar- 3Present address: Lisanto, Regensburg, Germany. aldi et al., 2008; Meinel et al., 2014, 2015; Sangal et al., 2014).

Abbreviations: PLD, phospholipase D; SEM, scanning electron micro- In 2011, two non-toxigenic C. ulcerans strains from the scopy. metropolitan area of Rio de Janeiro, Brazil, were sequenced:

1582 000121 G 2015 The Authors Printed in Great Britain

Hacker, E. 31 Publications

C. ulcerans–host interaction

BR-AD22, isolated from an asymptomatic dog, and 809, iso- Bacterial strains, cell lines and plasmids used in this study are listed in lated from an 80-year-old woman with fatal pulmonary Table 1. infection (Mattos-Guaraldi et al., 2008; Dias et al., 2010; Molecular biology techniques and construction of mutant Trost et al., 2011). Based on these genome sequences and strains. Standard techniques were used for plasmid isolation, trans- comparative genomics approaches, a number of putative formation of E. coli and cloning (Sambrook et al., 1989). For chro- virulence factors were annotated; however, functional data mosomal disruption of the C. ulcerans BR-AD22 genes pld and * were scarce. Up to now, data on fibrinogen, fibronectin CULC22_00609, internal DNA fragments of 500 bp were amplified by PCR using chromosomal DNA of strain BR-AD22 as template and and collagen binding, antimicrobial profiles, and arthrito- the following primers: pld-XmaI-s (59-CGCGCCCGGGACCTGGC- genic potential of isolates have been published (Dias et al., TCGATATTAAGAATCCTGA-39)/pld-XmaI-as (59-CGCGCCCGGG- 2011; Simpson-Louredo et al., 2014). In the study presented CCAAAGATCATTCCGTCTAC-39) for amplification of the pld gene here, C. ulcerans 809 and BR-AD22 were characterized for fragment and CULC22_00609-XmaI-s (59-CGCGCCCGGGGTATC- 9 9 the first time to the best of our knowledge with respect to TGGCCGCTCATGG-3 )/CULC22_00609-XmaI-as (5 -CGCGCCCG- GGGACTGCAAATCCGACACTG-39) for amplification of the adhesion to epithelial cells, invasion of epithelial cells and CULC22_00609 gene fragment. Using the XmaI sites introduced via the function of two putative virulence factors in these initial the PCR primers (shown in italic), the DNA fragment was ligated to processes of infection. XmaI-restricted and dephosphorylated pK18mob DNA (Scha¨fer et al., 1994). The resulting plasmids pK18mob-pld9 and pK18mob-006099 were amplified in E. coli DH5aMCR. Unmethylated plasmid isolated METHODS from this E. coli strain (1 mg) was used to transform C. ulcerans BR- AD22 using a GenePulser II (Bio-Rad). Electroporated cells were Bacterial strains and growth. Non-toxigenic C. ulcerans strains BR- added to 1 ml HI broth and incubated for 2 h at 37 uC under shaking. AD22 and 809 were grown in Heart Infusion (HI) broth at 37 uC, An appropriate volume of culture was plated on medium containing Escherichia coli DH5aMCR and Salmonella enterica serovar Typhi- kanamycin. As pK18mob cannot be replicated in C. ulcerans, kana- murium NCTC 12023 were grown in Luria Broth (LB) (Sambrook mycin-resistant C. ulcerans carried the vector integrated by homolo- et al., 1989) at 37 uC. If appropriate, kanamycin (60 mgml–1 for E. coli; gous recombination in the respective chromosomal genes, which were 50 mgml–1 for C. ulcerans) or chloramphenicol (25 mgml–1) was added. designated ELHA1 ( pld) and ELHA3 (CULC22_00609). The insertion

Table 1. Strains, cell lines and plasmids used in this study

PLD, phospholipase D.

Strain, cell line or plasmid Description/genotype Reference/source C. ulcerans BR-AD22 Isolated from an asymptomatic dog, non-toxigenic (tox –) Mattos-Guaraldi et al. (2008) 809 Isolated from an 80-year-old woman with fatal pulmonary Dias et al. (2010) infection, non-toxigenic (tox –) ELHA1 BR-AD22 pld : : pK18mob-pld9, PLD-deficient mutant Ott et al. (2012) ELHA3 BR-AD22 CULC22_00609 : : pK18mob-CULC22_006099 This study E. coli DH5aMCR endA1 supE44 thi-1 l – recA1 gyrA96 relA1 Grant et al. (1990) deoR D(lacZYA–argF) U196 w80 DlacZDM15 mcrA D(mmrhsdRMS mcrBC) S. Typhimurium NCTC 12023 WT identical to ATCC 14028 National Collection of Type Cultures (Colindale, UK) Cell lines Detroit562 Human hypopharyngeal carcinoma cells Peterson et al. (1968) HeLa Human cervical carcinoma cells Gey et al. (1952); Scherer et al. (1953) Plasmids pK18mob ori pUC mob,KmR Scha¨fer et al. (1994) pK18mob-006099 pK18mob carrying a 540 bp internal fragment This study of CULC22_00609 for gene disruption pK18mob-pld9 pK18mob carrying a 476 bp internal fragment This study of pld for gene disruption R pXMJ19 ori colE1 oricg ptac,Cm Jakoby et al. (1999) R pXMJ19-00609 ori colE1 oricg ptac CULC22_00609, Cm This study R pXMJ19-pld ori colE1 oricg ptac pld,Cm This study pEPR1-p45gfp P45 gfpuv rep per T1 T2, KmR Knoppova´ et al. (2007)

http://mic.sgmjournals.org 1583

Hacker, E. 32 Publications

E. Hacker and others of the plasmid into the chromosome of C. ulcerans and consequently For invasion assays, after 90 min of infection, cells were not detached, the disruption of the corresponding genes was confirmed by Southern but incubated for 2 h in DMEM (500 ml per well), containing 100 mg blotting using specific DNA probes (data not shown). gentamicin ml–1 to kill remaining extracellular bacteria. After this incubation, the cell layers were washed once with PBS, detached, lysed For overexpression of pld and CULC22_00609 in C. ulcerans BR- and plated as described above, and the c.f.u. count was determined. AD22, the complete genes were amplified by PCR using chromosomal The ratio of bacteria used for infection (number of colonies on DNA of strain BR-AD22 as template and the following primers: OE- 9 9 inoculum plates) and bacteria in the lysate (number of colonies on the pld-PstI-s (5 -CGCGCTGCAGTGTAGAGGGATCAACGATG-3 )/OE- lysate plates) multiplied by 100 gave the invasion efficiency in percent. pld-EcoRI-as (59-CGCGGAATTCTTGGCGGTGTCTAAACTCAC-39) 9 for amplification of the pld gene and OE-00609-XbaI-s (5 -CGCG- Immunofluorescence microscopy. Detroit562 and HeLa cells were 9 9 TCTAGAGGAGTTCGAGTTGGCGAACG-3 )/OE-00609-XmaI-as (5 - seeded on round coverslips in 24-well plates at a density of 5|104 9 CGCGCCCGGGGCACTAGGGATCTGATTAGC-3 ) for amplification and after 24 h infected at m.o.i. 50 with GFP-expressing C. ulcerans of the CULC22_00609 gene. Using the restriction sites introduced via for 90 min. To analyse adhesion, cells were washed six (Detroit562) or the PCR primers (shown in italic), the DNA fragments were ligated to three (HeLa) times and then fixed with 4 % paraformaldehyde in PBS likewise restricted and dephosphorylated pXMJ19 DNA (Jakoby et al., for 20 min at 37 uCandstoredinPBSat4uC until staining. For 1999). The resulting plasmids pXMJ19-pld and pXMJ19-00609 were analysis of invasion, cells were washed three times after 90 min of a amplified in E. coli DH5 MCR and used for electroporation of infection and then further incubated for 2 h with DMEM containing C. ulcerans BR-AD22 as described above. Positive clones were selected 100 mgml–1 gentamicin,thenwashedagainandfixedasdescribed on HI agar containing chloramphenicol. RNA hybridization exper- above. iments with specific RNA probes proved increased expression levels of the corresponding genes in the resulting C. ulcerans strains (data not Fixed cells were stained with 30 ml Alexa Fluor 647 Phalloidin (Life shown). Technologies) diluted 1 : 200 in an Image-iT FX signal enhancer (Life Technologies) for 45 min in the dark. Subsequently, cells were washed For fluorescence microscopy, electrocompetent C. ulcerans were twice with PBS, dried for 5–10 min at 37 uC and then mounted on transformed with gfp-carrying plasmid pEPR1-p45gfp, which leads to glass slides using ProLong Gold Antifade Mountant with DAPI (Life constitutive expression of GFP controlled by the p45 corynephage Technologies) and stored at 4 uC. Visualization was carried out on the promoter (Knoppova´ et al., 2007). confocal laser scanning microscope Leica SP5 II and analysed with the LAS software suite. Cell cultures. Detroit562 were cultivated in Dulbecco’s modified Eagle’s medium (DMEM), high glucose with L-glutamine and Measurement of transepithelial resistance. Detroit562 cells were m sodium pyruvate (PAA Laboratories), supplemented with 120 g seeded in transwells (12 mm, 0.4 mm, polyester membrane, 12-well –1 m –1 penicillin ml , 120 g streptomycin ml and 10 % heat-inactivated plate; Corning Costar) at a density of 1|105 cells per well and cul- FCS in a CO2 incubator. Cells were passaged at a ratio of 1 : 10 twice tivated in DMEM (high glucose, 10 % FCS, 2 mM glutamine) for per week. HeLa cells were cultured in DMEM, high glucose with L- 14 days until they built a transepithelial resistance of at least m glutamine (PAA Laboratories) supplemented with 100 ggentami- 800 V cm–2. Bacteria were subcultured (OD 0.1 from overnight –1 m –1 600 cin ml ,12 g ciprofloxacin ml and 10 % heat-inactivated FCS in cultures) in 20 ml HI broth (C. ulcerans) or LB medium (S. Typhi- aCO2 incubator. Cells were passaged at a ratio of 1 : 10 twice per murium) until OD600 0.4–0.6 was reached, harvested by centrifu- week. gation and the pellet was resuspended in 500 ml PBS. An OD600 of 5 was adjusted in 500 ml PBS and 100 ml of this suspension was used to Scanning electron microscopy (SEM). Infected monolayer cultures infect one well. Measurements of transepithelial resistance of were first fixed with 0.1 % glutaraldehyde, 2 % paraformaldehyde, Detroit562 cells after infection were carried out with an EVOM2 u 5 % sucrose in 0.2 M sodium cacodylate solution for 1 h at 37 C. voltohmmeter (World Precision Instruments) every 60 min. After Subsequently, samples were fixed with 0.3 % glutaraldehyde, 3 % 180 min the supernatant of infected Detroit562 cells was removed and paraformaldehyde in 0.2 M sodium cacodylate solution for 1 h at the cells were incubated in fresh DMEM overnight to avoid detri- u 37 C. Samples were subsequently dehydrated using a graded series of mental effects of excessive bacterial growth. acetone, critical point dried and gold sputtered, and examined using a FEI Quanta 200 scanning electron microscope.

Adhesion and invasion assays. Detroit562 and HeLa cells were RESULTS seeded in 24-well plates (Nunc) at a density of 1|105 at 24 h prior to infection. Bacteria were inoculated to OD600 0.1 from overnight SEM cultures and grown in HI broth to OD600 0.4–0.6. Subsequently, the bacteria were harvested by centrifugation and cell density was As a first approach, binding of C. ulcerans to epithelial cells was adjusted to OD600 0.2. A master mix of the inoculum with m.o.i. 50 characterized by SEM. For this purpose, adhesion of strains was prepared in DMEM and 500 ml per well was used to infect the BR-AD22 and 809 to Detroit562 and HeLa cells was analysed. cells. The plates were centrifuged for 5 min at 500 r.p.m. to syn- Independent of the C. ulcerans strain applied and the tested chronize infection and subsequently incubated for 90 min. The cells were washed with PBS three (HeLa cells) or six (Detroit562 cells) epithelial cell line, a localized, clustered adhesion pattern was times, detached with 500 ml trypsin solution (0.12 % trypsin, 0.01 % found (Fig. 1). As tested by microscopic inspection of the EDTA in PBS) per well (5 min, 37 uC, 5 % CO2, 90 % humidity) and inoculum and determination of c.f.u. before and after wash- lysed with 0.025 % Tween 20 for 5 min at 37 uC. Serial dilutions of ing steps, these clusters were not the result of bacterial aggre- the lysates and the inoculi in 1| PBS were plated on Columbia agar gation (data not shown), but hinted at the presence of a with sheep blood (Oxoid) using an Eddy Jet Version 1.22 (IUL u limited number of specific cell surface receptor sites. Instruments). After incubation at 37 C for 2 days, the c.f.u. count A similar pattern of adhesion of bacteria to the cell surface was determined. The ratio of bacteria used for infection (number of colonies on inoculum plates) and bacteria in the lysate (number of was observed for different C. diphtheriae strains (Hirata colonies on the lysate plates) multiplied by 100 gave the adhesion et al., 2004). In the case of Detroit562 cells, in rare cases efficiency in percent. an insertion of bacteria in the membrane was observed,

1584 Microbiology 161

Hacker, E. 33 Publications

C. ulcerans–host interaction

suggesting invasion of bacteria into epithelial cells, which are constant up to 21 h, infection of Detroit562 monolayers typically non-phagocytotic (Fig. 1c). with S. Typhimurium caused a fast breakdown of transe- pithelial resistance within 2 h. Compared with S. Typhi- murium, effects on monolayers caused by C. ulcerans Fluorescence microscopy strains BR-AD22 and 809 were considerably slower. Fluorescence microscopy was used as an independent Whilst in the case of infection with BR-AD22 w11 h was method to investigate interaction of C. ulcerans strains needed for breakdown of transepithelial resistance, in the with epithelial cells. For this purpose, both C. ulcerans case of strain 809 a low basal resistance level had already strains were marked by GFP expression, nuclei were stained been reached at 4 h (Fig. 4). The strong detrimental effect with DAPI and the cytoskeleton with Alexa Fluor 647 of 809 might be due to the ribosome-binding protein with phalloidin (Fig. 2). Fluorescence microscopy images revea- similarity to Shiga-like toxin encoded by this strain. led adhesion of the two strains to the surface of Detroit562 and HeLa cells. Internalization of bacteria in the two cell Influence of putative C. ulcerans virulence lines was clearly demonstrated by orthogonal z-stacks. Typi- factors cal V-shaped bacteria due to snapping division indicated growth both on the surface and inside the epithelial cells. Despite increasing numbers of infections and sometimes fatal cases, virulence factors of C. ulcerans are widely uncharacterized. However, based especially on homology Quantitative analysis of adhesion and to closely related pathogenic corynebacteria, a number of invasion C. ulcerans proteins have been annotated as putative viru- For a quantitative analysis, adhesion and invasion of lence factors (Trost et al., 2011). Phospholipase D (PLD) is C. ulcerans BR-AD22 and 809 were analysed by determi- the most important virulence factor of Corynebacterium nation of c.f.u. Whilst adhering bacteria were directly pseudotuberculosis. This so-called ovis toxin is involved in counted after 90 min, a gentamicin protection assay was dissemination of this animal pathogen from the initial used to analyse the rate of internalization (Fig. 3). The two site of infection to other parts of the body. Only two strains were able to adhere to and invade epithelial cells, other bacterial species are known to produce a PLD protein although to different degrees. In the case of adhesion to similar to that from C. pseudotuberculosis: Arcanobacterium Detroit562 cells, a fivefold higher rate was determined for haemolyticum and C. ulcerans (for comparison, see Table 2). 809 compared with BR-AD22 (187+62 versus 31+15 %; Genomic organization of the pld gene is similar in Fig. 3a), whilst no differences were observed for adhesion C. ulcerans and C. pseudotuberculosis with a gene encoding to HeLa cells (179+72 versus 186+76 %; Fig. 3c). In the a transposase protein and a fagC gene flanking pld, whilst case of Detroit562 infection with strain 809 and the two the A. haemolyticum pld is flanked by genes encoding a HeLa infection approaches, strong growth of bacteria was hypothetical protein on the one side and three tRNAs on indicated by c.f.u. exceeding the m.o.i. initially applied. the other. Invasion rates determined were identical for the two A BR-AD22 strain lacking PLD was generated and desig- C. ulcerans strains in the case of Detroit562 cells (0.4+ nated ELHA1. Compared with the WT, no significant + 0.1 for BR-AD22 versus 0.3+0.1 % for 809; Fig. 3b), changes in adhesion (BR-AD22 36.1 15.6 versus ELHA1 46.5+16.8 %) or invasion (BR-AD22 0.09+0.04 versus whilst different invasion rates for were found for HeLa + infections. In this case, BR-AD22 showed an invasion ELHA1 0.1 0.04 %) rate were observed when Detroit562 rate of 2.2+0.5 compared with 5.8+0.7 % for strain 809 cells were infected. Overexpression of the protein in (Fig. 3d). strain BR-AD22 pXMJ19-pld also did not influence the adhesion (29.7+8.8 %) or invasion (0.08+0.04 %) rate. Similar results were obtained with HeLa cells (data Transepithelial resistance not shown). Epithelial cells infected with C. ulcerans appeared to be Another example of a putative virulence factor encoded damaged sometimes; however, whether this was caused in the C. ulcerans genome is CULC22_00609, a homologue by handling effects or a reaction specific for C. ulcerans of C. diphtheriae DIP0733, which was found recently contact remained unclear. Some pathogens, such as to be involved in adhesion to and invasion of epithelial S. Typhimurium, can cause severe damage on cell mem- cells (Antunes et al., 2015). The corresponding protein com- branes and due to the resulting loss of cell integrity, the prises 81 % identity to DIP0733 from strain C. diphtheriae transepithelial resistance of monolayers is dramatically NCTC 13129. A CULC22_00609 gene disruption was gener- reduced (e.g. see Gerlach et al., 2008; Ott et al., 2010). ated in C. ulcerans BR-AD22, leading to strain ELHA3, In this study, we used S. Typhimurium NCTC 12023 as and analysed with respect to adhesion and invasion. Disrup- a positive control to test the influence of C. ulcerans BR- tion of the DIP0733 homologue CULC22_00609 had AD22 and 809 on transepithelial resistance. Whilst the no effect with respect to adhesion (BR-AD22 33.0+11.7 negative control without bacteria showed an even, increas- versus ELHA3 17.5+7.1 %) to Detroit562 cells as well ing transepithelial resistance for 4 h, which later stayed as with respect to invasion of this cell line (BR-AD22

http://mic.sgmjournals.org 1585

Hacker, E. 34 Publications

E. Hacker and others

C. ulcerans BR-AD22 C. ulcerans 809 (a) (b)

(c) (d) Detroit562

(e) (f)

HeLa (g) (h)

Fig. 1. SEM of C. ulcerans and epithelial cells. Interaction of C. ulcerans strains with Detroit562 and HeLa cells viewed by SEM at 90 min post-infection. Infected monolayer cultures were fixed, dehydrated, sputtered with gold and examined using a FEI Quanta 200 scanning electron microscope. (a–h) Detroit562 (a–d) and HeLa (e–h) cells infected with BR-AD22 (a, c, e, g) and 809 (b, d, f, h). Putative invasion of bacteria is marked by a white circle in (c). Scale bars represent 10 mm.

1586 Microbiology 161

Hacker, E. 35 Publications

C. ulcerans–host interaction

C. ulcerans BR-AD22 C. ulcerans 809 (a) (b) Detroit562

(c) (d) HeLa

Fig. 2. Fluorescence microscopy of C. ulcerans and epithelial cells. (a–d) Detroit562 (a, b) and HeLa (c, d) cells were infected with C. ulcerans BR-AD22 pEPR1p45gfp (a, c) and C. ulcerans 809 pEPR1p45gfp (b, d). Nuclei were stained with DAPI and the cytoskeleton with Alexa Fluor 647 phalloidin, and z-stack micrographs were taken using a Leica SP5 II confocal laser scanning microscope.

0.07+0.02 % versus ELHA3 0.05+0.02 %). Overexpression mechanisms of host–pathogen interaction and specific viru- also showed no significant effects with an adhesion rate lence factors of C. ulcerans have been widely uncharacterized of strain BR-AD22 pXMJ19-00609 of 19.1+8.0 % and an until now. + invasion rate of 0.08 0.04 %. Again, with HeLa cells also, The two C. ulcerans strains investigated in this study were no influence of the putative virulence factor was found able to bind to epithelial cells – most likely at specific (data not shown). receptor sites – and multiply during in vitro infection. Fur- thermore, the bacteria were able to invade the cytoplasm of DISCUSSION epithelial cells. Compared with adhesion rates found for different C. diphtheriae strains (for review, see Ott & C. ulcerans is an emerging pathogen, which is transmitted Burkovski, 2014) and invasion rates determined (e.g. Ott by a zoonotic pathway to the human host (Meinel et al., et al., 2013), C. ulcerans rates observed here were at least 2014, 2015). A wide variety of mammals can be infected two- to fivefold higher, indicating the high infectious and provide a reservoir for the pathogen. In recent years potential of this pathogen. This is further supported by det- the cases of respiratory diphtheria caused by C. diphtheriae rimental effects on epithelial cells as shown by the break- were outnumbered by C. ulcerans infections in Western down of transepithelial resistance of cell monolayers after Europe (Bonmarin et al., 2009; Wagner et al., 2010; infection with C. ulcerans BR-AD22 and 809. Again, also Zakikhany & Efstratiou, 2012). Despite increasing numbers in this case, C. ulcerans showed strong effects, whilst an of infections and the occurrence of fatal cases, basic influence on transepithelial resistance was not observed

http://mic.sgmjournals.org 1587

Hacker, E. 36 Publications

E. Hacker and others

(a) (b) 300 0.6

250 0.5

200 0.4

150 0.3 Invasion (%) Adhesion (%) 100 0.2

50 0.1

0 0 BR-AD22 809 BR-AD22 809 (c) (d) 300 7

250 6

5 200 4 150 3 Invasion (%) Adhesion (%) 100 2 50 1

0 0 BR-AD22 809 BR-AD22 809

Fig. 3. Quantitative analysis of C. ulcerans adhesion and invasion rates to epithelial cells. Epithelial cells were infected with C. ulcerans at m.o.i. 50 for 90 min. (a–d) For analysis of adhesion rates (a, c), the cells were washed six times, detached, lysed and dilutions of the lysate were plated on agar plates to determine the number of attached c.f.u.; for analysis of intra- cellular c.f.u. (b, d), cells were further incubated for 120 min with medium containing 100 mg gentamicin ml–1 to kill extracellu- lar bacteria, then washed, detached, lysed and plated on agar plates to recover intracellular c.f.u.. (a, b) Detroit562 and (c, d) HeLa. Data shown are mean¡SD values of at least three independent biological replicates. for C. diphtheriae strains (Ott et al., 2010). The observed (Trost et al., 2011; Meinel et al., 2014, 2015; Antunes internalization into epithelial cells might be an effective et al., 2015). mechanism for immune evasion of C. ulcerans, supporting Two putative virulence factors of C. ulcerans were tested the establishment and progress of infections. in this study: the DIP0733 homologue CULC22_00609 Both strain 809, from a fatal case of pneumonia, and BR- and PLD. In C. diphtheriae CDC-E8392, DIP0733 is import- AD22, isolated from an asymptomatic dog, were able to ant for adhesion and invasion, and a mutation of the corres- colonize human epithelial cells, which might be explained ponding gene influenced these processes negatively (Antunes by a close taxonomic relationship as found by Meinel et al.,2015).InC. ulcerans, an effect on adhesion and inva- et al. (2014). Nevertheless, strain- and cell line-specific diff- sion by CULC22_00609, the corresponding homologue of erences observed hint at the existence of more than one DIP0733, was not observed under the experimental con- mechanism for adhesion and invasion. In line with this ditions tested. The reason for this result is unclear. Possible idea, initial sequencing of strains 809 and BR-AD22 explanations might be species- and strain-specific differences (Trost et al., 2011) as well as next-generation sequencing with respect to adhesion mechanisms, as also observed for of nine C. ulcerans isolates revealed a number of different 809 and BR-AD22 in this study. However, as we were not virulence factors which might influence colonization by able to generate a CULC22_00609 deletion – or any other different isolates of this pathogen. Proteins found included kind of mutation – in the strain 809 background, this PLD, neuraminidase H, endoglycosidase E, SpaDEF-type hypothesis could not be verified. adhesive pili, diphtheria toxin, a ribosome-binding protein Whilst studies on the influence of PLD on colonization of with similarity to Shiga-like toxin, a haemagglutinin pro- epithelial cell lines are lacking for C. pseudotuberculosis, tein, a Fic toxin homologue and a RhuM homologue interaction of a PLD mutant with HeLa cells was tested

1588 Microbiology 161

Hacker, E. 37 Publications

C. ulcerans–host interaction

1200 In summary, the emerging pathogen C. ulcerans includes a broad and varying set of virulence factors (Trost et al., 2011). Furthermore, by acquisition of new genes, this 1000 repertoire can change quickly (Meinel et al., 2014), sup-

) porting the necessity of constant surveillance of this patho- 2 – gen (Both et al., 2015). The contribution of putative 800 virulence proteins identified so far and of other com- ponents of the cell, like glycolipids, which play an import- ant role in virulence of C. diphtheriae, Mycobacterium 600 tuberculosis and Rhodococcus equi (Indrigo et al., 2003; Axelrod et al., 2008; Moreira et al., 2008; Sydor et al., 2013), has to be elucidated in the future at the molecular 400 level by genetic experiments with defined mutants. Transepithelial resistance ( Ω cm Transepithelial 200 ACKNOWLEDGEMENTS

The financial support of L. O. by the Friedrich-Alexander-Universita¨t 0 Erlangen-Nu¨rnberg in frame of the FFL program ‘‘Fo¨rderung der 0 5 10 15 20 Chancengleichheit fu¨r Frauen in Forschung und Lehre’’ is gratefully Time (h post-infection) acknowledged.

Fig. 4. Influence of C. ulcerans on transepithelial resistance. REFERENCES Polarized Detroit562 monolayers were grown on transwells and infected with C. ulcerans strains (&) BR-AD22 and (m) 809. For Antunes, C. A., Sanches dos Santos, L., Hacker, E., Ko¨ hler, S., Bo¨ sl, control, cells were incubated without (D) bacteria and (%) K., Ott, L., de Luna, M., Hirata, R., Jr, Azevedo, V. A. & other authors S. Typhimurium. Experiments were carried out independently at (2015). Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae. Microbiology 161, 639–647. least twice and typical results are shown. Axelrod, S., Oschkinat, H., Enders, J., Schlegel, B., Brinkmann, V., Kaufmann, S. H., Haas, A. & Schaible, U. E. (2008). Delay of phagosome maturation by a mycobacterial lipid is reversed by nitric previously in A. haemolyticum. Adhesion was diminished oxide. Cell Microbiol 10, 1530–1545. and invasion increased compared with the WT, due to Bergin, I. L., Chien, C. C., Marini, R. P. & Fox, J. G. (2000). Isolation the detrimental effect of PLD secretion in the latter species and characterization of Corynebacterium ulcerans from cephalic (Lucas et al., 2010). In a recent study, PLD deficiency in the implants in macaques. Comp Med 50, 530–535. C. ulcerans background had a positive influence on survival Bonmarin, I., Guiso, N., Le Fle` che-Mate´ os, A., Patey, O., Grimont Patrick, A. D. & Levy-Bruhl, D. (2009). Diphtheria: a zoonotic of wax moth larvae. As in this study, an influence on disease in France? Vaccine 27, 4196–4200. adhesion was not observed (Ott et al., 2012). These results Bostock, A. D., Gilbert, F. R., Lewis, D. & Smith, D. C. (1984). might indicate that C. ulcerans PLD is more important for Corynebacterium ulcerans infection associated with untreated milk. dissemination within the host than for adhesion to host J Infect 9, 286–288. cells – an idea that is in principle also in accordance with Both, L., Collins, S., de Zoysa, A., White, J., Mandal, S. & Efstratiou, A. the observation made in A. haemolyticum (Lucas et al., (2015). Molecular and epidemiological review of toxigenic diphtheria infections in England between 2007 and 2013. J Clin Microbiol 53, 2010). 567–572. Burkovski, A. (2013). Cell envelope of corynebacteria: structure and influence on pathogenicity. ISRN Microbiol 2013, 935736. Table 2. Sequence homology of pld genes and correspond- Contzen, M., Sting, R., Blazey, B. & Rau, J. (2011). Corynebacterium ing PLD proteins of C. ulcerans, C. pseudotuberculosis and ulcerans from diseased wild boars. Zoonoses Public Health 58, A. haemolyticum compared with CULC22_00038, the pld 479–488. gene of C. ulcerans BR-AD22 Dias, A. A., Silva, F. C., Jr, Pereira, G. A., Souza, M. C., Camello, T. C., Damasceno, J. A., Pacheco, L. G., Miyoshi, A., Azevedo, V. A. & other Strain Nucleotide Amino acid authors (2010). Corynebacterium ulcerans isolated from an sequence sequence asymptomatic dog kept in an animal shelter in the metropolitan identity (%) identity (%) area of Rio de Janeiro, Brazil. Vector Borne Zoonotic Dis 10, 743–748. Dias, A. A., Silva, F. C., Jr, Santos, L. S., Ribeiro-Carvalho, M. M., C. ulcerans 809 97 97 Sabbadini, P. S., Santos, C. S., Filardy, A. A., Myioshi, A., Azevedo, C. pseudotuberculosis FRC41 81 87 V. A. & other authors (2011). Strain-dependent arthritogenic potential A. haemolyticum DSM 20595 66 63 of the zoonotic pathogen Corynebacterium ulcerans. Vet Microbiol 153, 323–331. http://mic.sgmjournals.org 1589

Hacker, E. 38 Publications

E. Hacker and others

Eisenberg, T., Mauder, N., Contzen, M., Rau, J., Ewers, C., Schlez, K., bacterium ulcerans isolates reveals zoonotic transmission and a Althoff, G., Schauerte, N., Geiger, C. & other authors (2015). novel putative diphtheria toxin-encoding pathogenicity island. Outbreak with clonally related isolates of Corynebacterium ulcerans Genome Med 6, 113. in a group of water rats. BMC Microbiol 15, 42. Meinel, D. M., Konrad, R., Berger, A., Ko¨ nig, C., Schmidt-Wieland, T., Foster, G., Patterson, T., Howie, F., Simpson, V., Davison, N., Hogardt, M., Bischoff, H., Ackermann, N., Ho¨ rmansdorfer, S. & other Efstratiou, A. & Lai, S. (2002). Corynebacterium ulcerans in free- authors (2015). Zoonotic transmission of toxigenic Corynebacterium ranging otters. Vet Rec 150, 524. ulcerans strain, Germany, 2012. Emerg Infect Dis 21, 356–358. Gerlach, R. G., Cla´ udio, N., Rohde, M., Ja¨ ckel, D., Wagner, C. & Moreira, L. O., Mattos-Guaraldi, A. L. & Andrade, A. F. (2008). Novel Hensel, M. (2008). Cooperation of Salmonella pathogenicity islands lipoarabinomannan-like lipoglycan (CdiLAM) contributes to the 1 and 4 is required to breach epithelial barriers. Cell Microbiol 10, adherence of Corynebacterium diphtheriae to epithelial cells. Arch 2364–2376. Microbiol 190, 521–530. Gey, G. O., Coffman, W. D. & Kubicek, M. T. (1952). Tissue culture Morris, W. E., Uzal, F. A. & Cipolla, A. L. (2005). Pyogranulomatous studies of the proliferative capacity of cervical carcinoma and meningoencephalitis in a goat due to Corynebacterium ulcerans. Vet normal epithelium. Cancer Res 12, 264–265. Rec 156, 317–318. Gilbert, R. & Stewart, F. C. (1927). Corynebacterium ulcerans;a Olson, M. E., Goemans, I., Bolingbroke, D. & Lundberg, S. (1988). pathogenic microorganism resembling Corynebacterium diphtheriae. Gangrenous dermatitis caused by Corynebacterium ulcerans in J Lab Clin Med 12, 756–761. Richardson ground squirrels. J Am Vet Med Assoc 193, 367–368. Grant, S. G., Jessee, J., Bloom, F. R. & Hanahan, D. (1990). Ott, L. & Burkovski, A. (2014). Toxigenic corynebacteria: adhesion, Differential plasmid rescue from transgenic mouse DNAs into invasion and host response. In Corynebacterium diphtheriae and Escherichia coli methylation-restriction mutants. Proc Natl Acad Sci Related Toxigenic Species, pp. 143–170. Edited by A. Burkovski. USA87, 4645–4649. Dordrecht: Springer. Hart, R. J. (1984). Corynebacterium ulcerans in humans and cattle in Ott, L., Ho¨ ller, M., Rheinlaender, J., Scha¨ ffer, T. E., Hensel, M. & North Devon. J Hyg (Lond) 92, 161–164. Burkovski, A. (2010). Strain-specific differences in pili formation Hirai-Yuki, A., Komiya, T., Suzaki, Y., Ami, Y., Katsukawa, C., and the interaction of Corynebacterium diphtheriae with host cells. Takahashi, M., Yamamoto, A. & Yamada, Y. K. (2013). Isolation BMC Microbiol 10, 257. and characterization of toxigenic Corynebacterium ulcerans from 2 Ott, L., McKenzie, A., Baltazar, M. T., Britting, S., Bischof, A., closed colonies of cynomolgus macaques (Macaca fascicularis)in Burkovski, A. & Hoskisson, P. A. (2012). Evaluation of invertebrate Japan. Comp Med 63, 272–278. infection models for pathogenic corynebacteria. FEMS Immunol Hirata, R., Jr, Souza, S. M., Rocha-de-Souza, C. M., Andrade, A. F., Med Microbiol 65, 413–421. Monteiro-Leal, L. H., Formiga, L. C. & Mattos-Guaraldi, A. L. Ott, L., Scholz, B., Ho¨ ller, M., Hasselt, K., Ensser, A. & Burkovski, A. (2004). Patterns of adherence to HEp-2 cells and actin polymeri- (2013). Induction of the NFk-B signal transduction pathway in sation by toxigenic Corynebacterium diphtheriae strains. Microb response to Corynebacterium diphtheriae infection. Microbiology 159, Pathog 36, 125–130. 126–135. Hogg, R. A., Wessels, J., Hart, J., Efstratiou, A., De Zoysa, A., Mann, Pappenheimer, A. M., Jr (1977). Diphtheria toxin. Annu Rev Biochem G., Allen, T. & Pritchard, G. C. (2009). Possible zoonotic transmission of toxigenic Corynebacterium ulcerans from companion animals in a 46, 69–94. human case of fatal diphtheria. Vet Rec 165, 691–692. Peterson, W. D., Jr, Stulberg, C. S., Swanborg, N. K. & Robinson, Hommez, J., Devriese, L. A., Vaneechoutte, M., Riegel, P., Butaye, P. & A. R. (1968). Glucose-6-phosphate dehydrogenase isoenzymes in Haesebrouck, F. (1999). Identification of nonlipophilic coryne-bacteria human cell cultures determined by sucrose-agar gel and cellulose isolated from dairy cows with mastitis. JClinMicrobiol37, 954–957. acetate zymograms. Proc Soc Exp Biol Med 128, 772–776. Indrigo, J., Hunter, R. L., Jr & Actor, J. K. (2003). Cord factor trehalose Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: 6,69-dimycolate (TDM) mediates trafficking events during mycobact- a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring erial infection of murine macrophages. Microbiology 149, 2049–2059. Harbor Laboratory. Jakoby, M., Ngouoto-Nkili, C.-E. & Burkovski, A. (1999). Cons- Sangal, V. & Hoskisson, P. A. (2014). Infection of infectors. truction and application of new Corynebacterium glutamicum vect- In Corynebacterium diphtheriae and Related Toxigenic Species, ors. Biotechnol Tech 13, 437–441. pp. 67–81. Edited by A. Burkovski. Dordrecht: Springer. Knoppova´, M., Phensaijai, M., Vesely´, M., Zemanova´ , M., Nesvera, J. Sangal, V., Nieminen, L., Weinhardt, B., Raeside, J., Tucker, N. P., &Pa´ tek, M. (2007). Plasmid vectors for testing in vivo promoter Florea, C. D., Pollock, K. G. & Hoskisson, P. A. (2014). Diphtheria- activities in Corynebacterium glutamicum and Rhodococcus erythro- like disease caused by toxigenic Corynebacterium ulcerans strain. polis. Curr Microbiol 55, 234–239. Emerg Infect Dis 20, 1257–1258. Lucas, E. A., Billington, S. J., Carlson, P., McGee, D. J. & Jost, B. H. Scha¨ fer, A., Tauch, A., Ja¨ ger, W., Kalinowski, J., Thierbach, G. & (2010). Phospholipase D promotes Arcanobacterium haemolyticum Pu¨hler, A. (1994). Small mobilizable multi-purpose cloning vectors adhesion via lipid raft remodeling and host cell death following derived from the Escherichia coli plasmids pK18 and pK19: selection bacterial invasion. BMC Microbiol 10, 270. of defined deletions in the chromosome of Corynebacterium Mattos-Guaraldi, A. L., Sampaio, J. L., Santos, C. S., Pimenta, F. P., glutamicum. Gene 145, 69–73. Pereira, G. A., Pacheco, L. G., Miyoshi, A., Azevedo, V., Moreira, Scherer, W. F., Syverton, J. T. & Gey, G. O. (1953). Studies on the L. O. & other authors (2008). First detection of Corynebacterium propagation in vitro of poliomyelitis viruses. IV. Viral multipli- ulcerans producing a diphtheria-like toxin in a case of human with cation in a stable strain of human malignant epithelial cells (strain pulmonary infection in the Rio de Janeiro metropolitan area, Brazil. HeLa) derived from an epidermoid carcinoma of the cervix. J Exp Mem Inst Oswaldo Cruz 103, 396–400. Med 97, 695–710. Meinel, D. M., Margos, G., Konrad, R., Krebs, S., Blum, H. & Sing, Schuhegger, R., Schoerner, C., Dlugaiczyk, J., Lichtenfeld, I., A. (2014). Next generation sequencing analysis of nine Coryne- Trouillier, A., Zeller-Peronnet, V., Busch, U., Berger, A., Kugler, R.

1590 Microbiology 161

Hacker, E. 39 Publications

C. ulcerans–host interaction

& other authors (2009). Pigs as source for toxigenic Corynebact-erium Tauch, A. & Sandbote, J. (2014). The family Corynebacteriaceae. ulcerans. Emerg Infect Dis 15, 1314–1315. In The Prokaryotes, pp. 239–277. Edited by E. Rosenberg, Seto, Y., Komiya, T., Iwaki, M., Kohda, T., Mukamoto, M., E. DeLong, S. Lory, E. Stackebrandt & F. Thompson. Berlin: Springer. Takahashi, M. & Kozaki, S. (2008). Properties of corynephage Tejedor, M. T., Martin, J. L., Lupiola, P. & Gutierrez, C. (2000). Caseous attachment site and molecular epidemiology of Corynebacterium lymphadenitis caused by Corynebacterium ulcerans in the dromedary ulcerans isolated from humans and animals in Japan. Jpn J Infect camel. Can Vet J 41, 126–127. Dis 61, 116–122. Trost, E., Al-Dilaimi, A., Papavasiliou, P., Schneider, J., Viehoever, P., Simpson-Louredo, L., Ramos, J. N., Peixoto, R. S., Santos, L. S., Burkovski, A., Soares, S. C., Almeida, S. S., Dorella, F. A. & other Antunes,C.A.,Ladeira,E.M.,Santos,C.S.,Vieira,V.V.,Boˆ as,M.H.& authors (2011). Comparative analysis of two complete Corynebact- other authors (2014). Corynebacterium ulcerans isolates from humans erium ulcerans genomes and detection of candidate virulence fact- and dogs: fibrinogen, fibronectin and collagen-binding, antimicrobial ors. BMC Genomics 12, 383. and PFGE profiles. Antonie van Leeuwenhoek 105, 343–352. Wagner, K. S., White, J. M., Crowcroft, N. S., De Martin, S., Mann, G. & Sydor, T., von Bargen, K., Hsu, F. F., Huth, G., Holst, O., Wohlmann, Efstratiou, A. (2010). Diphtheria in the United Kingdom, 1986–2008: J., Becken, U., Dykstra, T., So¨ hl, K. & other authors (2013). Diversion the increasing role of Corynebacterium ulcerans. Epidemiol Infect 138, of phagosome trafficking by pathogenic Rhodococcus equi depends on 1519–1530. mycolic acid chain length. Cell Microbiol 15, 458–473. Zakikhany, K. & Efstratiou, A. (2012). Diphtheria in Europe: current Sykes, J. E., Mapes, S., Lindsay, L. L., Samitz, E. & Byrne, B. A. problems and new challenges. Future Microbiol 7, 595–607. (2010). Corynebacterium ulcerans bronchopneumonia in a dog. J Vet Intern Med 24, 973–976. Edited by: J. Simon

http://mic.sgmjournals.org 1591

Hacker, E. 40 Publications

4.2 The killing of macrophages by Corynebacterium ulcerans (Hacker et al., 2015b)

Hacker, E. 41

Publications

The killing of macrophages by Corynebacterium ulcerans

Elena Hacker1), Lisa Ott1), Jan Schulze-Luehrmann2), Anja Lührmann2), Veit Wiesmann3), Thomas Wittenberg3) and Andreas Burkovski1)

1) Friedrich-Alexander-Universität Erlangen-Nürnberg, Professur für Mikrobiologie, Staudtstr. 5, 91058 Erlangen, Germany 2) Friedrich-Alexander-Universität Erlangen-Nürnberg, Universitätsklinikum Erlangen, Mikrobiologisches Institut – Klinische Mikrobiologie, Immunologie und Hygiene, Wasserturmstr. 3-5, 91054 Erlangen, Germany 3) Fraunhofer Institut für Integrierte Schaltungen (IIS), Am Wolfsmantel 33, 91058 Erlangen, Germany

* Corresponding author: Andreas Burkovski, Professur für Mikrobiologie, Friedrich-Alexander-Universität Erlangen- Nürnberg, Staudtstr. 5, 91058 Erlangen, Germany. Phone: +49 9131 85 28086. Fax: +49 9131 85 28082. E-mail: [email protected]

Keywords: Corynebacterium ulcerans, host-pathogen interaction, necrotic cell death, THP-1 cells, FACS

Abbreviations: CFU: colony forming units; G-CSF: granulocyte-colony stimulating factor; GFP: green fluorescent protein, IL-6: Interleukin 6, JAK: Janus kinase, MOI: multiplicity of infection, STAT3: signal transducers and activators of transcription-3

Hacker, E. 42

Publications

ABSTRACT Corynebacterium ulcerans is an emerging pathogen transmitted by a zoonotic pathway with a very broad host spectrum to humans. Despite rising numbers of infections and potentially fatal outcomes, data on the molecular basis of pathogenicity are scarce. In this study, the interaction of two C. ulcerans isolates - one from an asymptomatic dog, one from a fatal case of human infection - with human macrophages was investigated. C. ulcerans strains were able to survive in macrophages for at least 20 hours. Uptake led to delay of phagolysosome maturation and detrimental effects on the macrophages as deduced from cytotoxicity measurements and FACS analyses. The data presented here indicate a high infectious potential of this emerging pathogen.

INTRODUCTION Corynebacterium ulcerans is a pathogenic member of the genus Corynebacterium, which is part of the family Corynebacteriaceae, the order Actinomycetales and the phylum Actinobacteria.1 Within the taxon Actinobacteria, the genera Corynebacterium, Nocardia and Mycobacterium form a monophyletic branch, the CMN group, based on their unusual cell envelope composition.2 The genus Rhodococcus shares this features and is sometimes also regarded as a member of this group, then designated as CMNR group.3, 4 C. ulcerans was first described by Gilbert and Stewart, who isolated the bacteria from the throat of a patient with respiratory diphtheria-like illness.5 In fact, when lysogenized by a tox gene-carrying corynephage, C. ulcerans can – as Corynebacterium diphtheriae – produce diphtheria toxin and during the past decade, diphtheria-like infections with toxigenic C. ulcerans have outnumbered those caused by toxigenic C. diphtheriae in many industrialized countries.6 Moreover, during the last years, human infections associated with C. ulcerans appear to be increasing in various countries and can most often be ascribed to zoonotic transmission.7, 8 The range of mammals that may serve as a reservoir for human infections is extremely broad. C. ulcerans was e.g. isolated from cattle, goats, pigs, wild boars, dogs, cats, ground squirrels, otters, camels, monkeys, orcas and water rats.9-22 In 2011, two C. ulcerans strains from the metropolitan area of Rio de Janeiro, Brazil, were sequenced: BR-AD22, isolated from a nasal swab of an asymptomatic dog, and 809, isolated from a bronchoalveolar lavage (BAL) sample of an 80-year-old woman with fatal pulmonary infection.23-25 Based on these genome sequences and comparative genomics approaches, a number of putative virulence factors were annotated, while functional data were scarce. Up to now, data on adhesion and invasion of epithelial cells, fibrinogen, fibronectin and collagen binding, antimicrobial profiles and arthritogenic potential of isolates were published.26- 28 In the study presented here, C. ulcerans 809 and BR-AD22 were characterized for the first

Hacker, E. 43

Publications time to our knowledge in respect to their interaction with macrophages as important part of the innate immune system in order to get deeper insight into the pathogenicity of this emerging pathogen. The interaction with macrophages was of special interest, since the closely related animal pathogen Corynebacterium pseudotuberculosis, a facultative intracellular bacterium, is able to survive and grow in macrophages in order to disseminate within the host, 3, 29 and also the human pathogen C. diphtheriae is able to survive within human macrophages.30

RESULTS Internalization of C. ulcerans by human macrophages As a first approach, the interaction of human macrophages with C. ulcerans was analyzed by fluorescence microscopy, together with Corynebacterium glutamicum used as control. This bacterium is non-pathogenic and was expected to be eliminated quickly by the macrophages. GFP-labeled bacteria were readily internalized by human THP-1 macrophages without opsonization or additional external primining e.g. by lipopolysaccharide. The fate of the bacteria was monitored for 2, 8 and 20 hours. The number of fluorescent bacteria within the macrophages appeared to be constant or declining within the 20 hours for C. glutamicum ATCC 13032. In contrast, a time-dependent increase of C. ulcerans 809 and BR-AD22 within macrophages was observed by fluorescence microscopy within this time interval (Fig. 1). Similar results were obtained with the murine macrophage cell line RAW 264.7 (data not shown).

Hacker, E. 44

Publications

C. glutamicum C. ulcerans C. ulcerans ATCC 13032 809 BR-AD22 2 h 8 h 20 h

Fig. 1. Fluorescence microscopy of C. ulcerans and THP-1 cells. THP-1 cells were infected with C. glutamicum ATCC 13032 pEPR1p45gfp, C. ulcerans 809 pEPR1p45gfp and C. ulcerans BR-AD22 pEPR1p45gfp at an MOI of 10 for 30 min. Extracellular bacteria were killed by the addition of gentamicin and after different time points, cells were fixed. Nuclei were stained with DAPI, the cytoskeleton with Alexa Fluor® 647 Phalloidin and z-stack micrographs were taken using the confocal laser-scanning microscope Leica SP5 II and analyzed with the LAS software suite to proof that bacteria are located inside of the macrophages. Representative pictures are shown.

Survival of C. ulcerans after internalization by human macrophages For a more quantitative analysis of C. ulcerans – macrophage interaction, a replication assay was carried out and CFU were determined after adhering bacteria were killed by gentamicin addition. When the number of bacteria within THP-1 cells was analyzed, already after 2 hours almost no viable C. glutamicum were detectable (Fig. 2A), indicating that the fluorescent bacteria observed at the same and later time points by microscopy were already irreversibly inactivated or killed by the macrophages. When the low number of viable C. glutamicum at 2 hours post-infection were set to 100 %, a constant decline of CFU could be observed at the later time points (Fig. 2B). For C. ulcerans strains 809 and BR-AD22 already after 2 hours post-infection higher CFU were detected compared to the non-pathogenic C. glutamicum

Hacker, E. 45

Publications strain. After 8 hours of macrophage infection, even a strong increase of CFU was observed both, in relation to the number of bacteria used for infection and in relation to the number of intracellular CFU determined at 2 hours post infection (Fig. 2A, B). This result indicates that – in contrast to C. glutamicum - C. ulcerans is able to proliferate in the macrophages within the first hours of internalization. After 20 hours, the number of CFU declined for the two C. ulcerans strains, indicating successful inactivation of bacteria at this late time point by the macrophages. Interestingly, with longer incubation time, cells challenged by C. ulcerans in this experimental set-up began to detach from the wells and showed signs of cell death, which was analyzed in more detail in subsequent approaches (see below). Taken together, in contrast to C. glutamicum, C. ulcerans strains seem to be able to interfere with macrophage function independent of their origin from animal or human sources. Similar results were obtained with murine RAW 264.7 and J774E macrophages indicating that these findings are not limited to a single macrophage cell line or exclusively to human macrophages (data not shown).

180 160 A 140 120 100 80 60 40 intracellular CFU [%] 20 0 C.ATCC13032 glutamicum C. ulcerans809 C.BR ulcerans-AD22 ATCC 13032 809 BR-AD22

500 450 B 400 350 300 250 200 150 100

intracellular survival [%] 50 0 C.ATCC13032 glutamicum C. ulcerans809 C.BR ulcerans-AD22 ATCC 13032 809 BR-AD22

Fig. 2. Quantitative analysis of viable intracellular C. ulcerans in THP-1 cells. THP-1 cells were infected with C. ulcerans wild type strains 809 and BR-AD22 at an MOI of 1 for 30 min. To kill extracellular bacteria, cells were incubated with medium containing gentamicin and after 2 (white bars),

Hacker, E. 46

Publications

8 (grey bars) and 20 h (black bars), cells were harvested, lysed and lysates were plated on blood agar plates to recover intracellular CFU. A) Intracellular CFU in percent referred to the inoculum, B) intracellular survival in percent referred to the bacteria that were taken up after 2 h. Data shown are mean values of four independent biological replicates each performed in triplicates ± standard deviation.

Delay of phagolysosome maturation by C. ulcerans A hallmark of macrophage function after phagocytosis is the formation of phagolysosomes by fusion of phagosomes and acidic lysosomes. In order to investigate the reason for the observed growth and delay of degradation of C. ulcerans, formation of acidic compartments within macrophages was monitored using a LysoTracker dye. In parallel, Alexa Fluor® 647 Phalloidin and DAPI staining visualized cytoskeleton and nuclei, respectively, and corynebacteria were labeled by GFP (Fig. 3). In case of C. glutamicum ATCC 13032, a clear co-localization of bacteria and acidic compartments was observed already after 2 hours. After 8 and 20 hours, this co-localization was even more pronounced with almost all bacteria located together with acidic compartments. This situation differed strongly in case of C. ulcerans isolates. After 2 hours, no co-localization of strain 809 or BR-AD22 with lysosomes was detected. After 8 hours, only about half of the bacteria detected were co-localizing with lysosomes, and even after 20 hours, a considerable number of GFP-labeled bacteria were not co-localizing with acidic compartments. The data indicate a delay of phagolysosome formation by C. ulcerans. In accordance with the CFU determined (Fig. 2), the delaying effect of 809 seemed to be stronger than in case of BR-AD22. For an unbiased and more quantitative approach, pictures were automatically evaluated either on the level of pixels or bacterial cells. Both methods showed consistently that C. ulcerans impairs phagosome maturation. In contrast to the non-pathogenic C. glutamicum, which already shows approximately 60 % co-localization after 2 hours post-infection and at least 80 % after 20 hours, the C. ulcerans strains applied show lower co-localization at all time points investigated with less than half of the bacteria co-localized with acidic compartments after 20 hours (Table 1).

Hacker, E. 47

Publications

C. glutamicum C. ulcerans C. ulcerans ATCC 13032 809 BR-AD22 2 h 8 h 20 h

Fig. 3. Labeling and tracking of acidic organelles in THP-1 cells infected with C. ulcerans. THP-1 cells were incubated with LysoTracker® Red DND-99 for 120 min before cells were infected with C. glutamicum ATCC 13032 pEPR1p45gfp, C. ulcerans 809 pEPR1p45gfp and C. ulcerans BR-AD22 pEPR1p45gfp at an MOI of 10 for 30 min. Extracellular bacteria were killed by the addition of gentamicin and after 2, 8 and 20 h, cells were fixed. Nuclei were stained with DAPI, the cytoskeleton with Alexa Fluor® 647 Phalloidin and micrographs were taken using the confocal laser-scanning microscope Leica SP5 II and analyzed with the LAS software suite. Non-pathogenic C. glutamicum immediately co-localize with acidic compartments (2 h) whereas pathogenic C. ulcerans only show co-localization after longer incubation time (8 h to 20 h). Representative pictures are shown. Scale bars: 10 µm.

Hacker, E. 48

Publications

Table 1. Automated analysis of co-localization of bacteria with acidic compartments. At least 19 fluorescence microscopy pictures were analyzed for each data set as described in the Materials and Methods section.

% co-localization after evaluated on pixel level 2 h 8 h 20 h C. glutamicum 59 82 91 ATCC 13032 C. ulcerans 809 8 35 49 C. ulcerans BR- 19 29 47 AD22 % co-localization after evaluated on bacterium level 2 h 8 h 20 h C. glutamicum 54 80 79 ATCC 13032 C. ulcerans 809 11 34 49 C. ulcerans BR- 20 29 47 AD22

Response of human macrophages to C. ulcerans contact The observed delay in phagolysosome maturation gave rise to the question whether macrophage function is impaired in general or specifically. To address this, NFκB induction and cytokine secretion were analyzed in response to infection with C. ulcerans. Cells of the reporter cell line THP1-Blue NF-κB were incubated for 20 hours with viable (Fig. 4A) and UV- killed (Fig. 4B) bacteria of non-pathogenic C. glutamicum ATCC 13032 and pathogenic C. ulcerans strains 809 and BR-AD22. Viable corynebacteria led to strong NF-κB activation when MOI 1 and 10 were tested independent of their pathogenicity (Fig. 4A). When an MOI of 100 was applied, NF-κB activation by C. ulcerans strains was decreased (data not shown) most likely due to detrimental effects of these pathogenic bacteria on the reporter cells (see below). Infection with dead bacteria led to a weaker NF-κB activation, which reached the values obtained for viable cells only in case of 10-fold higher MOIs were applied. The activation by dead bacteria was dose-dependent; detrimental effects of UV-inactivated pathogens were not observed (Fig. 4B), indicating that macrophage damage is the result of an active process induced by C. ulcerans.

Hacker, E. 49

Publications

3.5 A 3 2.5 levels 2 1.5 Rel. SEAP Rel. SEAP

rel. SEAP levels 1 0.5 0 C.ATCC13032 glutamicum C. ulcerans809 C.BR ulcerans-AD22 ATCC 13032 809 BR-AD22

3.5 B 3 2.5 levels 2 1.5 Rel. SEAP Rel. SEAP

rel. SEAP levels 1 0.5 0 C.ATCC13032 glutamicum C. ulcerans809 C.BR ulcerans-AD22 ATCC 13032 809 BR-AD22

Fig. 4. NF-κB activation in THP1-Blue™ NF-κB reporter cells after C. ulcerans infection. THP1- Blue™ NF-κB cells were incubated for 20 h with A) viable and B) UV-killed bacteria of the non- pathogenic C. glutamicum ATCC 13032 and pathogenic C. ulcerans strains 809 and BR-AD22 at an MOI of 1 (white bars), 10 (grey bars) and 100 (black bars). For viable C. ulcerans, MOI 100 led to detrimental effects on cells and data are not shown. Subsequently, supernatants were taken and mixed with QuantiBlue SEAP detection solution leading to a change in color upon NF-κB activation. Living C. glutamicum led to high SEAP levels in all MOIs, whereas living C. ulcerans showed lower levels at higher MOIs. Infection with dead bacteria led to a dose-dependent NF-κB activation in all cases. Data shown are mean values of three independent biological replicates each performed in triplicates ± standard deviation.

Additionally, supernatants of THP-1 cells infected with C. ulcerans with an MOI of 1 and 10, respectively, were collected at 2, 8 and 20 hours post-infection and used for determination of IL-6 (Fig. 5A) and G-CSF (Fig. 5B) secretion. Upon receptor binding, IL-6 and G-CSF activate JAK kinases, which then activate STAT3 signaling leading to increased expression of the anti- apoptotic proteins Bcl-xL and Bcl-2.31, 32 C. ulcerans 809 and BR-AD22 led to rising cytokine levels with infection time and reached concentrations of 1,000 to 1,400 pg ml-1 for IL-6 and 4,500 to 6,500 pg ml-1 for G-CSF, respectively.

Hacker, E. 50

Publications

1800 1600 A 1400

/ml] 1200

pg 1000 IL - 6 [

6 [pg/ml] 800 -

IL 600 400 200 0 C. ulcerans809 C.BR ulcerans-AD22 809 BR-AD22

9000 8000 B 7000

/ml] 6000

pg 5000 4000 - CSF [ CSF [pg/ml] CSF G

- 3000 G 2000 1000 0 C. ulcerans809 C.BR ulcerans-AD22 809 BR-AD22

Fig. 5. Cytokine ELISA of THP-1 cells after infection with C. ulcerans. Supernatants of THP-1 cells infected with C. ulcerans were collected at 2 (white bars), 8 (grey bars) and 20 h (black bars) post- infection and used as samples for determination of A) IL-6 and B) G-CSF concentrations. Data shown are mean values of three independent biological replicates each performed in triplicates ± standard deviation.

Taken together, these data indicate functional NF-κB and STAT3 signal transduction pathways in the macrophages, indicating that delay of phagolysosome formation is caused by a specific mechanism and not a general detrimental effect of C. ulcerans.

Killing of macrophages by C. ulcerans As described above, a partial detachment of cells from the surface of cell culture wells was observed in response to C. ulcerans contact. This phenomenon was analyzed in more detail in subsequent experiments. First, different MOIs and times of infection were tested. Detachment of cells was not observed with untreated cells and those infected with an MOI of 1. In contrast, C. ulcerans strains 809 and BR-AD22 applied with an MOI of 10 led to malformation and detachment of cells after 8 hours. An effect, which was not observed with C. glutamicum ATCC 13032 (Fig. 6).

Hacker, E. 51

Publications

untreated C. glutamicum C. ulcerans C. ulcerans cells ATCC 13032 809 BR-AD22 20 h MOI = 1 2 h 8 h MOI = 10 20 h

Fig. 6. Detachment of THP-1 cells infected with C. ulcerans. Cells infected with non-pathogenic C. glutamicum or pathogenic C. ulcerans at an MOI of 10 and 1 were analyzed microscopically at 2, 8 and 20 h post-infection, untreated cells served as negative control. Both C. ulcerans wild type strains caused detachment of cells at MOI 10 after 8 h, and an increasing effect with longer incubation time. Cells treated with C. glutamicum looked healthy even after 20 h, as well as cells incubated with a lower amount of bacteria (MOI 1).

Microscopic analysis of DAPI stained fragmented nuclei as a sign of apoptotic cell death of THP-1 cells after infection with C. ulcerans revealed only single apoptotic cells independent of the time point analyzed, which is in accordance with the activation of anti-apoptotic cell signaling as described above. To further analyze the mode of cell death, FACS experiments were carried out. First, the SubG1 technique was used as assay to measure apoptosis rates. However, no typical SubG1 fractions were detectable for THP-1 cells infected with C. ulcerans at MOI 50 further supporting the idea that apoptosis is not involved in macrophage death (data not shown). As second approach, lysis of cells was tested. To quantify the number of lytic cells upon C. ulcerans infection, cells were harvested at different time points, stained with 7-AAD and analyzed by flow cytometry after gating and exclusion of cell debris (Fig. 7A). 7-AAD is a dye

Hacker, E. 52

Publications that is excluded from viable cells, but can penetrate damaged cell membranes. In the DNA histogram, 7-AAD positive cells show a shift to the right and can be easily separated from viable cells (Fig. 7B). Uninfected cells show a background level of around 10 % 7-AAD positive cells over all points in time investigated. Handling and infection of cells increased this background to about 30 % at 2 hours post-infection. At 8 hours post-infection, the count of positive cells increased for all samples (Fig. 7B); however, when all events are considered (Fig. 7A), the amount of small fragments detected was much higher for cells infected with C. ulcerans compared to cells infected with C. glutamicum, indicating that many cells were already fragmented due to C. ulcerans infection. These were gated out and not considered in the count of 7-AAD positive cells, resulting in a relatively low difference between the strains. After 20 hours, cells infected with C. glutamicum recover and drop down to background level of uninfected cells, whereas cells infected with C. ulcerans strains show about 50 % positive cells. Furthermore, again a high amount of cell debris was observed. In summary, a lytic cell death mechanism resembling necrosis upon infection with C. ulcerans was indicated by this approach (Fig. 7B,C).

Hacker, E. 53

Publications

A 2 h 8 h 20 h SSC

FSC

B 2 h 8 h 20 h count

7-AAD C [%] cells - AAD positive 7

uninfected C. glutamicum C. ulcerans C. ulcerans ATCC 13032 809 BR-AD22

Hacker, E. 54

Publications

Fig. 7. 7-AAD staining and FACS analysis of THP-1 cells infected with C. ulcerans. THP-1 cells were infected with the non-pathogenic C. glutamicum ATCC 13032 and pathogenic C. ulcerans strains 809 and BR-AD22 at MOI 50, uninfected cells were carried along as negative control. After different points in time, cells were harvested and stained with 7-AAD, a dye that only penetrates dead cells, and analyzed by flow cytometry. A) Dot plots of representative samples show forward scatter versus side scatter. Gate A excludes cell debris and small fragments like bacteria. B) Histograms (7-AAD signal versus cell count) of 10,000 cells after gating. Gate B represents 7-AAD positive, dead cells (light gray in gate A). C) Quantitative analysis of 7-AAD positive cells at 2 (white bars), 8 (grey bars) and 20 h (black bars). Mean values and standard deviations of at least two independent assays performed in triplicate are depicted.

In addition to the FACS assays, the release of cytosolic lactate dehydrogenase into the supernatant as a sign of host cell damage was measured during infection of THP1-Blue NF- κB cells with C. ulcerans. Cells were infected for 20 hours with viable and UV-killed C. glutamicum ATCC 13032 and C. ulcerans strains 809 and BR-AD22 at an MOI of 1 and 10. The non-pathogenic C. glutamicum as well as dead C. ulcerans had no damaging effect, while viable C. ulcerans strains 809 and BR-AD22 induced cell lysis as indicated by strong LDH release already at an MOI of 1 (Fig. 8).

120 A 100 A

[%] 80

60

40 Cytotoxicity

cytotoxicity [%] cytotoxicity 20

0 C. glutamicum C. ulcerans C. ulcerans -20 CGATCC ATCC13032 13032 CU809 809 CUBR BR-AD22-AD22

B [%] Cytotoxicity

C. ulcerans C. ulcerans C. ulcerans BR-AD22 ELHA1 ELHA3

Hacker, E. 55

Publications

Fig. 8. LDH release of THP1-Blue cells infected with C. ulcerans. The release of lactate dehydrogenase (LDH) as a sign of host cell damage during infection of THP1-Blue™ NF-κB cells with C. ulcerans was measured using the cytotoxicity detection kit (Roche). A) Cells were infected for 20 h with the non-pathogenic C. glutamicum ATCC 13032 and pathogenic C. ulcerans wild type strains 809 and BR-AD22 at an MOI of 1 (white bars) and 10 (grey bars). The non-pathogenic C. glutamicum had no damaging effect. However, C. ulcerans led to LDH release at MOI 1 and 10. B) Cells were infected with C. ulcerans mutant strains ELHA1 and ELHA3 at MOI 1 and the LDH release was compared to the corresponding wild type strain BR-AD22 which was set to 100 %. Data shown are mean values of three independent biological replicates each performed in triplicates ± standard deviation.

Factors involved in the killing of macrophages by C. ulcerans were not characterized before. However, based on homology to closely related pathogenic corynebacteria, different C. ulcerans proteins have been annotated as putative virulence factors.25 Phospholipase D (PLD) is the most important virulence factor of the closely related C. pseudotuberculosis.33 This so- called ovis toxin is involved in dissemination of this animal pathogen from the initial site of infection to other parts of the body. Another example of a putative virulence factor encoded in the C. ulcerans genome is CULC22_00609, a homolog of C. diphtheriae DIP0733, which was found recently to be involved in adhesion to and invasion of epithelial cells and in the survival inside macrophages.34 BR-AD22 strains carrying a pld and CULC22_00609 gene disruption, respectively, were generated and designated ELHA1 and ELHA3.27, 35 When cells were infected with these mutant strains and LDH release was measured, at MOI 1 strain ELHA1 showed 51 ± 16 % and ELHA3 34 ± 13 % cytotoxicity compared to the wild type, indicating a multifactorial mechanism of macrophage damage, which has to be analyzed in more detail in future. A complex interaction is also indicated by the fact that the defect of mutant strains observed at MOI 1 could be overcome by higher MOIs (data not shown).

DISCUSSION C. ulcerans is an emerging pathogen, which is transmitted by a zoonotic pathway from a wide number of mammals to the human host.7, 8 In fact, in the last decade cases of respiratory diphtheria caused by C. diphtheriae were outnumbered by C. ulcerans infections in Western Europe.6, 36, 37 Despite increasing numbers of infections and the occurrence of fatal cases, basic mechanisms of host-pathogen interaction and specific virulence factors of C. ulcerans are widely uncharacterized. In the study presented here, C. ulcerans BR-AD22, isolated from an asymptomatic dog, and 809, isolated from an 80-year-old woman with fatal pulmonary infection were characterized for the first time to our knowledge in respect to their interaction with macrophages. Interestingly, the human and animal isolate did not differ significantly in respect to internalization and survival within macrophages and showed higher uptake and survival rates

Hacker, E. 56

Publications compared to C. diphtheriae strains in U937 macrophages,30 supporting the high infectious potential of this emerging pathogen. In contrast to non-pathogenic C. glutamicum strain ATCC 13032, the C. ulcerans strains were able to multiply within macrophages at least 8 hours post-infection and were able to delay phagolysosome formation, which might be an effective mechanism for immune evasion and spreading of C. ulcerans within the body supporting the establishment and progress of infections. Corynebacterium species are closely related to the genera Mycobacterium, Nocardia and Rhodoccoccus, forming the CMNR group of Actinobacteria. The members of this group are characterized by a complex cell wall architecture with an additional layer of mycolic acids. Interaction with macrophages and the influence of mycolic acids is best investigated for Mycobacterium tuberculosis. Here, trehalose dimycolate, the cord factor, was shown to inhibit fusion events between phospholipid vesicles inside the host macrophage and to be involved in macrophage activation.38 Furthermore, Rhodococcus equi, a facultative intracellular pathogen, is able to arrest phagosome maturation in macrophages before the late endocytic stage.39 For corynebacteria, data on macrophage interaction and the role of mycolic acids are scarce. As shown previously, outer membrane lipids of C. pseudotuberculosis have a lethal effect on murine and caprine macrophages.40, 41 It might therefore be worth to investigate the role of mycolic acids in the interaction of C. ulcerans with host macrophages, besides the influence of PLD and CULC22_00609. Flow cytometry analysis revealed a high amount of cell debris when THP-1 cells were infected with C. ulcerans 809 and BR-AD22 and a strong lytic effect was found when LDH release was tested. Together with the absence of evidence for nucleus fragmentation a lytic cell death mechanism is favored by these results rather than apoptosis. In the past few years, several mechanisms of programmed necrotic cell death besides apoptosis have been described. These include besides others pyroptosis and necroptosis, which result in plasma membrane rupture. In contrast to classical necrosis, they are initiated in response to various intrinsic or extrinsic signals and underlie specific signaling cascades resulting finally in cell death (for review see 42, 43). To unravel the molecular background and exact mode of cell death induction by C. ulcerans in THP-1 cells, further investigations need to be carried out. A host cell damaging effect of C. ulcerans to eukaryotic cells was already described for the epithelial cell line Detroit562. Here, the transepithelial resistance of cell monolayers was rapidly reduced in response infection with C. ulcerans 809 and in a less severe mode with C. ulcerans BR-AD22. The more aggressive behavior of C. ulcerans 809 towards epithelial cells might be due to the additional Shiga-like toxin this strain carries.27 Taken together, the observations made in this study hint to the existence of more than one active mechanism of C. ulcerans to cope with macrophage function. In line with this idea, initial sequencing of strains 809 and BR-AD22 25 as well as next generation sequencing of nine

Hacker, E. 57

Publications

C. ulcerans isolates 8 revealed a number of different virulence factors, which might influence survival of this pathogen besides the Shiga-like toxin of strain 809 as well as PLD and DIP0733 homolog CULC22_00609 of BR-AD22. Mutations in the genes encoding PLD and CULC22_00609 led reduced cytotoxicity towards THP1-Blue™ NF-κB cells compared to the wild type strain BR-AD22. To further investigate the effect of these putative virulence factors, studies analyzing the effect of purified proteins on macrophage survival would be of high interest. Further proteins found for C. ulcerans were e.g. neuraminidase H, endoglycosidase E, a Fic toxin homolog and a RhuM homolog.7, 8, 25 In conclusion, the emerging pathogen C. ulcerans is a potent pathogen, which comprises a broad and varying set of virulence factors.25 Furthermore, by acquisition of new genes, this repertoire can change quickly,8 supporting the necessity of a constant surveillance of this pathogen.44 The contribution of putative virulence proteins identified so far and of other components of the cell, like glycolipids, which play an important role in virulence of C. diphtheriae, M. tuberculosis or R. equi,38, 39, 45, 46 has to be elucidated at the molecular level by genetic experiments with defined mutants in future.

MATERIALS AND METHODS Bacterial strains and growth Strains and plasmids used in this study are shown in Table 2. C. glutamicum and C. ulcerans strains were grown in Heart Infusion (HI) broth at 37°C. If appropriate, 50 μg ml-1 kanamycin was added. For generation of GFP expressing strains used in fluorescence microscopy studies, electrocompetent C. ulcerans cells were transformed with the plasmid pEPR1-p45gfp and positive clones were selected on HI agar with kanamycin.

Hacker, E. 58

Publications

Table 2. Strains, cell lines and plasmids used in this study.

Strain, cell line or plasmid Description/genotype Reference/source

Strains

C. glutamicum

ATCC 13032 type-strain, non-pathogenic 50

Corynebacterium ulcerans

809 isolated from an 80-year-old woman 24 with fatal pulmonary infection

BR-AD22 isolated from an asymptomatic dog 23

ELHA1 BR-AD22 pld::pK18mob-pld’, 35 phospholipase D-deficient mutant

ELHA3 BR-AD22 27 CULC22_00609::pK18mob-

CULC22_00609’

Cell lines

THP-1 human leukemic monocytic cells 51

THP1-Blue NFκB THP-1 cells with stable integrated Invivogen NFκB inducible SEAP reporter construct

RAW 246.7 murine leukemic macrophages 52

J774E mannose receptor-expressing clone 53 of the J774 mouse macrophage cell line Plasmids

pEPR1-p45gfp P45, gfpuv, KmR, rep, per, T1, T2 54

Hacker, E. 59

Publications

Replication assay THP-1 human monocytic cells cultures in 10 % FCS supplemented RPMI medium 1640 -1 -1 (containing 100 U penicillin ml and 100 µg streptomycin ml ) at 37°C in 5 % CO2 in a humified cell culture incubator. For replication assays, cells were seeded in 24-well plates (Nunc) at a density of 2 x 105 and differentiated by addition of 10 ng ml-1 phorbol 12-myristate 13-acetate (PMA) 24 h prior to infection. Overnight cultures of C. ulcerans grown in HI were re-inoculated to an OD600 of 0.1 in fresh medium and grown to an OD600 of 0.4 to 0.6. An inoculum with an MOI of 1 or 10 was prepared in RPMI without antibiotics and 500 µl per well were used to infect the cells. The plates were centrifuged for 5 min at 350 x g to synchronize infection and incubated for 30 min (37°C, 5 % CO2, 90 % humidity) to allow phagocytosis of bacteria. Subsequently, the supernatant containing non-engulfed bacteria was aspirated, cells were washed once with PBS and remaining extracellular bacteria were killed by addition of 100 µg ml-1 gentamicin in cell culture medium. After 2 h, cells were either lysed and intracellular bacteria were recovered or further incubated with medium containing 10 µg ml-1 gentamicin for analysis at later time points (8 h and 20 h). To recover intracellular bacteria, the medium was aspirated and cells were lysed by adding 500 µl of 0.1 % TritonX-100 in PBS. Serial dilutions of the lysate and the inoculi were plated on blood agar plates (Oxoid) using an Eddy Jet Version 1.22 (IUL Instruments). After incubation at 37°C for two days, the number of colony forming units (CFU) was determined. The ratio of bacteria used for infection (number of colonies on inoculum plates) and bacteria in the lysate (number of colonies on the lysate plates) multiplied with 100 gave the percentage of viable intracellular bacteria at different time points. When the survival of intracellular bacteria in THP-1 cells was analyzed over the time, the number of CFU at 2 h was set to 100 % and later time points were calculated based on this value. The assay was performed in four biological replicates each performed in triplicates and means and standard deviations were calculated.

Fluorescence microscopy For qualitative analysis of intracellular CFU, THP-1 cells were infected with GFP-expressing bacteria and analyzed by fluorescence microscopy. THP-1 cells were seeded one day prior to infection in a density of 1 x 105 cells on sterile coverslips in 24-well plates. Overnight cultures of C. glutamicum or C. ulcerans strains transformed with plasmids encoding gfp cultivated in

HI medium containing kanamycin were re-inoculated to an OD600 of 0.1 in fresh medium, harvested at the beginning of the exponential growth phase and used to infect macrophages as described above. After different time points, the medium was aspirated and cells were fixed by addition of 500 µl 4 % paraformaldehyde in PBS and incubated for 20 min at 37°C. Until further staining, cells were stored in PBS at 4°C. For subsequent analysis by microscopy, coverslips were incubated with 30 µl of Alexa Fluor® 647 Phalloidin diluted 1:200 in Image-iT™

Hacker, E. 60

Publications

FX Signal Enhancer (Molecular Probes, Life Technologies) for 45 min in the dark to stain the cytoskeleton of THP-1 cells. After washing twice with PBS, the coverslips were dried and embedded on glass slides in ProLong® Gold antifade mountant with DAPI (Molecular Probes, Life Technologies) and samples were stored in the dark at 4°C. Micrographs were taken with the confocal laser scanning microscope Leica SP5 CLSM-1P (Leica Microsystems) and analyzed with the LAS software suite.

Staining of acidic compartments in infected THP-1 cells To analyze if C. ulcerans co-localizes with acidic compartments, THP-1 cells we were treated with 200 nM LysoTracker® Red DND-99 (Molecular Probes, Life Technologies), a red fluorescent dye that stains acidic compartments in live cells, 2 h before infection. Then, cells were infected and further treated as described for fluorescence microscopy above.

Automated analysis of fluorescence microscopic images The analysis of bacteria on fluorescence images poses a demanding challenge as the bacteria tend to stick together and form clusters. To circumvent this problem, i.e. to avoid the problem of segmentation of single bacteria, data were analyzed with the software tool CaeT 47 in two ways. In the first approach the area of all bacteria was analyzed on pixel level. For this purpose, an adaptive threshold algorithm 48 based on k-means clustering 49 was applied to the images of the GFP channel to determine pixels, which belong to bacteria. If intensities of corresponding pixels in the LysoTracker® Red DND-99 channel exceeded an empirically determined threshold, the pixels were regarded as co-localized. For all images the proportion between the number of co-localizing pixels and the total number of pixels of bacteria was determined. The second method was based on the evaluation of images at bacterial cell level. For this purpose, pixels belonging to bacteria which had been calculated in the first method, were grouped into regions. To prevent the analysis of cell clusters, these regions were filtered according to a statistical shape model to analyze exclusively single bacteria. The single bacteria had to reach a minimum average intensity in the LysoTracker® Red DND-99 channel to be positive for co- localization with acidic compartments. Then, the proportion between the number of positive bacteria and the total number of bacteria was determined.

FACS apoptosis assay (SubG1 analysis) For analysis of apoptotic cell death induced by infection with C. ulcerans, cells were stained with propidium iodide (PI) and the SubG1 peak was analyzed by FACS. 4 x 105 THP-1 cells were seeded 24 h prior to infection in 12-well plates and differentiated by addition of 10 ng ml- 1 PMA. Cells were infected with C. ulcerans at an MOI of 50 as described above. After different time points, cells were harvested and fixed. To collect all cells, the supernatant with potentially

Hacker, E. 61

Publications detached cells was saved. Cells were washed once with 50 mM EDTA in PBS, detached by incubation for 5 to 10 min with trypsin at 37°C, and washed with 10 % FCS in PBS. Then everything was combined and centrifuged at 350 x g for 10 min. The supernatant was carefully removed and the cell pellet was washed once with 10 % FCS in PBS. The cell pellets were resuspended in 500 µl ice-cold PBS and 3 ml of 70 % ethanol (-20°C) were added dropwise under vortexing. Fixed cells were stored up to two weeks at 4°C. For staining, cells were harvested by centrifugation (500 x g, 10 min), and cell pellets were resuspended in 1 ml extraction buffer (9 parts 50 mM Na2HPO4, 1 part 25 mM citric acid, 0.1 % TritonX-100, 0.01

% NaN3, pH 7.8) to extract low molecular weight DNA, transferred to a FACS tube and centrifuged again (500 x g, 10 min). Finally, cells were stained by adding 400 µl of the staining buffer (10 mM PIPES, 0.1 N NaCl, 2 mM MgCl2, 0.1 % TritonX-100, 0.02 % NaN3, pH 6.8), 25 -1 -1 µl 10 mg ml RNase in PBS and 20 µl 1 mg ml PI dissolved in ddH2O. Until flow cytometric analysis on a BD FACS Canto II, samples were incubated in the dark.

FACS necrosis assay (7-AAD staining) For analysis of necrotic cell death forms in response to infection with C. ulcerans, cells were stained with 7-AAD. 4 x 105 THP-1 cells were seeded 24 h prior to infection in 12-well plates and differentiated by addition of 10 ng ml-1 PMA. Cells were infected with C. ulcerans at an MOI of 50 as described above. At different time points after infection, cells were harvested as described above for the SubG1 analysis, the resulting pellet was resuspended in 250 µl PBS but then not fixed but immediately stained with 5 µl of the 7-AAD staining solution (BD Bioscience). Uninfected cells served as negative control and cells treated with 0.01 % TritonX- 100 as positive control. Analysis was performed within 10 minutes by flow cytometry on a BD FACS Canto II after gating and excluding cell debris. Data were analyzed using KaluzaTM flow cytometry analysis software V1.1 (Beckmann Coulter).

NF-κB reporter assay THP1-Blue™ NF-κB cells (InvivoGen) carrying a stable integrated NF-κB-inducible secreted embryonic alkaline phosphatase (SEAP) reporter construct were used to analyze NF-κB induction by C. ulcerans. C. ulcerans strains were inoculated from an overnight culture to an

OD600 of 0.1 in fresh medium and grown to an OD600 of 0.4 to 0.6. An inoculum with an OD600 of 1.25 in 1,000 µl PBS was prepared and 20 µl of this inoculum or of the 10-1 and 10-2 dilutions were mixed with 180 µl of a suspension with 5 x 105 THP1-Blue™ NF-κB cells in cell culture medium resulting in an MOI of 100, 10 or 1. UV-killed bacteria in the same concentrations were also carried along. After incubation for 20 h at cell culture conditions, the 96-well plates were centrifuged (350 x g, 5 min) and 20 µl of the cell free supernatant was mixed with 180 µl prewarmed SEAP detection reagent QUANTI-Blue™ (InvivoGen). After further incubation at

Hacker, E. 62

Publications cell culture conditions for 3 h, the levels of NF-κB-induced SEAP resulting in a color change from pink to blue was measured in a microplate reader (TECAN Infinite 200 PRO) at 620 nm.

Determination of cytokine excretion For determination of cytokine activation through C. ulcerans, supernatants of infected THP-1 cells were collected after different time points and stored at -20°C. IL-6 and G-CSF concentrations were measured using the DuoSet ELISA Kits according to the manufacturer’s recommendations (R&D systems). Briefly, the ELISA plates were coated over night with a capture antibody at room temperature, washed three times with 0.05 % Tween20 in PBS, blocked for 1 h at room temperature with 1 % BSA in PBS and washed again three times. Subsequently, 100 µl supernatant of infected cells or standard dilutions were added and the plates were incubated for 2 h, washed again three times and further incubated for 2 h at room temperature. After another washing step, a streptavidin-HRP solution was added and the plates were stored for 20 min under light exclusion, washed again and incubated for another

20 min in the dark with substrate solution. To stop the color reaction, 2 N H2SO4 was added to the wells and the optical density was determined using a microplate reader (TECAN Infinite 200 PRO) set to 450 nm with wavelength correction at 550 nm.

LDH release The release of cytosolic lactate dehydrogenase (LDH) as a sign of host cell damage during infection was measured using the cytotoxicity detection kit (LDH) according to the supplier (Roche). Briefly, 100 µl supernatant of infected cells were mixed with 2.5 µl of the provided catalyst solution and 112.5 µl of the provided dye solution in 96 well plates, incubated in the dark for 30 minutes and the absorbance was measured at 490 nm and wavelength correction at 620 nm in a multiplate reader (TECAN Infinite 200 PRO). Cells treated with 2 % TritonX-100 served as positive control for maximal LDH release and were set to 100 %, untreated cells served as negative control.

ACKNOWLEDGEMENTS The project was supported by the Deutsche Forschungsgemeinschaft in frame of SFB796 (A4, B8 and MGK).

Hacker, E. 63

Publications

REFERENCES

1. Tauch A, Sandbote J. The family Corynebacteriaceae. In: Rosenberg E, DeLong E, Lory S, Stackebrandt E, Thompson F, eds. The Prokaryotes. Berlin Heidelberg: Springer 2014:239-77. 2. Burkovski A. Cell envelope of corynebacteria: structure and influence on pathogenicity. ISRN Microbiol 2013; 2013:935736. 3. Dorella FA, Pacheco LG, Oliveira SC, Miyoshi A, Azevedo V. Corynebacterium pseudotuberculosis: microbiology, biochemical properties, pathogenesis and molecular studies of virulence. Vet Res 2006; 37:201-18. 4. Silva AS, Baraúna RA, de Sá PC, das Gracas DA, Carneiro AR, Thouvenin M, Azevedo V, Badell E, Guiso N, da Silva AL, Ramos RT. Draft genome sequence of Corynebacterium ulcerans FRC58, isolated from the bronchitic aspiration of a patient in France. Genome Announc 2014; 2. 5. Gilbert R, Stewart FC. Corynebacterium ulcerans; a pathogenic microorganism resembling Corynebacterium diphtheriae. J Lab Clin Med 1927; 12:756-61. 6. Zakikhany K, Efstratiou A. Diphtheria in Europe: current problems and new challenges. Future Microbiol 2012; 7:595-607. 7. Meinel DM, Konrad R, Berger A, König C, Schmidt-Wieland T, Hogardt M, Bischoff H, Ackermann N, Hörmansdorfer S, Krebs S, Blum H, Margos G, Sing A. Zoonotic transmission of toxigenic Corynebacterium ulcerans strain, Germany, 2012. Emerg Infect Dis 2015; 21:356- 8. 8. Meinel DM, Margos G, Konrad R, Krebs S, Blum H, Sing A. Next generation sequencing analysis of nine Corynebacterium ulcerans isolates reveals zoonotic transmission and a novel putative diphtheria toxin-encoding pathogenicity island. Genome Med 2014; 6:113. 9. Berger A, Huber I, Merbecks SS, Ehrhard I, Konrad R, Hörmansdorfer S, Hogardt M, Sing A. Toxigenic Corynebacterium ulcerans in woman and cat. Emerg Infect Dis 2011; 17:1767-9. 10. Bergin IL, Chien CC, Marini RP, Fox JG. Isolation and characterization of Corynebacterium ulcerans from cephalic implants in macaques. Comp Med 2000; 50:530-5. 11. Contzen M, Sting R, Blazey B, Rau J. Corynebacterium ulcerans from diseased wild boars. Zoonoses Public Health 2011; 58:479-88. 12. Eisenberg T, Mauder N, Contzen M, Rau J, Ewers C, Schlez K, Althoff G, Schauerte N, Geiger C, Margos G, Konrad R, Sing A. Outbreak with clonally related isolates of Corynebacterium ulcerans in a group of water rats. BMC Microbiol 2015; 15:42. 13. Foster G, Patterson T, Howie F, Simpson V, Davison N, Efstratiou A, Lai S. Corynebacterium ulcerans in free-ranging otters. Vet Rec 2002; 150:524.

Hacker, E. 64

Publications

14. Hirai-Yuki A, Komiya T, Suzaki Y, Ami Y, Katsukawa C, Takahashi M, Yamamoto A, Yamada YK. Isolation and characterization of toxigenic Corynebacterium ulcerans from 2 closed colonies of cynomolgus macaques (Macaca fascicularis) in Japan. Comp Med 2013; 63:272-8. 15. Hogg RA, Wessels J, Hart J, Efstratiou A, De Zoysa A, Mann G, Allen T, Pritchard GC. Possible zoonotic transmission of toxigenic Corynebacterium ulcerans from companion animals in a human case of fatal diphtheria. Vet Rec 2009; 165:691-2. 16. Hommez J, Devriese LA, Vaneechoutte M, Riegel P, Butaye P, Haesebrouck F. Identification of nonlipophilic corynebacteria isolated from dairy cows with mastitis. J Clin Microbiol 1999; 37:954-7. 17. Morris WE, Uzal FA, Cipolla AL. Pyogranulomatous meningoencephalitis in a goat due to Corynebacterium ulcerans. Vet Rec 2005; 156:317-8. 18. Olson ME, Goemans I, Bolingbroke D, Lundberg S. Gangrenous dermatitis caused by Corynebacterium ulcerans in Richardson ground squirrels. J Am Vet Med Assoc 1988; 193:367-8. 19. Schuhegger R, Schoerner C, Dlugaiczyk J, Lichtenfeld I, Trouillier A, Zeller-Peronnet V, Busch U, Berger A, Kugler R, Hörmansdorfer S, Sing A. Pigs as source for toxigenic Corynebacterium ulcerans. Emerg Infect Dis 2009; 15:1314-5. 20. Seto Y, Komiya T, Iwaki M, Kohda T, Mukamoto M, Takahashi M, Kozaki S. Properties of corynephage attachment site and molecular epidemiology of Corynebacterium ulcerans isolated from humans and animals in Japan. Jpn J Infect Dis 2008; 61:116-22. 21. Sykes JE, Mapes S, Lindsay LL, Samitz E, Byrne BA. Corynebacterium ulcerans bronchopneumonia in a dog. J Vet Intern Med 2010; 24:973-6. 22. Tejedor MT, Martin JL, Lupiola P, Gutierrez C. Caseous lymphadenitis caused by Corynebacterium ulcerans in the dromedary camel. Can Vet J 2000; 41:126-7. 23. Dias AA, Silva FC, Jr., Pereira GA, Souza MC, Camello TC, Damasceno JA, Pacheco LG, Miyoshi A, Azevedo VA, Hirata R, Jr., Boas MH, Mattos-Guaraldi AL. Corynebacterium ulcerans isolated from an asymptomatic dog kept in an animal shelter in the metropolitan area of Rio de Janeiro, Brazil. Vector Borne Zoonotic Dis 2010; 10:743-8. 24. Mattos-Guaraldi AL, Sampaio JL, Santos CS, Pimenta FP, Pereira GA, Pacheco LG, Miyoshi A, Azevedo V, Moreira LO, Gutierrez FL, Costa JL, Costa-Filho R, Damasco PV, Camello TC, Hirata R, Jr. First detection of Corynebacterium ulcerans producing a diphtheria- like toxin in a case of human with pulmonary infection in the Rio de Janeiro metropolitan area, Brazil. Mem Inst Oswaldo Cruz 2008; 103:396-400. 25. Trost E, Al-Dilaimi A, Papavasiliou P, Schneider J, Viehoever P, Burkovski A, Soares SC, Almeida SS, Dorella FA, Miyoshi A, Azevedo V, Schneider MP, Silva A, Santos CS, Santos LS, Sabbadini P, Dias AA, Hirata R, Jr., Mattos-Guaraldi AL, Tauch A. Comparative analysis

Hacker, E. 65

Publications of two complete Corynebacterium ulcerans genomes and detection of candidate virulence factors. BMC Genomics 2011; 12:383. 26. Dias AA, Silva FC, Jr., Santos LS, Ribeiro-Carvalho MM, Sabbadini PS, Santos CS, Filardy AA, Myioshi A, Azevedo VA, Hirata R, Jr., Villas-Boas MH, Mattos-Guaraldi AL. Strain- dependent arthritogenic potential of the zoonotic pathogen Corynebacterium ulcerans. Vet Microbiol 2011; 153:323-31. 27. Hacker E, Ott L, Hasselt K, Mattos-Guaraldi AL, Tauch A, Burkovski A. Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources. Microbiology 2015; 161:1582-91. 28. Simpson-Louredo L, Ramos JN, Peixoto RS, Santos LS, Antunes CA, Ladeira EM, Santos CS, Vieira VV, Boas MH, Hirata R, Jr., Mattos-Guaraldi AL. Corynebacterium ulcerans isolates from humans and dogs: fibrinogen, fibronectin and collagen-binding, antimicrobial and PFGE profiles. Antonie Van Leeuwenhoek 2014; 105:343-52. 29. McKean S, Davies J, Moore R. Identification of macrophage induced genes of Corynebacterium pseudotuberculosis by differential fluorescence induction. Microbes Infect 2005; 7:1352-63. 30. dos Santos CS, dos Santos LS, de Souza MC, dos Santos Dourado F, de Souza de Oliveira Dias AA, Sabbadini PS, Pereira GA, Cabral MC, Hirata R, Jr., Mattos-Guaraldi AL. Non-opsonic phagocytosis of homologous non-toxigenic and toxigenic Corynebacterium diphtheriae strains by human U-937 macrophages. Microbiol Immunol 2010; 54:1-10. 31. Hirano T, Ishihara K, Hibi M. Roles of STAT3 in mediating the cell growth, differentiation and survival signals relayed through the IL-6 family of cytokine receptors. Oncogene 2000; 19:2548-56. 32. Solaroglu I, Cahill J, Jadhav V, Zhang JH. A novel neuroprotectant granulocyte-colony stimulating factor. Stroke 2006; 37:1123-8. 33. Hodgson AL, Krywult J, Corner LA, Rothel JS, Radford AJ. Rational attenuation of Corynebacterium pseudotuberculosis: potential cheesy gland vaccine and live delivery vehicle. Infect Immun 1992; 60:2900-5. 34. Antunes CA, Sanches Dos Santos L, Hacker E, Köhler S, Bösl K, Ott L, de Luna M, Hirata R, Jr., Azevedo VA, Mattos-Guaraldi AL, Burkovski A. Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae. Microbiology 2015; 161:639- 47. 35. Ott L, McKenzie A, Baltazar MT, Britting S, Bischof A, Burkovski A, Hoskisson PA. Evaluation of invertebrate infection models for pathogenic corynebacteria. FEMS Immunol Med Microbiol 2012; 65:413-21. 36. Bonmarin I, Guiso N, Le Fleche-Mateos A, Patey O, Patrick AD, Levy-Bruhl D. Diphtheria: A zoonotic disease in France? Vaccine 2009; 27:4196-200.

Hacker, E. 66

Publications

37. Wagner KS, White JM, Crowcroft NS, De Martin S, Mann G, Efstratiou A. Diphtheria in the United Kingdom, 1986-2008: the increasing role of Corynebacterium ulcerans. Epidemiol Infect 2010; 138:1519-30. 38. Indrigo J, Hunter RL, Jr., Actor JK. Cord factor trehalose 6,6'-dimycolate (TDM) mediates trafficking events during mycobacterial infection of murine macrophages. Microbiology 2003; 149:2049-59. 39. Sydor T, von Bargen K, Hsu FF, Huth G, Holst O, Wohlmann J, Becken U, Dykstra T, Söhl K, Lindner B, Prescott JF, Schaible UE, Utermöhlen O, Haas A. Diversion of phagosome trafficking by pathogenic Rhodococcus equi depends on mycolic acid chain length. Cell Microbiol 2013; 15:458-73. 40. Hard GC. Comparative toxic effect of the surface lipid of Corynebacterium ovis on peritoneal macrophages. Infect Immun 1975; 12:1439-49. 41. Tashjian JJ, Campbell SG. Interaction between caprine macrophages and Corynebacterium pseudotuberculosis: an electron microscopic study. Am J Vet Res 1983; 44:690-3. 42. Aki T, Funakoshi T, Uemura K. Regulated necrosis and its implications in toxicology. Toxicology 2015; 333:118-26. 43. Sridharan H, Upton JW. Programmed necrosis in microbial pathogenesis. Trends Microbiol 2014; 22:199-207. 44. Both L, Collins S, de Zoysa A, White J, Mandal S, Efstratiou A. Molecular and epidemiological review of toxigenic diphtheria infections in England between 2007 and 2013. J Clin Microbiol 2015; 53:567-72. 45. Axelrod S, Oschkinat H, Enders J, Schlegel B, Brinkmann V, Kaufmann SH, Haas A, Schaible UE. Delay of phagosome maturation by a mycobacterial lipid is reversed by nitric oxide. Cell Microbiol 2008; 10:1530-45. 46. Moreira LO, Mattos-Guaraldi AL, Andrade AF. Novel lipoarabinomannan-like lipoglycan (CdiLAM) contributes to the adherence of Corynebacterium diphtheriae to epithelial cells. Arch Microbiol 2008; 190:521-30. 47. Wiesmann V, Reimer D, Franz D, Hüttmayer H, Mielenz D, Wittenberg T. Automated high-throughput analysis of B cell spreading on immobilized antibodies with whole slide imaging. Current Directions in Biomedical Engineering 2015; 1:224–7. 48. Held C. Towards increased efficiency and automation in fluorescence micrograph analysis based on hand-labeled data. Bielefeld: Universität Bielefeld, 2013. 49. Hartigan JA, Wong MA. Algorithm AS 136: A K-means Clustering Algorithm. Journal of the Royal Statistical Society Series C (applied Statistics) 1979; 28:100-8. 50. Abe S, Takayama K, Kinoshita S. Taxonomical studies on glutamic acid producing bacteria. J Gen Microbiol 1967; 13:279-301.

Hacker, E. 67

Publications

51. Tsuchiya S, Yamabe M, Yamaguchi Y, Kobayashi Y, Konno T, Tada K. Establishment and characterization of a human acute monocytic leukemia cell line (THP-1). Int J Cancer 1980; 26:171-6. 52. Ralph P, Nakoinz I. Antibody-dependent killing of erythrocyte and tumor targets by macrophage-related cell lines: enhancement by PPD and LPS. J Immunol 1977; 119:950-54. 53. Fiani ML, Beitz J, Turvy D, Blum JS, Stahl PD. Regulation of mannose receptor synthesis and turnover in mouse J774 macrophages. J Leukoc Biol 1998; 64:85-91. 54. Knoppová M, Phensaijai M, Veselý M, Zemanová M, Nesvera J, Pátek M. Plasmid vectors for testing in vivo promoter activities in Corynebacterium glutamicum and Rhodococcus erythropolis. Curr Microbiol 2007; 55:234-9.

Hacker, E. 68

Publications

4.3 Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae (Antunes et al., 2015a)

Hacker, E. 69

Publications

Microbiology (2015), 161, 639–647 DOI 10.1099/mic.0.000020

Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae

Camila Azevedo Antunes,1,23 Louisy Sanches dos Santos,33 Elena Hacker,1 Stefanie Ko¨hler,1 Korbinian Bo¨sl,1 Lisa Ott,1 Maria das Grac¸as de Luna,3 Raphael Hirata, Jr,3 Vasco Ariston de Carvalho Azevedo,2 Ana-Luı´za Mattos-Guaraldi3 and Andreas Burkovski1

Correspondence 1Mikrobiologie, Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg, Erlangen, Germany Andreas Burkovski 2Departmento de Biologia, Instituto de Cieˆncias Biolo´gicas, Universidade Federal de Minas Gerais, [email protected] Belo Horizonte, Minas Gerais, Brazil 3Faculdade de Cieˆncias Me´dicas, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, Brazil

Corynebacterium diphtheriae is typically recognized as an extracellular pathogen. However, a number of studies revealed its ability to invade epithelial cells, indicating a more complex pathogen–host interaction. The molecular mechanisms controlling and facilitating internalization of Cor. diphtheriae are poorly understood. In this study, we investigated the role of DIP0733 as virulence factor to elucidate how it contributes to the process of pathogen–host cell interaction. Based on in vitro experiments, it was suggested recently that the DIP0733 protein might be involved in adhesion, invasion of epithelial cells and induction of apoptosis. A corresponding Cor. diphtheriae mutant strain generated in this study was attenuated in its ability to colonize and kill the host in a Caenorhabditis elegans infection model system. Furthermore, the mutant showed an altered adhesion pattern and a drastically reduced ability to adhere and invade epithelial cells. Subsequent experiments showed an influence of DIP0733 on binding of Cor. diphtheriae to extracellular matrix proteins such as collagen and fibronectin. Furthermore, based on its fibrinogen-binding activity, DIP0733 may play a role in avoiding recognition of Cor. diphtheriae by Received 5 December 2014 the immune system. In summary, our findings support the idea that DIP0733 is a multi-functional Accepted 19 December 2014 virulence factor of Cor. diphtheriae.

INTRODUCTION toxaemic infection of respiratory tract that can be fatal. Moreover, it can cause skin ulcers (cutaneous diphtheria) as The genus Corynebacterium belongs to the class of Actino- well as systemic infections such as endocarditis, meningitis, bacteria (i.e. Gram-positive bacteria with high G+CDNA pneumonia and others (Murphy, 1996; Burkovski, 2013b). content) and comprises a collection of irregular- or club- This indicates that Cor. diphtheriae is able to colonize not shaped (micro)aerobic bacteria (Ventura et al., 2007; Zhi only epithelia but also deeper parts of the body and that the et al., 2009). Today, 88 Corynebacterium species are described bacteria interact with various types of host cells. The ability with about 53 having a more or less pronounced medical of Cor. diphtheriae to enter cultured respiratory epithelial importance (Bernard, 2012). The most prominent and impor- cells was first shown by Hirata et al. (2002) and validated tant member of the pathogenic species is Corynebacterium by a number of further studies (Mattos-Guaraldi, 2002; diphtheriae, which is also the type species of the genus. Bertuccini et al., 2004; Ott et al., 2010a, b, 2013). Cor. diphtheriae is the classical aetiological agent of diph- Due to its medical importance, Cor. diphtheriae might theria (Hadfield et al., 2000; Burkovski, 2013a, b), a localized be the best-investigated pathogenic member of the genus; however, even for this species only a few virulence factors 3These authors contributed equally to this paper. have been characterized in detail. Beside the diphtheria Abbreviations: ECM, extracellular matrix; MSCRAMM, microbial surface toxin these include mainly pili and a few other adhesion components recognizing adhesive matrix molecule; PBST, PBS contain- factors (see Collier, 2001; Rogers et al., 2011; Readon- ing Tween. Robinson & Ton-That, 2013; Ott & Burkovski, 2013 for Three supplementary figures and two supplementary tables are available reviews), while little is known about the factors mediating with the online Supplementary Material. the entry processes in the host cell and the receptors

000020 G 2015 The Authors Printed in Great Britain 639

Hacker, E. 70 Publications

C. A. Antunes and others recognized. An interesting protein in respect to its role in pCR2.1-TOPO-DIP0733 was used to transform E. coli TOP10 Cor. diphtheriae pathogenicity is DIP0733, which was Electrocomp cells. Subsequently, 1 mg unmethylated plasmid DNA initially described as non-fimbrial surface protein 67-72p isolated from this E. coli strain was used to transform Cor. diphtheriae using a GenePulser II apparatus (Bio-Rad). Electroporated cells were according to its appearance as two bands with distinct added to 1 ml TSB and incubated for 4 h at 37 uC. This culture was apparent masses in SDS-PAGE (Colombo et al., 2001). Both plated on TSB agar plates containing kanamycin. Since pCR2.1-TOPO the 67 and 72 kDa polypeptide bind to erythrocyte recep- cannot autonomously replicate in Cor. diphtheriae, kanamycin- tors, leading to haemagglutination. Subsequently, Hirata et al. resistant Cor. diphtheriae carried the vector integrated via homologous (2004) demonstrated that 67-72p binds not only human recombination in the chromosomal CDC-E8392_0678 gene. The erythrocytes but also the human epidermoid laryngeal insertion of the plasmid into the chromosome of Cor. diphtheriae carcinoma cell line HEp-2 and that binding was effectively CDC-E8392 strain was confirmed by PCR employing different combinations of primers and Southern blotting (see Fig. S1, available blocked by anti-67-72p IgG antibodies. A recent MS study in the online Supplementary Material). The confirmed strain was indicated that the 67-72p protein is encoded by a single gene, designated CAM-1. dip0733 (Sabbadini et al., 2012). Furthermore, an effect of the protein on internalization and induction of apoptosis was For overexpression and complementation, the complete DIP0733 gene sequence was amplified by PCR (primers: 59-CGCGCCTGCAGGTT- suggested based on experiments using 67-72p protein-coated GGCGACCGGTTTTACGCG-39 and 59-CGCGCCCGGGTT ACTGC- latex beads (Sabbadini et al., 2012). Despite the different CCAGCAACCTTGC-39) and the product was digested with the properties attributed to the protein, i.e. haemagglutination, restriction enzymes SbfI and XmaI for 1.5 h at 37 uC. The vector binding to host cell receptors and induction of apoptosis, the pXMJ19 (Jakoby et al., 1999) was digested with the same enzymes in protein sequence is lacking obvious functional domains. As a parallel and dephosphorylated with Rapid Phosphatase (Roche) for u further step of a molecular characterization of this 30 min at 37 C. Ligation of the insert into the pXMJ19 plasmid was carried out with T4 DNA ligase overnight at room temperature. After functionally astonishingly diverse protein, we studied the transformation of E. coli DH5a MCR, the positive clones were selected 2 interaction of 67-72p wild-type and mutant strains with on LB medium containing 25 mg chloramphenicol ml 1. The resulting Caenorhabditis elegans, an invertebrate model system, as well plasmid pXMJ19-DIP0733 was isolated, DNA sequenced and trans- as with epithelial cells and macrophages. In addition, we formed in Cor. diphtheriae as described above. For verification of investigated if DIP0733 is involved in interactions with DIP0733 transcription, see Fig. S2. extracellular matrix components fibronectin and type I For fluorescence microscopy experiments, plasmid pXMJ19-DIP0733- collagen and with human plasma fibrinogen. In summary, mCherry was constructed carrying the DIP0733 gene between the our results indicate that DIP0733 can be considered as a tac promoter and mCherry. For this purpose, a 0.7 kb KpnI/EcoRI microbial surface components recognizing adhesive matrix DNA fragment carrying the mCherry gene was isolated from molecule (MSCRAMM) with virulence properties. pXMJ19mCherry and ligated downstream of the DIP0733 gene in plasmid pXMJ19-DIP0733.

Adherence and internalization assays. HeLa cells were seeded in METHODS 24-well plates (Nunc) with 56104 cells per well 48 h prior to infec- tion. Bacteria were inoculated to an OD of 0.1 from overnight Reconstruction of phylogenetic trees. Sequences with amino acid 600 cultures and grown in HI broth to an OD of 0.4 to 0.6. Sub- identities of 45 % or higher for multiple alignment were obtained by a 600 sequently, the bacteria were harvested by centrifugation and cell BLAST of DIP0733 sequence versus the UniProtKB database using density was adjusted to an OD600 of 0.5. A master mix of the default values. The multiple alignment was performed by CLUSTAL W inoculum with a m.o.i. of 50 was prepared in Dulbecco’s modified 2.0 (Larkin et al., 2007). Phylogenetic trees were calculated by the Eagle’s medium (DMEM) and 500 ml per well was used to infect the neighbour joining method using the Jones–Taylor–Thirnton distance cells. The plates were centrifuged for 5 min at 500 r.p.m. to matrix model with a transition/transversion ratio of 2.0. Evolutionary synchronize infection and subsequently incubated for 90 min. The distance correction was used as indicated. cells were washed with PBS three times, detached with 500 ml Accutase per well (5 min at 37 uC with 5 % CO and 95 % humidity) Bacterial strains and culture conditions. Strains used in this study 2 and lysed with 0.025 % Tween 20 under same conditions. Serial are listed in Table 1. Escherichia coli OP50 and TOP10 Electrocomp dilutions were made in pre-chilled 16 PBS and plated on Columbia were grown in Luria–Bertani (LB) medium at 37 uC (Sambrook et al., agar with sheep blood (Oxoid) to determine the number of c.f.u. 1989). Cor. diphtheriae strains were grown in heart infusion (HI) broth or trypticase soy broth (TSB) at 37 uC and stored in TSB For invasion analysis, the cells were washed three times with PBS to medium with 20 % glycerol at 280 uC. When appropriate, kanamy- remove planktonic and loosely attached bacteria. Subsequently, the 2 cin was added (50 mg kanamycin ml 1 final concentration). cells were incubated for 2 h in DMEM (500 ml per well), containing 2 100 mg gentamicin ml 1 to kill remaining extracellular bacteria. After Molecular biology methods. Standard techniques were used for this incubation, the cell layers were washed three times with PBS, plasmid isolation, transformation and cloning (Sambrook et al., detached by adding 500 ml trypsin solution (0.12 % trypsin, 0.01 % 1989). For chromosomal disruption of the Cor. diphtheriae DIP0733 EDTA in PBS) per well (5 min at 37 uC with 5 % CO2 and 95 % gene, a 1068 bp internal DNA fragment was amplified by PCR. humidity) and lysed for 5 min at 37 uC with 0.025 % Tween 20 to Chromosomal DNA of strain CDCE8392 was used as template and liberate the intracellular bacteria. Serial dilutions of the inocula and the following primers: 59-CACCATGGAGCGTTTCTCTGTTTC-39, the lysates were plated out on Columbia agar with sheep blood 59-CTGCCGTCGTAGCTGTCCAC-39. The DNA fragment was (Oxoid) to determine the number of c.f.u. ligated to the overhanging 39 deoxythymidine (T) of pCR2.1-TOPO linearized vector exploiting the Taq polymerase’s nontemplate- Infection of THP-1 cells. Human THP-1 cells were cultured in 10 % dependent terminal transferase activity that adds a single deoxyade- FBS supplemented RPMI medium 1640 (containing 100 U penicillin 21 21 nosine (A) to the 39 ends of PCR products. The resulting plasmid ml and 100 mg streptomycin ml )at37uCin5%CO2 in a

640 Microbiology 161

Hacker, E. 71 Publications

Characterization of C. diphtheriae DIP0733 protein

Table 1. Strains, cell lines and plasmids used in this study

Strains Description/genotype Reference/source Corynebacterium diphtheriae + CDC-E8392 Biovar mitis; tox Trost et al. (2012) CAM-1 CDC-E8392 DIP0733 :: pCR2.1-TOPO ‘DIP0733’ This study Escherichia coli 2 TOP10 Electrocomp F mcrA D(mrr-hsdRMS-mcrBC) w80lacZDM15 DlacX74 recA1 Invitrogen 2 araD139 D(ara-leu) 7697 galU galK rpsL (StrR) endA1 nupG l OP50 Uracil-auxotrophic E. coli B strain Brenner (1974) Cell lines HEp-2 Human epidermoid larynx carcinoma cells Hirata et al. (2004) HeLa Human cervical carcinoma cells Gey et al. (1952); Scherer et al. (1953) THP-1 Human monocytic cell line derived from an acute monocytic Tsuchiya et al. (1980) leukaemia Plasmids R R pCR 2.1-TOPO oripUC, orif1, lacZa, pM13, pT7, Km ,Ap Invitrogen pCR2.1-TOPO-DIP0733 pCR 2.1-TOPO containing a 1068 bp internal DNA fragment This study R pXMJ19 ori colE1, oricg, ptac,Cm Jakoby et al. (1999) R pXMJ19-DIP0733 ori colE1, oricg, ptac, CDCE8392_0678, Cm Kindly provided by J. Kecskes, Erlangen R pXMJ19mCherry ori colE1, oricg, ptac, mCherry,Cm Ott et al. (2012) pXMJ19-DIP0733-mCherry pXMJ19 carrying the DIP0733 gene cloned between ptac and This study mCherry

humidified cell culture incubator. Before infection, THP-1 cells were Extracellular matrix (ECM) and plasma protein binding assays. cultured in antibiotic-free medium containing 10 % FBS and were Bacterial binding to type I collagen, fibronectin and biotinylated 2 differentiated into macrophage-like cells using 10 ng PMA ml 1 for fibrinogen (Sigma) was performed in 96-well ELISA microtitre plates 24 h. The cells were infected with Cor. diphtheriae CDC-E8392 and (Costar 96 Well EIA/RIA plate; Corning). Bacterial cultures grown the corresponding mutant CAM-1 at m.o.i. of 10 or left uninfected. for 24 h at 37 uC in TSB medium were washed twice with PBS and The plates were centrifuged for 5 min at 500 r.p.m. to synchronize resuspended in 0.1M NaHCO3, pH 9.6, to an OD650 of 0.2 2 infection. After incubation for 30 min, the medium was aspirated and (equivalent to 56109 c.f.u. ml 1). The wells were incubated with cells were treated first with medium containing 100 mg gentamicin 100 ml of bacterial suspensions for 1 h at 37 uC and then overnight at 21 ml for 1 h. Then, the medium with a lower gentamicin concentration 8 uC. A standard curve was performed by diluting the biotinylated 21 2 (10 mg gentamicin ml ) was added and cells were incubated until they protein solutions to concentrations varying from 5 to 0.05 mg ml 1 were harvested (1, 7 or 19 h). The supernatant was removed and cells (1 h at 37 uC). After blocking with 2 % BSA (BSA type V; Sigma) in were lysed with 500 ml 0.1 % Triton X-100. Serial dilutions of the PBS containing 0.05 Tween-20 (PBST) for 1 h at 37 uC, the wells were inocula and the lysates were plated out on Columbia agar with sheep washed three times with PBST. The bacterial strains were subsequently 2 blood (Oxoid) to determine the number of c.f.u. incubated with 20 mg biotinylated ECM/plasma proteins ml 1 at 37 uC. After three washes with PBST, the wells were incubated for Cae. elegans. 2 Infection and colonization of Cae. elegans N2 was 30 min at 37 uC with 0.001 g extravidin-peroxidase ml 1 (Sigma) maintained on E. coli OP50 for 6 to 7 days until the worms became prepared in PBST containing 1 % BSA. After washing three times, starved, as indicated by clumping behaviour (de Bono & Bargmann, 3,39,5,59-tetramethylbenzidine liquid substrate was added. After 1998). Subsequently, the nematodes were infected with different Cor. 20 min the reaction was stopped with 50 ml of 1 M HCl and colour diphtheriae strains. Infection of L4 stage larval worms was carried out development was measured at l5450 nm in a microtitre plate reader. with 20 ml of each bacterial strain (from an overnight culture) on u The colour intensity of the wells sensitized with the micro-organisms NGM plates at 21 C for 24 h. Worms were assessed each day was compared to the standard curve by GraphPad Prism, version 6.0. following infection and the dead nematodes were counted and The results were expressed in micrograms of adhered proteins, with a removed every 24 h. For each strain, approximately 60 nematodes mean±SD of three independent assays, each performed in triplicate. were used and the assays were performed three times. The Kaplan– The mean of the binding was compared by Tukey’s multiple Meier survival analysis was used and all statistical analyses were comparison test (Simpson-Louredo et al., 2014). performed with Prism 5.0 (GraphPad), with P values of less than 0.05 considered significant. The isolation of bacteria after colonization of worms was carried out RESULTS as previously described (Garsin et al., 2001). In short, to remove the surface bacteria, the worms were washed twice in 4 ml M9 medium and transferred to new NGM plates supplemented with 25 mg Distribution of DIP0733 genes 21 nalidixic acid ml . Subsequently, 10 nematodes were placed in Using MALDI-TOF analyses 67-72p was identified with 1.5 ml reaction tubes containing 10 ml PBS with 1 % Triton X-100 and were mechanically disrupted by using a pestle. Then, 200 mlof significant scores as the protein DIP0733 (Sabbadini et al., 1 % Triton X-100 was added and 100 ml aliquots of the liquid were 2012). Corresponding genes are widely distributed among plated on HI agar with and without antibiotics. members of the genera Corynebacterium, Mycobacterium http://mic.sgmjournals.org 641

Hacker, E. 72 Publications

C. A. Antunes and others

and Nocardia (CMN group of Actinobacteria), but corre- investigation of Cor. diphtheriae pathogenicity and viru- sponding genes were also found in the genomes of Amyco- lence factors (Ott et al., 2012; Broadway et al., 2013). latopsis, Dietzia and Rhodococcus species. Taken together, When Cae. elegans were fed with E. coli OP50 carrying more than 250 homologous genes were found in genome plasmid pXMJ19mCherry, no fluorescence was detectable sequence data collections, both in pathogenic and in non- after the nematodes were set back on agar plates containing pathogenic Actinobacteria. Phylogenetic analyses revealed unlabelled E. coli, since the bacteria were unable to colonize distinct clusters of closely related members of this group the nematodes and were completely digested (Fig. 2a). within the genera Corynebacterium, Gordonia, Amyco- When the nematodes were fed with mCherry-labelled Cor. latopsis, Nocardia and Rhodococcus, while members of the diphtheriae wild-type CDC-E8392, the complete gut was genus Mycobacterium form two distinct clusters [overview fluorescent even after Cae. elegans were set back on agar Fig. 1, see Fig. S3 for the reconstructed phylogenetic tree plates containing unlabelled OP50 (Fig. 2b), thus indicating (1.000 actinobacterial DIP0733-like proteins extracted persistent colonization. In contrast, Cae. elegans infected from databases)]. In several Mycobacterium and Rhodo- with DIP0733 gene disruption mutant CAM-1 revealed an coccus species two or three paralogues were found (Table almost complete clearance of mCherry-expressing bacteria S1), while a number of Actinobacteria lack DIP0733 genes after transfer to plates with unlabelled E. coli. Only slight (Table S2). background fluorescence was detectable in this case, indicating a significantly impaired colonization by CAM-1 Interaction with Cae. elegans pXMJ19mCherry (Fig. 2c). This defect was complemented at least partially, when strain CAM-1 was transformed with A number of different model systems including amoeba, plasmid pXMJ19-DIP0733-mCherry (Fig. 2d). nematodes, insects, fishes and mammals have been introduced to study host–pathogen interactions. Besides As a more quantitative assay and to avoid putative guinea pigs already applied by Loeffler, wax moth larvae misinterpretation caused by fluorescent protein, which (Galleria mellonella) and the nematode Cae. elegans were might remain in the gut after digestion of bacteria, c.f.u. established as simple surrogate model systems for the were determined from Cae. elegans infected with Cor.

Corynebacterium pseudotuberculosis (Cp258_0559) Corynebacterium ulcerans (D881_04070) 0.015 Corynebacterium diphtheriae (DIP0733) Corynebacterium glutamicum (KIQ_14142) Corynebacterium ammoniagenes (HMPREF0281_01354) Corynebacterium jeikeium (HMPREF0297_0391) Corynebacterium resistens (CRES_1693) Corynebacterium kroppenstedtii (ckrop_1465) Gordonia otitidis (GOOTI_232_00130) Gordonia bronchialis (Gbro_3487) Mycobacterium smegmatis (MSMEG_1959/MSMEI_1915) Mycobacterium marinum (MMEU_5025) Mycobacterium abscessus (MAB_3498c) Amycolatopsis mediterranei (AMES_1239) Nocardia brasiliensis (O3I_034585) Nocardia farcinica (NFA_45260) Rhodococcus erythropolis (RER_22310) Rhodococcus equi (HMPREF0724_13235) Mycobacterium tuberculosis (CCDC5180_0058) Mycobacterium bovis BCG (BCG_0095)

Fig. 1. Phylogenetic tree of the DIP0733 family of proteins. Unrooted phylogenetic trees were reconstructed with the CLUSTAL W software making use of the implemented neighbour joining method with the function for evolutionary distance correction. Evolutionary distances are proportional to the branch length.

642 Microbiology 161

Hacker, E. 73 Publications

Characterization of C. diphtheriae DIP0733 protein

(a) (a) 100

80

60

Survival (%) 40

(b) 20

0 0321 7654 8 Time (days) (b) 100

80 (c) 60

40 Survival (%)

20

0 (d) 0321 7654 8 Time (days)

Fig. 3. Nematode survival assay. (a) Infection with E. coli OP50 (&), Cor. diphtheriae CDC-E8392 (m) and CAM-1 ($). (b) Complementation and overexpression assay. E. coli OP50 (&), CAM-1 pXMJ19 (#), CAM-1 pXMJ19-DIP0733 (7), CDC- E8392 pXMJ19 (D) and CDC-E8392 pXMJ19-DIP0733 (2). Data shown are the mean of three independent experiments; error Fig. 2. Fluorescence microscopy of Cae. elegans colonization. bars represent SD. Nematodes infected with bacteria were mounted onto agar pads, paralysed with 0.6 % 2-phenoxy-2-propanol and photographed using a Leica SP5 II laser microscope. In each of three independent experiments, approximately 20 worms were infected. Representative results are shown. Cae. elegans fed with (a) E. successful complementation and overexpression of DIP0733 coli OP50 pXMJ19mCherry, (b) Cor. diphtheriae CDC-E8392 (Fig. 3b). In summary, disruption of the DIP0733 gene pXMJ19mCherry, (c) CAM-1 pXMJ19mCherry and (d) CAM-1 reduced host colonization and killing dramatically in the pXMJ19-DIP0733-mCherry. Tails of nematodes are indicated by arrows. Cae. elegans model system.

Influence on adhesion to epithelial cells diphtheriae strains. After 24 h of colonization, about 10- As a more quantitative approach to characterize adhesion, fold higher numbers of c.f.u. were obtained from lysed the adhesion rate of wild-type CDC-E8392 and DIP0733 worms infected with wild-type CDC-E8392 compared to mutant strain CAM-1 to HeLa cells was determined. While mutant strain CAM-1 (254±189 versus 30±13 c.f.u. CDC-E8392 reached an adhesion rate of 25.97±7.06 %, 2 ml 1). Moreover, the mutant bacteria were, in contrast CAM-1 attachment was reduced to only 6.93±2.98 %. to the wild-type, less detrimental to Cae. elegans.Ina Similar rates were obtained with HEp-2 cells (data not nematode survival assay, about 90 % of the nematodes shown). The rates were unchanged when the corresponding survived contact with the mutant strain CAM-1, while strains were transformed with Cor. diphtheriae plasmid 70 % were killed by the wild-type CDC-E8392 within pXMJ19 (24.72±3.91 % for the wild-type and 9.43±3.61 % 7 days (Fig. 3a). The mutant strain and wild-type showed for CAM-1), while transformation with DIP0733 expression enhanced nematode toxicity when carrying DIP0733 expres- vector pXMJ19-DIP0733 led to a significantly enhanced sion vector pXMJ19-DIP0733, while the empty vector adhesion in wild-type CDC-E8392 and mutant strain control pXMJ19 had no significant influence, indicating compared to the plasmid-free strains. Overexpression of

http://mic.sgmjournals.org 643

Hacker, E. 74 Publications

C. A. Antunes and others

DIP0733 led to an adhesion rate of 59.63±9.81 % for the Binding to extracellular matrix proteins and wild-type and 49.87±8.81 % for CAM-1. Obviously, fibrinogen complementation of the DIP0733 mutation was possible Four main fibrous proteins are constituents of the extra- and overexpression resulted in strongly enhanced adherence cellular matrix surrounding eukaryotic cells, collagen, fibro- to epithelial cells (Fig. 4a). nectin, laminin and elastin. Pathogenic bacteria developed so-called MSCRAMMs, which interact with these proteins (Chagnot et al., 2012). Also for Cor. diphtheriae an inter- action with collagen and fibronectin was reported n=22, n=6, n=8 (Sabbadini et al., 2010). (a) 80.00 ** ND To test binding to extracellular matrix components, bacteria * were coupled to microtitre plate wells and incubated with 60.00 type I collagen and fibronectin. Both strains adhered to the plates with the same coating efficiency as determined by 40.00 Sorensen & Brodbeck (1986). Compared to the wild-type CDC-E8392, collagen binding of a DIP0733 mutant CAM-1 Adhesion (%) 20.00 was significantly impaired. Binding to collagen, which might be especially crucial for Cor. diphtheriae strains causing septic arthritis (Puliti et al., 2006), was reduced by about 0.00 40 % (0.197±0.001 mg/56108 of c.f.u. wild-type CDC- 8 (b) 0.8 E8392 and 0.104±0.005 mg/5610 c.f.u. of CAM-1). Furthermore, the binding to fibronectin was at least slightly reduced (0.232±0.002 mg/56108 c.f.u. of the wild-type and 0.6 0.182±0.001 mg/56108 c.f.u. of CAM-1) (Fig. 5). In summary, DIP0733 seems to function as a Cor. diphtheriae 0.4 MSCRAMM and the impaired adhesion to epithelial cells might be at least partially the result of disturbed ECM Adhesion (%) 0.2 recognition by the mutant. Fibrinogen is a major component of the human plasma 0.0 and is crucial for blood clot formation due to its con- version into insoluble fibrin, a process that is hijacked by

CAM-1 many pathogens. Sabbadini et al. (2010) already demon- strated that Cor. diphtheriae is able to bind to fibrinogen CDC-E8392 CAM-1 pXMJ19MJ19-DIP0733 and convert it to fibrin. In an approach similar to the pX binding assays described above, DIP0733 disruption strain CDC-E8392 pXMJ19 3 CAM-1 pXMJ19-DIP0733 CAM-1 showed ~50 % reduced binding to fibrinogen compared to the wild-type (0.417±0.029 mg/56108 c.f.u. CDC-E8 92

Fig. 4. Adhesion to and invasion of epithelial cells. (a) HeLa cells 0.5 were infected with Cor. diphtheriae strains, washed with PBS, detached with trypsin solution, and lysed with Tween 20, and the 0.4 number of c.f.u. was determined. Adhesion is expressed as percentage of the inoculum, showing means and SD of 22, 6 and 8 independent measurements (biological replicates) with three 0.3 samples each (technical replicates) for the strains indicated. (b) Invasion of epithelial cells. HeLa cells were infected for 2 h with 0.2 * *

” Binding ( m g) Cor. diphtheriae strains, treated with 100 mg gentamicin ml 1, * washed with PBS, detached with trypsin solution, and lysed with 0.1 Tween 20, and the number of c.f.u. was determined. Invasion is expressed as percentage of the inoculum, showing means and SD 0.0 of 22, 6 and 8 independent measurements (biological replicates) Collagen Fibronectin Fibrinogen with three samples each (technical replicates) for wild-type and mutant strains without plasmid, transformed with the empty plasmid and overexpression plasmid, respectively. Statistically Fig. 5. Binding to extracellular matrix proteins and fibrinogen. relevant differences between the strains (based on Student’s t- Binding to type I collagen, fibronectin and human fibrinogen by test) are indicated by asterisks above columns (**P,0.005 and Cor. diphtheriae. Black and white bars represent CDC-E8392 and *P,0.05). ND, Not determined. CAM-1, respectively. For statistic evaluation, see legend to figure 4.

644 Microbiology 161

Hacker, E. 75 Publications

Characterization of C. diphtheriae DIP0733 protein

of the wild-type and 0.192±0.003 mg/56108 c.f.u. of summary, the data obtained indicate that the protein has a CAM-1) (Fig. 5), indicating the important role of the beneficial effect on the establishment of Cor. diphtheriae DIP0733 in fibrinogen binding. within macrophages.

Invasion of epithelial cells DISCUSSION Fibronectin is used as a surface receptor by dif- ferent pathogenic bacteria, but also triggers internalization In the study presented here, we investigated the role of of various species, e.g. Staphylococcus aureus or Strep- DIP0733 as virulence factor in Cor. diphtheriae. The tococcus pyogenes (see Hauck & Ohlsen, 2006; Schwarz- corresponding gene is widely distributed among patho- Linek et al., 2006 for reviews). The underlying mechanisms genic and non-pathogenic corynebacteria, but also found of internalization of Cor. diphtheriae by epithelial cells are currently not clear; however, a positive influence of DIP0733 on internalization was proposed based on studies (a) 4 using isolated proteins and antibodies directed against the protein (Sabbadini et al., 2012). When the internalization 3 of wild-type and DIP0733 mutant strain was determined using gentamicin protection assays, the mutant showed a significantly decreased internalization by HeLa cells. Invasion 2 rates of 0.19±0.03 % were reached for the wild-type CDC- 1 E8392 and 0.03±0.01 % for DIP0733 mutant strain CAM-1. Internalization (%) Similar results were obtained for HEp-2 cells (data not shown). As in the case of adhesion assays, transformation of 0 wild-type or strain CAM-1 with vector pXMJ19 (empty vector control) had no influence on invasion rate (0.16± (b) 2.5 0.07 % and 0.04±0.03 %, respectively). In contrast, trans- formation with DIP0733 expression plasmid pXMJ19- 2.0 DIP0733 resulted in increased invasion rates, which 1.5 exceeded wild-type rates by a factor of about three with 0.57±0.11 % and 0.48±0.15 %, respectively (Fig. 4b). The 1.0 results obtained support a direct functional role of the DIP0733 protein in respect to the internalization of Cor. Internalization (%) 0.5 diphtheriae by epithelial cells. 0

(c) Interaction with macrophages 10

%) 8 Previous studies with isolated DIP0733 suggested an influ- –5 ence of the protein not only on the interaction of Cor. 6 diphtheriae with extracellular matrix and blood plasma proteins or with epithelial cells but also with macrophages. 4 When the intracellular number of wild-type CDC-E8392 and DIP0733 mutant strain CAM-1 was determined after 2 Internalization (×10 uptake by THP1 macrophage cells, a clear detrimental effect 0 of the mutation was observed. Two hours after infection, an internalization rate of 1.65±0.48 % was determined for wild-type strain CDC-E8392, while strain CAM-1 reached CAM-1 only 0.69±0.19 %. Transformation of wild-type or strain CDC-E8392 CAM-1 pXMJ19 CAM-1 with vector pXMJ19 (empty vector control) had no influence on survival or replication within the macrophages CDC-E8392 pXMJ19 CAM-1 pXMJ19-DIP0733 (1.52±0.66 % and 1.07±0.06 %, respectively), while trans- formation with DIP0733 expression plasmid pXMJ19- CDC-E8392 pXMJ19-DIP0733 DIP0733 resulted in increased numbers of bacteria, which ± exceeded wild-type rates by a factor of about two (2.92 Fig. 6. Interaction of Cor. diphtheriae with macrophages. Human 0.82 % and 1.69±0.57 %, respectively). The number of c.f.u. THP-1 cells were differentiated into macrophage-like cells using decreased over time for all strains; however, after 8 and 20 h PMA (phorbol 12-myristate 13-acetate) and infected at an m.o.i. of of infection viable bacteria were still detectable. Again, 10 with Cor. diphtheriae CDC-E8392 and DIP0733 mutant CAM- deletion of DIP0733 decreased and overexpression increased 1. Intracellular survival (a) 2, (b) 8 and (c) 20 h after infection the number of c.f.u. within the macrophage (Fig. 6). In expressed as a percentage of bacteria used for infection.

http://mic.sgmjournals.org 645

Hacker, E. 76 Publications

C. A. Antunes and others in other members of the Actinobacteria. This distribution Only seven transmembrane helices were identified and also implicates that DIP0733 has other physiological functions these make handling of the protein more difficult. Future besides its role as virulence factor in pathogenic coryne- work might therefore rely on extensive mutagenesis studies bacteria such as Cor. diphtheriae or the closely related Cor. including truncated DIP0733 forms. ulcerans. As a first approach, an invertebrate model system was ACKNOWLEDGEMENTS applied to address the influence of DIP0733 on host inter- action. Compared to the wild-type CDC-E8392, DIP0733 The CAPES fellowship of the Brasilian Science Foundation CA is mutant CAM-1 exhibited significant attenuation in nema- gratefully acknowledged. tode colonization, proliferation inside the worms and killing of the host. The effect of DIP0733 on host colonization was REFERENCES further investigated and confirmed by experiments using epithelial cells. Adherence and internalization assays re- Bernard, K. (2012). The genus corynebacterium and other medically vealed decreasing numbers of c.f.u.s of the mutant strain relevant coryneform-like bacteria. J Clin Microbiol 50, 3152–3158. compared to the wild-type. In addition, complementation Bertuccini, L., Baldassarri, L. & von Hunolstein, C. (2004). and overexpression approaches were successful and resulted Internalization of non-toxigenic Corynebacterium diphtheriae by in strains with increased adhesion and internalization rates cultured human respiratory epithelial cells. Microb Pathog 37, 111– implicating a direct role of DIP0733 in these processes. 118. Furthermore, our findings support the idea that invasion of Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics epithelial cells by Cor. diphtheriae represents an important 77, 71–94. step of infection and that it is advantageous for Cor. Broadway, M. M., Rogers, E. A., Chang, C., Huang, I. H., Dwivedi, P., diphtheriae to maintain an intracellular location. Yildirim, S., Schmitt, M. P., Das, A. & Ton-That, H. (2013). Pilus gene pool variation and the virulence of Corynebacterium diphtheriae At a biochemical level, we observed that DIP0733 deletion clinical isolates during infection of a nematode. J Bacteriol 195, 3774– effected collagen and fibronectin binding; however, no 3783. complete loss was observed but a decrease to about 50 % Burkovski, A. (2013a). Diphtheria. In The Prokaryotes, 4th ed., vol. 5, was observed. Many pathogens colonize human tissues or pp. 237–246. Edited by E. Rosenberg, E. F. DeLong, F. Thompson, evadeimmunemechanismsofbacterialclearanceexploiting S. Lory & E. Stackebrandt, New York: Springer. extracellular matrix and/or plasma proteins to colonize Burkovski, A. (2013b). Diphtheria and its etiological agents. In human tissues or to evade immune mechanisms for clearance Corynebacterium diphtheriae and Related Toxigenic Species, pp. 1–14. of bacteria (Simpson-Louredo et al.,2014).Sabbadiniet al. Edited by A. Burkovski. Dordrecht: Springer. (2010) suggested that the conversion of fibrinogen to fibrin Chagnot, C., Listrat, A., Astruc, T. & Desvaux, M. (2012). Bacterial may be connected to pseudomembrane formation, since adhesion to animal tissues: protein determinants for recognition of differences in the abilities to bind and convert fibrinogen extracellular matrix components. Cell Microbiol 14, 1687–1696. may partially explain differences in the extent of pseudo- Collier, R. J. (2001). Understanding the mode of action of diphtheria membrane formation during diphtheria. Obviously, DIP0733 toxin: a perspective on progress during the 20th century. Toxicon 39, 1793–1803. is involved in these processes, but is not exclusively responsible for binding. Colombo, A. V., Hirata, R., Jr, de Souza, C. M., Monteiro-Leal, L. H., Previato, J. O., Formiga, L. C., Andrade, A. F. & Mattos-Guaraldi, A. L. Interestingly, Cor. diphtheriae also exhibit strategies to (2001). Corynebacterium diphtheriae surface proteins as adhesins to survive within phagocytic cells independent of the tox gene, human erythrocytes. FEMS Microbiol Lett 197, 235–239. since also non-toxigenic strains can survive for a consid- de Bono, M. & Bargmann, C. I. (1998). Natural variation in a erable time within macrophages (dos Santos et al., 2010). As neuropeptide Y receptor homolog modifies social behavior and food shown by Sabbadini et al. (2012) and in this study, DIP0733 response in C. elegans. Cell 94, 679–689. is involved in processes to avoid and/or delay host defence dos Santos, C. S., dos Santos, L. S., de Souza, M. C., dos Santos Dourado, F., de Souza de Oliveira Dias, A. A., Sabbadini, P. S., mechanisms as well as to shield itself from extracellular Pereira, G. A., Cabral, M. C., Hirata Junior, R. & de Mattos-Guaraldi, antibiotics and efficiently persist, disseminate and infect A. L. (2010). Non-opsonic phagocytosis of homologous non-toxigenic deep tissues. Since phagocytosis by macrophages is independ- and toxigenic Corynebacterium diphtheriae strains by human U-937 ent of bacterial adhesion, it is likely that DIP0733 influences macrophages. Microbiol Immunol 54, 1–10. the survival of Cor. diphtheriae within macrophages directly. Garsin, D. A., Sifri, C. D., Mylonakis, E., Qin, X., Singh, K. V., Murray, A putative mechanism in this respect might be the induction B. E., Calderwood, S. B. & Ausubel, F. M. (2001). A simple model of apoptosis, which was indicated by in vitro experiments host for identifying Gram-positive virulence factors. Proc Natl Acad using isolated DIP0733 protein (Sabbadini et al., 2012). SciUSA98, 10892–10897. Gey, G. O., Coffmann, W. D. & Kubicek, M. T. (1952). Tissue culture In summary, DIP0733 seems to be a multi-functional pro- studies of the proliferative capacity of cervical carcinoma and normal tein with an important role in pathogenicity of Cor. epithelium. Cancer Res 12, 264–265. diphtheriae. Unfortunately, a detailed molecular analysis is Hadfield, T. L., McEvoy, P., Polotsky, Y., Tzinserling, V. A. & Yakovlev, complicated by a completely lacking annotation of func- A. A. (2000). The pathology of diphtheria. J Infect Dis 181 (Suppl. 1), tional domains within the protein (Sabbadini et al., 2012). S116–S120.

646 Microbiology 161

Hacker, E. 77 Publications

Characterization of C. diphtheriae DIP0733 protein

Hauck, C. R. & Ohlsen, K. (2006). Sticky connections: extracellular Readon-Robinson, M. E. & Ton-That, H. (2013). Assembly and matrix protein recognition and integrin-mediated cellular invasion by function of Corynebacterium diphtheriae pili. In Corynebacterium Staphylococcus aureus. Curr Opin Microbiol 9, 5–11. diphtheriae and Related Toxigenic Species, pp. 123–141. Edited by Hirata, R., Napolea˜ o, F., Monteiro-Leal, L. H., Andrade, A. F., Nagao, A. Burkovski. Dordrecht: Springer. P. E., Formiga, L. C., Fonseca, L. S. & Mattos-Guaraldi, A. L. (2002). Sabbadini, P. S., Genovez, M. R., Silva, C. F., Adelino, T. L., Santos, Intracellular viability of toxigenic Corynebacterium diphtheriae strains C. S., Pereira, G. A., Nagao, P. E., Dias, A. A., Mattos-Guaraldi, A. L., in HEp-2 cells. FEMS Microbiol Lett 215, 115–119. Hirata & Junior,´ R. (2010). Fibrinogen binds to nontoxigenic and Hirata, R., Jr, Souza, S. M., Rocha-de-Souza, C. M., Andrade, A. F., toxigenic Corynebacterium diphtheriae strains. Mem Inst Oswaldo Monteiro-Leal, L. H., Formiga, L. C. & Mattos-Guaraldi, A. L. (2004). Cruz 105, 706–711. Patterns of adherence to HEp-2 cells and actin polymerisation by Sabbadini, P. S., Assis, M. C., Trost, E., Gomes, D. L., Moreira, L. O., toxigenic Corynebacterium diphtheriae strains. Microb Pathog 36, 125– Dos Santos, C. S., Pereira, G. A., Nagao, P. E., Azevedo, V. A. & other 130. authors (2012). Corynebacterium diphtheriae 67-72p hemagglutinin, Jakoby, M., Ngouoto-Nkili, C.-E. & Burkovski, A. (1999). characterized as the protein DIP0733, contributes to invasion and Construction and application of new Corynebacterium glutamicum induction of apoptosis in HEp-2 cells. Microb Pathog 52, 165–176. vectors. Biotechnol Tech 13, 437–441. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, a Laboratory Manual. 2nd edn. Cold Spring Harbor, NY: Cold Spring P. A., McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A. & other Harbor Laboratory. authors (2007). CLUSTAL W and CLUSTAL_X version 2.0. Bioinformatics Scherer, W. F., Syverton, J. T. & Gey, G. O. (1953). Studies on the 23, 2947–2948. propagation in vitro of poliomyelitis viruses. IV. Viral multiplication Mattos-Guaraldi, A. L. (2002). Patterns of adherence to HEp-2 cells in a stable strain of human malignant epithelial cells (strain HeLa) and ability to induce actin polymerisation by toxigenic Coryne- derived from an epidermoid carcinoma of the cervix. J Exp Med 97, bacterium diphtheriae strains. In Seventh International Meeting of the 695–710. European Laboratory Working Group on Diphtheria, ELWGD. Abstract Schwarz-Linek, U., Ho¨ o¨ k, M. & Potts, J. R. (2006). Fibronectin- Book 1, pp. 72–73. Vienna. binding proteins of Gram-positive cocci. Microbes Infect 8, 2291– Murphy, J. R. (1996). Corynebacterium diphtheriae.InMedical Micro- 2298. biology, 4th edn, pp. 99–100. Edited by S. Baron. Galveston, TX: Simpson-Louredo, L., Ramos, J. N., Peixoto, R. S., Santos, L. S., University of Texas Medical Branch at Galveston. Antunes, C. A., Ladeira, E. M., Santos, C. S., Vieira, V. V., Boˆ as, M. H. Ott, L. & Burkovski, A. (2013). Toxigenic corynebacteria: adhesion, & other authors (2014). Corynebacterium ulcerans isolates from invasion and host response. In Corynebacterium diphtheriae and humans and dogs: fibrinogen, fibronectin and collagen-binding, Related Toxigenic Species, pp. 143–170. Edited by A. Burkovski. antimicrobial and PFGE profiles. Antonie van Leeuwenhoek 105, 343– Dordrecht: Springer. 352. Ott, L., Ho¨ ller, M., Gerlach, R. G., Hensel, M., Rheinlaender, J., Sorensen, K. & Brodbeck, U. (1986). Assessment of coating- Scha¨ ffer, T. E. & Burkovski, A. (2010a). Corynebacterium diphtheriae efficiency in ELISA plates by direct protein determination. invasion-associated protein (DIP1281) is involved in cell surface J Immunol Methods 95, 291–293. organization, adhesion and internalization in epithelial cells. BMC Trost, E., Blom, J., de Castro Soares, S., Huang, I. H., Al-Dilaimi, A., Microbiol 10,2. Schro¨ der, J., Jaenicke, S., Dorella, F. A., Rocha, F. S. & other authors Ott, L., Ho¨ ller, M., Rheinlaender, J., Scha¨ ffer, T. E., Hensel, M. & (2012). Pangenomic study of Corynebacterium diphtheriae that Burkovski, A. (2010b). Strain-specific differences in pili formation provides insights into the genomic diversity of pathogenic isolates and the interaction of Corynebacterium diphtheriae with host cells. from cases of classical diphtheria, endocarditis, and pneumonia. BMC Microbiol 10, 257. J Bacteriol 194, 3199–3215. Ott, L., McKenzie, A., Baltazar, M. T., Britting, S., Bischof, A., Tsuchiya, S., Yamabe, M., Yamaguchi, Y., Kobayashi, Y., Konno, T. & Burkovski, A. & Hoskisson, P. A. (2012). Evaluation of invertebrate Tada, K. (1980). Establishment and characterization of a human acute infection models for pathogenic corynebacteria. FEMS Immunol Med monocytic leukemia cell line (THP-1). Int J Cancer 26, 171–176. Microbiol 65, 413–421. Ventura, M., Canchaya, C., Tauch, A., Chandra, G., Fitzgerald, G. F., Ott, L., Scholz, B., Ho¨ ller, M., Hasselt, K., Ensser, A. & Burkovski, A. Chater, K. F. & van Sinderen, D. (2007). Genomics of Actinobacteria: k (2013). Induction of the NF -B signal transduction pathway in tracing the evolutionary history of an ancient phylum. Microbiol Mol response to Corynebacterium diphtheriae infection. Microbiology 159, Biol Rev 71, 495–548. 126–135. Zhi, X. Y., Li, W. J. & Stackebrandt, E. (2009). An update of the Puliti, M., von Hunolstein, C., Marangi, M., Bistoni, F. & Tissi, L. structure and 16S rRNA gene sequence-based definition of higher (2006). Experimental model of infection with non-toxigenic strains of ranks of the class Actinobacteria, with the proposal of two new Corynebacterium diphtheriae and development of septic arthritis. suborders and four new families and emended descriptions of the J Med Microbiol 55, 229–235. existing higher taxa. Int J Syst Evol Microbiol 59, 589–608. Rogers, E. A., Das, A. & Ton-That, H. (2011). Adhesion by pathogenic corynebacteria. Adv Exp Med Biol 715, 91–103. Edited by: T. Msadek

http://mic.sgmjournals.org 647

Hacker, E. 78 Publications

4.4 Caenorhabditis elegans star formation and negative chemotaxis induced by infection with corynebacteria (Antunes et al., 2015b)

Hacker, E. 79

Publications

Microbiology (2015), 00, 1–10 DOI 10.1099/mic.0.000201

Caenorhabditis elegans star formation and negative chemotaxis induced by infection with corynebacteria Camila Azevedo Antunes,1,2,3 Laura Clark,4 Marie-There`s Wanuske,1 Elena Hacker,1 Lisa Ott,1 Liliane Simpson-Louredo,3 Maria das Gracas de Luna,3 Raphael Hirata Jr,3 Ana Luı´za Mattos-Guaraldi,3 Jonathan Hodgkin4 and Andreas Burkovski1

Correspondence 1Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg, Erlangen, Germany Andreas Burkovski 2Universidade Federal de Minas Gerais, Instituto de Cieˆncias Biolo´gicas, Belo Horizonte, MG, Brazil [email protected] 3Faculty of Medical Sciences, Rio de Janeiro State University, UERJ, Rio de Janeiro, RJ, Brazil 4Department of Biochemistry, University of Oxford, Oxford, UK

Caenorhabditis elegans is one of the major model systems in biology based on advantageous properties such as short life span, transparency, genetic tractability and ease of culture using an Escherichia coli diet. In its natural habitat, compost and rotting plant material, this nematode lives on bacteria. However, C. elegans is not only a predator of bacteria, but can also be infected by nematopathogenic coryneform bacteria such Microbacterium and Leucobacter species, which display intriguing and diverse modes of pathogenicity. Depending on the nematode pathogen, aggregates of worms, termed worm-stars, can be formed, or severe rectal swelling, so-called Dar formation, can be induced. Using the human and animal pathogens Corynebacterium diphtheriae and Corynebacterium ulcerans as well as the non-pathogenic species Corynebacterium glutamicum, we show that these coryneform bacteria can also induce star formation slowly in WT worms, as well as a severe tail-swelling phenotype. While Received 4 August 2015 C. glutamicum had a significant, but minor influence on survival of C. elegans, nematodes were Revised 9 October 2015 killed after infection with C. diphtheriae and C. ulcerans. The two pathogenic species were Accepted 16 October 2015 avoided by the nematodes and induced aversive learning in C. elegans.

INTRODUCTION putative tellurite resistance protein (Santos et al., 2015) for virulence of Corynebacterium diphtheriae. Caenorhabditis elegans is a nematode with global distri- bution that lives by eating bacteria. Based on advantageous C. diphtheriae, the type species of the genus Corynebacterium, properties such as ease of culture, short life span, transpar- is the aetiological agent of diphtheria (Burkovski, 2013a, b), ency and genetic tractability, C. elegans has developed into a localized toxaemic infection of respiratory tract and skin one of the major model systems in biology. Introduced in that can be fatal (Burkovski, 2013b). Due to its medical the early 1970s as model for neural development (Brenner, importance, C. diphtheriae may be the best-investigated 1974), it was later developed as a model system for host– pathogenic member of the genus, which belongs to the class pathogen interactions, in respect to both innate immunity of Actinobacteria (high G+C Gram-positive bacteria) and and characterization of microbial virulence factors (Clark & comprises a collection of morphologically similar, irregular- Hodgkin, 2014). or club-shaped (micro)aerobic bacteria (Ventura et al., 2007; Zhi et al., 2009). Eighty-eight Corynebacterium species C. elegans was applied in a proof of principle study for have been described with about 53 having medical importance application of this nematode as an infection model for causing either human or zoonotic infections. The remaining pathogenic corynebacteria (Ott et al., 2012). Later, it was 35 species were isolated from different environments such as used to investigate the role of pili (Broadway et al., synthetic surfaces, foodstuff, water and soil (Bernard, 2012). 2013), the DIP0733 protein (Antunes et al., 2015) and a Interestingly, coryneform bacteria from different genera have been identified as natural pathogens of Abbreviation: Dar, deformed anal region. C. elegans. Microbacterium nematophilum was isolated

000201 G 2015 The Authors Printed in Great Britain 1

Hacker, E. 80 Publications

C. A. Antunes and others

from contaminated lab cultures of the nematode (Hodgkin In this study, we provide data on the species-specific et al., 2000). Infection of C. elegans with this coryneform colonization of C. elegans, localization of Corynebacterium worm pathogen results in a distinctive phenotype, a infection and chemotactic behaviour of the nematodes deformed anal region (Dar). C.elegans mutants with altered towards different corynebacteria. Three species were tested response to M. nematophilum infection were isolated from for this purpose: C. diphtheriae, an important pathogen and screens for a bacterially unswollen (Bus) phenotype. The the type strain of the genus; Corynebacterium ulcerans,a majority of these mutants exhibited little or no rectal infec- commensal corynebacterium isolated from a wide variety tion when challenged with the pathogen and are probably of domestic and wild animals, which shows zoonotic trans- altered in surface properties, which prevent colonization by mission to humans and is increasingly recognized as emerging the bacteria (Gravato-Nobre et al., 2005; Yook & Hodgkin, pathogen; and Corynebacterium glutamicum,anon- 2007). Manifestation of infection depends on colonization pathogenic member of the genus, applied as a biotechnology of the rectum of the nematode and involves tight adherence workhorse (Becker & Wittmann, 2012; Burkovski, 2015) and of the bacteria to rectal and anal regions. Infected animals isolated originally from a soil sample (Kinoshita et al., 1957). become mildly constipated and display reduced growth and fertility (Akimkina et al., 2006). Later, natural C. elegans populations, which showed Dar formation due METHODS to infection with Leucobacter strains, were identified. One bacterium isolated from an aviary in Japan caused symptoms Strains and growth conditions. Strains used in this study are listed similar to M. nematophilum infection, while two strains in Table 1. C. diphtheriae and C. ulcerans were grown in heart infusion (HI) broth at 37 uC, C. glutamicum was grown in brain heart infusion isolated from worms from rotting banana trunks in Cape broth at 30 uC and Escherichia coli OP50 was grown in Luria broth at 2 Verde exerted complementary virulence on C. elegans. In 37 uC (Sambrook et al., 1989). If appropriate, kanamycin 50 mgml 1 2 contrast to the former isolates, these strains, designated or chloramphenicol 25 mgml 1 was added. C. elegans N2, used as the Verde1 and Verde2, were highly virulent and lethal to the WT strain, and the respective mutant strains were maintained and nematodes (Hodgkin et al., 2013). A hallmark of Verde1 propagated on E. coli OP50 as described (Brenner, 1974). infection in liquid culture is the formation of aggregates of nematodes sticking together at their tails, termed worm- Infection of C. elegans. C. elegans N2 were maintained on agar stars. In contrast, Verde2 infection results in distorted plates inoculated with E. coli strain OP50 for 3–7 days until the worms internal organs and vacuole formation within the worms, become starved, indicated by clumping behaviour (de Bono & Bargmann, 1998). Subsequently, the nematodes were infected with as well as a Dar phenotype (Hodgkin et al., 2013). While different Corynebacterium strains transformed with pXMJ19mcherry, as these coryneform strains cause interesting disease pheno- well as E. coli OP50. Infection of 20 L4 stage larval worms was carried types in C. elegans, they are genetically intractable and little out with 20 ml of each bacteria strain (from an overnight culture) on is known about their biology. It is therefore convenient to NGM plates at 21 uC for 24 h. Worms were then transferred to plates study these phenotypes using a well-characterized bacterial with 20 ml of unlabelled E. coli OP50 for further 24 h, to allow the gut genus such as Corynebacterium. The inclusion of human- to clear of fluorescent organisms and cell debris. Worms were assessed each day following infection. Transfer back to E. coli OP50 is essential pathogenic coryneform strains in the C. elegans model also following infection, as corynebacteria are not a preferred prey source enhances the potential clinical relevance of these disease for C. elegans and continued feeding results in the nematodes phenotypes. attempting to leave the culture plates. Nematodes were paralysed with

Table 1. Strains, plasmids and C. elegans strains used in this study

Strain or plasmid Description Reference/source

Strains C. diphtheria + CDC-E8392 Biovar mitis; tox Hirata et al. (2008); Viguetti et al. (2012); Trost et al. (2012) C. glutamicum ATCC 13032 Type strain, non-pathogenic Abe et al. (1967) C. ulcerans 809 Bronchoalveolar lavage sample from an elderly Trost et al. (2011) woman with a fatal pulmonary infection BR-AD22 Nasal sample of an asymptomatic dog Trost et al. (2011) E. coli OP50 Uracil auxotrophic E. coli B strain Brenner (1974) Plasmids R pXMJ19mCherry ori colE1, oricg, ptac, mCherry,Cm Ott et al. (2012) C. elegans strains N2 Bristol WT strain Brenner (1974)

2 Microbiology 00

Hacker, E. 81 Publications

C. elegans infection by corynebacteria

(a) (b) (c)

(d) (e) (f)

(g) (h) (i)

(j) (k) (l)

Fig. 1. Induction of worm-star formation by corynebacteria. Aggregates of worms were observed after incubation for about 2–4 days in liquid medium in the presence of C. glutamicum ATCC13032 (a–c), C. diphtheriae CDC-E8392 (d–f), C. ulcerans 809 (g–i) and BR-AD22 (j–l).

5–10 ml of 0.6 % 2-phenoxy-2-propanol (Sigma; Wormbook.org) or each strain, approximately 60 nematodes were used and the assays were 20 mM sodium azide, mounted onto agar pads and photographed performed three times. Kaplan–Meier survival analysis was used and the using a Leica DMR fluorescence microscope. curves were compared with the logrank test. The statistical analysis was performed with Prism 5.0 (GraphPad), with P-valuesoflessthan0.05. To investigate worm-star formation, worms were exposed to bacterial cells in M9 buffer based on methods adapted from Hodgkin et al. (2013). Chemotactic behaviour. Chemotaxis and learning behaviour was Briefly, adult worms were washed three times in M9 buffer and aliquots of analysed as described previously (Zhang et al., 2005). The chemotactic 100 ml were placed into 500 mlofM9bufferineachwellofa24-well behaviour of the nematodes towards different corynebacteria were microplate. Aliquots of 0.05 ml, 0.1 ml and 0.2 ml of bacterial determined based on a choice index with 21.0 representing a com- strains grown in 20 ml HI medium overnight were placed in each well. plete preference for C. glutamicum or E. coli, the control bacterium Worm-star formation was initially observed 2 days post-incubation at used, an index of 1.0 representing complete preference for the test 21 uC. The interaction between nematodes and bacteria were inspected bacterium and an index of 0 representing an equal distribution. with light microscopy (Nikon C-DSD 230). Worm aggregates were picked out of liquid onto an agar plate and photographed as described above. Two bacterial suspensions (20 ml, OD60051.0) were spotted onto NGM plates for 24 h at 37 uC. Adult worms were washed twice in 2 Nematode killing assay. C. elegans N2 was maintained on E. coli OP50 small drops of M9 buffer containing 25 mgml 1 nalidixic acid to kill for 6–7 days until the worms became starved, as described above (de Bono adhering E. coli from the feeding plate. Twenty animals were placed & Bargmann, 1998). Infection of L4 stage larval worms was carried out near the centre of the plate at equal distance to the two spots of with 20 ml of each bacterial strain (from an overnight culture) on NGM bacteria. For training, a suspension of 200 ml of the test bacteria was plates at 21 uC for 24 h. Worms were assessed each day following infec- spread on a plate and 50 ml of OP50 suspension was used to place a tion, and the dead nematodes were counted and removed every 24 h. For small lawn on the side. Plates were incubated at 37 uC for 24 h before http://mic.microbiologyresearch.org 3

Hacker, E. 82 Publications

C. A. Antunes and others

of aggregates of worms sticking together at their tails. This (a) phenomenon of worm-star formation was described, until now, only for Leucobacter Verde1 infection of nematodes. Worms trapped in the stars usually cannot escape, and die within 24 h. The bacteria then use the worm carcasses as a nutrient source for growth. Interestingly, when C. elegans was brought in contact with the investigated corynebacteria, all species including the C. ulcerans strains isolated from a (b) human (809) and animal (BR-AD22) source were able to induce worm-star formation (Fig. 1). Aggregates of worms were observed after incubation for about 2–4 days in liquid culture in the presence of C. glutamicum ATCC13032 (Fig. 1a–c), C. diphtheriae CDC-E8392 (Fig. 1d–f), C. ulcerans 809 (Fig. 1g–i) and BR-AD22 (Fig. 1j–l). Compared to Leucobacter Verde1 infection, star formation was markedly (c) slower and tail-to-tail attachment due to corynebacterial colonization was weaker, since worms were able to escape the stars at least some of the time. The stars observed were formed by adult and larval state nematodes (Fig. 1b, f, i, k). No differences in respect to the period of time for or the extent of worm-star formation were observed. In all cases, nematodes started to break and lyse after prolonged incubation (d) and stars dissolved due to the destruction of the worms (Fig. 1c, l). Worm-stars were not observed during cultivation with E. coli OP50 (data not shown).

Dar induction by corynebacteria Compared to worm-star formation, the induction of tail (e) swelling (Dar phenotype), previously described as a result of M. nematophilum and Leucobacter Verde2 infections, is a more subtle symptom. In this case, adherence of bacteria to the rectal and post-anal cuticle of C. elegans induces local swelling of the underlying hypodermal tissue resulting in significant morphological changes (Hodgkin et al., 2000). Corynebacterial biofilms are already visible by light microscopy (Fig. 2b–d). However, the molecular background of colonization is unknown and needs further investigation, Fig. 2. Dar formation. Morphological changes induced by infec- e.g. scanning electron microscopy approaches, fluorescence tion of nematodes with C. glutamicum ATCC13032 (b), C. diphtheriae CDC-E8392 (c), C. ulcerans 809 (d) and BR- microscopy of biofilm formation and mutant analyses. AD22 (e). E. coli OP50 was used as control (a). The anal region As in the case of worm-star formation, all Corynebacterium of worms is indicated by white arrows. species tested induced Dars approximately after 2 days of infection of worms cultivated on NGM plates or in liquid medium (Fig. 2). However, significant species-specific differ- use. The nematodes were allowed to move freely on the plate and ences were observed. Dar formation in liquid medium was counted after 1, 2 and 24 h for the training assay and counted after most pronounced in the case of C. ulcerans BR-AD22 and 24 h for the index choice measurement. For each strain, approxi- 809 (Fig. 2d, e), followed by C. glutamicum ATCC 13032 mately 20 nematodes were used and the assays were performed in (Fig. 2b) and C. diphtheriae CDC-E8392 (Fig. 2c). The high- three independent experiments. Analysis by t-test was performed with Prism 5.0 (GraphPad), with P-values of less than 0.05. est rates were observed upon infection with C. ulcerans BR-AD22, with 5–43 % of worms showing Dars, followed by RESULTS C. ulcerans 809 with 10–40 %, C. glutamicum ATCC 13032 with 6–26 % and C. diphtheriae CDC-E8392 with 0–4 % of nematodes showing rectal swelling after 72 h exposure. The Worm-star formation by Corynebacterium fluctuating values for Dar formation in the four independent infection biological replicates carried out might indicate an influence An obvious and striking symptom of surface colonization of worm development on susceptibility and colonization. of C. elegans by certain pathogenic bacteria is the formation Furthermore, Dar formation was more abundant on solid

4 Microbiology 00

Hacker, E. 83 Publications

C. elegans infection by corynebacteria

123 4

a1 a2 a3 a4 ATCC-1302 C. glutamicum

b1 b2 b3 b4 CDC-E8392 C. diphtheriae

c1 c2 c3 c4 809 C. ulcerans

d1 d2 d3 d4 BR-AD22 C. ulcerans

e1 e2 e3 e4 ATCC 1302 ATCC C. glutamicum

f1 f2 f3 f4 CDC-E8392 C. diphtheriae

g1 g2 g3 g4 809 C. ulcerans

h1 h2 h3 h4 BR-AD22 C. ulcerans

http://mic.microbiologyresearch.org 5

Hacker, E. 84 Publications

C. A. Antunes and others

Fig. 3. Detailed analysis of colonization and morphological effects on N2. Worms were infected with the indicated mCherry- labelled bacteria for 48 h and colonization of different parts of the worms was analysed by fluorescence microscopy (a–d). Subsequently, the colonized worms were placed on plates with unlabelled E. coli OP50 for 24 h (e–h). All pictures were taken using identical fluorescence microscope settings. Bar, 50 mm.

media, indicating that the pathogenicity of strains differed bacterial strains, with C. glutamicum inducing milder depending on the culture conditions. Dar formation was not symptoms than C. diphtheriae, and the two C. ulcerans observed during cultivation with E. coli OP50 (Fig. 2a). strains inducing the most severe symptoms across all worm strains. Examination by microscopy at the 72 h time point revealed a large number of bacteria in the Colonization of C. elegans N2 guts of the worms, and what appear to be bacteria adhering to the post-anal cuticle (data not shown). Light and fluorescence microscopy experiments were carried out in order to investigate morphological effects Worms proliferated most successfully on C. glutamicum, induced by colonization of C. elegans N2 by corynebacteria and by 10 days post-infection there were more progeny (Fig. 3). Worms were infected with the indicated mCherry- and fewer Dauer worms on C. glutamicum plates than on labelled bacteria for 48 h and colonization of different parts the other bacterial strains, suggesting the worms were of the worms was analysed by fluorescence microscopy more successful in obtaining nutrition from C. glutamicum using identical settings for all samples. Strong colonization than from C. diphtheriae or C. ulcerans (data not shown). of all regions of the worm was observed for C. diphtheriae In a subsequent assay, the worms infected for 48 h were CDC-E8392 and C. ulcerans BR-AD22 (Fig. 3b, d), fol- placed on plates with unlabelled E. coli OP50 for further lowed by C. ulcerans 809 (Fig. 3c), while worms colonized 24 h to investigate persistence of colonization in absence with C. glutamicum ATCC 13032 showed the weakest flu- of external corynebacteria (Fig. 3f–h). In this experimental orescence signals (Fig. 3a). At 48 h post-infection, worms set-up, C. glutamicum was unable to persist within the on all Corynebacterium strains began to exhibit distension worms (Fig. 3e), while colonization with C. diphtheriae of the gut and, perhaps because of this, reduced movement CDC-E8392 was completely resistant to the possibility of around the plate. Many adult worms additionally exhibited new infection and uptake of OP50 (Fig. 3f). C. ulcerans a Dar phenotype. If larvae were present on the plate, they strains 809 and BR-AD22 also persisted even in the pre- frequently also exhibited a Dar phenotype, but with less sence of E. coli, although slightly less successfully than apparent distension of the gut and no obvious impairment C. diphtheriae (Fig. 3g, h). of motility. In summary, at the 48 h time point, differences were discernible between the effects of the different Survival of C. elegans in response to corynebacterial contact As shown above, corynebacteria are able to colonize 100 C. elegans and evoke morphological changes as well as altered movement behaviour. In order to study putative detrimental 80 effects of C. elegans colonization, survival of nematodes in relation to bacterial contact was determined. As expected, 60 E. coli OP50 had no detrimental effects on the worms. A sig- nificant, but minor influence of C. glutamicum ATCC13032

Surv i v al ( % ) 40 was observed, while C. diphtheriae CDC-E8392 as well as C. ulcerans 809 and BR-AD22 impaired survival of 20 C. elegans dramatically. After 5 days post-infection, about 70 % mortality was observed for worms infected with 0 C. diphtheriae, and about 90 % in the case of the two 0 1 2 3 4 5 C. ulcerans strains (Fig. 4). These results correlated with Time (days) the extent of colonization and persistence within the worms (Fig. 3).

Fig. 4. Nematode survival assay. Infection with E. coli OP50 (&), Adult C. elegans worms normally lay eggs that hatch outside C. glutamicum ATCC 13032 (%), C. diphtheriae CDC-E8392 (X), the parental body, but internal egg hatching, so-called C. ulcerans BR-AD22 (.) and 809 (m). Data shown are the mean ‘worm bagging’, was reported to be induced at a high fre- of three parallel experiments with 20 worms per plate repeated two quency by exposure to pathogenic bacteria, e.g. virulent times independently, error bars represent deviations from mean E. coli strains and Enterococcus faecalis, and can be regarded values. as a reliable population-wide stress reporter (Mosser et al.,

6 Microbiology 00

Hacker, E. 85 Publications

C. elegans infection by corynebacteria

(a) (b) 100 100

80 80

60 60 Nem a to d e ( % ) Nem a to d e ( % ) 40 40

20 20

0 0 E. coli OP50 E. coli OP50 E. coli OP50 E. coli OP50 C. ulcerans 809 C. ulcerans 809 C. ulcerans BR-AD22 C. ulcerans BR-AD22 C. diphtheriae CDC-E8392 C. diphtheriae CDC-E8392 ATCC 13032 C. glutamicum ATCC 13032 C. glutamicum ATCC 13032 C. glutamicum ATCC ATCC 13032 C. glutamicum ATCC

1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 –0.2 0.2– –0.4 –0.4 –0.6 –0.6

C h o ic e i n d ex 0.8– C h o ic e i n d ex –0.8 –1.0 –1.0 –1.2 –1.2 –1.4 1.4– 1.6– –1.6 C. ulcerans 809 C. ulcerans 809 C. ulcerans BR-AD22 C. ulcerans BR-AD22 C. diphtheriae CDC-E8392 C. diphtheriae CDC-E8392 ATCC 13032 C. glutamicum ATCC

Fig. 5. Chemotactic behaviour of C. elegans. Nematodes were transferred to plates with combinations of C. glutamicum and pathogenic corynebacteria (a) or E. coli and different Corynebacterium species (b) as indicated. A choice index was calculated as described in Methods, with a value of 21.0 representing a complete preference for the control bacterium (C. glutamicum or E. coli), an index of 1.0 representing complete preference for the test bacterium and an index of 0 representing an equal distribution. For each strain, approximately 20 nematodes were used and the assays were performed in three independent experiments. t-test analysis was performed with P-values of less than 0.05. Bars represent mean values¡SEM.

http://mic.microbiologyresearch.org 7

Hacker, E. 86 Publications

C. A. Antunes and others

1 h 2 h 24 h 20 20 20 18 18 18 16 16 16 14 14 14 12 12 12 10 10 10 8 8 8 6 6 6 4 4 4 2 2 2 0 0 0 Worms on e ach bac ter i um ( % ) Worms Worms on e ach bac ter i um ( % ) Worms Worms on e ach bac ter i um ( % ) Worms Untrained Starved Trained Untrained Starved Trained Untrained Starved Trained 1.0 1.0 1.0

0.5 0.5 0.5

0 0.0 0.0

–0.5 –0.5 0.5–

–1.0 1.0– –1.0 C h o ic e i n d ex a t 1 C h o ic e i n d ex a t 2 C h o ic e i n d ex a t 24 1.5– 1.5– –1.5

Fig. 6. Learning behaviour of C. elegans. Nematodes were taken from maintenance plates with E. coli OP50 (untrained), from pathogenic bacteria (C. diphtheriae CDC-E8392) for 4 h (trained) or from an empty NGM plate supplemented with 2 25 mgml 1 nalidixic acid for 4 h (starved). White bars indicate incubation with E. coli OP50 and grey bars indicate infection with C. diphtheriae CDC-E8392.A choice index for 1, 2 and 24 h was calculated. For statistical analysis, see Fig. 5.

2011). This phenomenon was suggested to be an adaptive Moreover, the worms showed learning behaviour. For this response, as the parental body could provide enough food set of experiments, worms with different cultivation and physical protection for the small larvae under this con- histories (OP50 grown, unstarved/starved, without and dition (Chen & Caswell-Chen, 2003). Therefore, in addition with previous contact to C. diphtheriae) were brought in to the killed worms, worm bagging was also scored as dead contact with C. diphtheriae CDC-E8392. Worms that were for the survival assay in this study. previously in contact with pathogenic corynebacteria, i.e. trained specimens, avoided these at early time points of contact in contrast to untrained individuals. However, after Chemotactic behaviour of worms 24 h, all worms avoided C. diphtheriae (Fig. 6). Taken together, the experiments suggest learning of nematodes to As indicated above, corynebacteria are able to infect avoid tainted, detrimental food sources. C. elegans and induce morphological changes similar to those previously described for other nematode pathogens. Consequently, it would be beneficial for C. elegans to avoid DISCUSSION these bacteria. As shown previously, C. elegans is able to avoid pathogenic Pseudomonas aeruginosa strains and C. elegans lives on bacteria but can also be infected by nema- shows aversive olfactory learning (Zhang et al., 2005). topathogenic coryneform bacteria such as Microbacterium In order to investigate if C. elegans is able to distinguish and Leucobacter species. The nematodes are a well-established different Corynebacterium species in order to avoid more infection model system for many pathogenic bacteria (Clark pathogenic strains, the preference of nematodes for & Hodgkin, 2014) including C. diphtheriae (Ott et al., 2012; C. glutamicum ATCC13032 and C. diphtheriae CDC-E8392, Broadway et al., 2013; Antunes et al., 2015; Santos et al., C. ulcerans 809 or BR-AD22 was tested. It was found that 2015). Interestingly, in this study, worm-star and Dar the nematodes avoided the more detrimental pathogenic formation, morphological changes typical for nematopatho- species and preferred the C. glutamicum ATCC13032 strain genic bacteria, were observed with both pathogenic and (Fig. 5a). A similar behaviour was found for the choice non-pathogenic corynebacteria. To our knowledge, this has between E. coli and corynebacteria. Strain OP50 was clearly never been observed before for other pathogenic bacteria preferred when tested versus C. diphtheriae and C. ulcerans like Serratia or Pseudomonas species. The grade of morpho- strains, while no preference was observed compared to logical changes did not correlate with survival, since C. glutamicum (Fig. 5b), which showed only a minor effect C. glutamicum had only a minor effect compared to strong on nematode survival, most likely due to its poor persistence negative influences of C. diphtheriae and C. ulcerans. Interest- in the worm (Fig. 3). ingly, C. elegans is able to distinguish between E. coli or

8 Microbiology 00

Hacker, E. 87 Publications

C. elegans infection by corynebacteria

C. glutamicum, and strongly detrimental bacteria such as REFERENCES C. diphtheriae and C. ulcerans. The feeding strain OP50 and C. glutamicum, which has only a minor effect on survival Abe, S., Takayama, K. & Kinoshita, S. (1967). Taxonomical studies on and is a poor gut colonizer, are preferred in comparison to glutamic acid producing bacteria. J Gen Microbiol 13, 279–301. C. diphtheriae and C. ulcerans, which showed strong negative Akimkina, T., Yook, K., Curnock, S. & Hodgkin, J. (2006). Genome influences on nematode survival. The worms showed learn- characterization, analysis of virulence and transformation of Microbacterium nematophilum, a coryneform pathogen of the ing behaviour as described previously for other bacterial nematode Caenorhabditis elegans. FEMS Microbiol Lett 264, 145–151. species including Pseudomonas fluorescens and Serratia mar- Antunes, C. A., Sanches dos Santos, L., Hacker, E., Ko¨ hler, S., cescens (Zhang et al., 2005). The molecular background of Bo¨ sl, K., Ott, L., de Luna, M., Hirata, R., Jr, Azevedo, V. A. & other aversive learning with respect to pathogenic Corynebacterium authors (2015). Characterization of DIP0733, a multi-functional species needs to be elucidated, and stimuli as well as signal virulence factor of Corynebacterium diphtheriae. Microbiology 161, pathways have still to be unravelled. 639–647. Becker, J. & Wittmann, C. (2012). Interactions between coryneform bacteria and soil nema- Bio-based production of chemicals, materials and fuels - Corynebacterium glutamicum as versatile cell todes are likely to be ecologically important and diverse, factory. Curr Opin Biotechnol 23, 631–640. in view of the abundance of these bacteria in soil. The Bernard, K. (2012). The genus Corynebacterium and other medically results obtained in this study might indicate that distinct, relevant coryneform-like bacteria. J Clin Microbiol 50, 3152–3158. but general properties of the genus members are respon- Brenner, S. (1974). The genetics of Caenorhabditis elegans. Genetics sible for the symptoms, rather than specific virulence fac- 77, 71–94. tors as discussed recently (Tauch & Burkovski, 2015). Broadway, M. M., Rogers, E. A., Chang, C., Huang, I. H., Dwivedi, P., Candidates in this respect might be components of the Yildirim, S., Schmitt, M. P., Das, A. & Ton-That, H. (2013). Pilus gene unique corynebacterial cell envelope (Burkovski, 2013). pool variation and the virulence of Corynebacterium diphtheriae Based on the established molecular biology toolboxes for clinical isolates during infection of a nematode. J Bacteriol 195, C. glutamicum, C. diphtheriae and C. ulcerans, mutants 3774–3783. with altered cell surfaces might be generated. Burkovski, A. (2013a). Diphtheria. In The Prokaryotes, 4th ed., vol. 5, In combination with already existing strains with altered pp. 237–246. Edited by E. Rosenberg, E. F. DeLong, F. Thompson, surface properties (Wolf et al., 2003; Tropis et al., 2005; S. Lory & E. Stackebrandt. New York: Springer. Ott et al., 2010), these would allow a detailed characteriz- Burkovski, A. (2013b). Cell envelope of corynebacteria: structure ation of different biosynthetic pathways, proteins and gly- and influence on pathogenicity. ISRN Microbiol 2013, 935736. colipids in respect to their function for C. elegans Burkovski, A. (2014). Diphtheria and its etiological agents. colonization. It might be worthwhile to include additional In Corynebacterium Diphtheriae and Related Toxigenic Species, Corynebacterium species in such studies. pp. 1–14. Edited by A. Burkovski. Dordrecht: Springer. Burkovski, A. (2015). Corynebacterium Glutamicum: From Systems Biology Future work might also focus on the analysis of C. elegans to Biotechnological Applications. Norfolk: Caister Academic Press. mutants to address the question of which host factors and Chen, J. & Caswell-Chen, E. P. (2003). Why Caenorhabditis elegans host signalling pathways are important for, and involved in, adults sacrifice their bodies to progeny. Nematology 5, 641–645. the infection process. The immune response of C. elegans to Clark, L. C. & Hodgkin, J. (2014). Commensals, probiotics and pathogens Corynebacterium infection might also be studied in more in the Caenorhabditis elegans model. Cell Microbiol 16, 27–38. detail; preliminary data obtained with C. elegans lys-7 de Bono, M. & Bargmann, C. I. (1998). Natural variation in a mutants hint, for example, to an important function of lyso- neuropeptide Y receptor homolog modifies social behavior and zyme in defence against corynebacteria, since corresponding food response in C. elegans. Cell 94, 679–689. worm mutants revealed increased mortality on the Gravato-Nobre, M. J., Nicholas, H. R., Nijland, R., O’Rourke, D., Corynebacterium strains tested (data not shown). Whittington, D. E., Yook, K. J. & Hodgkin, J. (2005). Multiple genes affect sensitivity of Caenorhabditis elegans to the bacterial pathogen In summary, the interaction of C. elegans with Coryne- Microbacterium nematophilum. Genetics 171, 1033–1045. bacterium species observed here might give new interest- Hodgkin, J., Kuwabara, P. E. & Corneliussen, B. (2000). A novel bacterial ing insights into ecologically relevant processes in the pathogen, Microbacterium nematophilum, induces morphological change nematodes’ natural habitat, as well as aspects of in the nematode C. elegans. Curr Biol 10, 1615–1618. C. elegans interactions with bacterial pathogens at a mol- Hodgkin, J., Fe´ lix, M. A., Clark, L. C., Stroud, D. & Gravato-Nobre, ecular level. M. J. (2013). Two Leucobacter strains exert complementary virulence on Caenorhabditis including death by worm-star formation. Curr Biol 23, 2157–2161. Kinoshita, S., Udaka, S. & Shimono, M. (1957). Studies on the ACKNOWLEDGEMENTS amino acid fermentation: I. Production of L-glutamic acid by various microorganism. J Gen Appl Microbiol 3, 193–205. C. A. A. was supported by a CAPES fellowship (Capes - Proc. BEX Mattos Guaraldi, A. L., Hirata, R., Jr & Azevedo, V. A. C. 2874/13-0) and by the Deutsche Forschungsgemeinschaft in-frame (2014). Corynebacterium diphtheriae, Corynebacterium ulcerans and of SFB796 (MGK). The worm sketch was kindly provided by Susanne Corynebacterium pseudotuberculosis-General Aspects. In Corynebacterium Morbach (Erlangen). The work of L. C. and J. H. was supported by Diphtheriae and Related Toxigenic Species, pp. 15–37. Edited by MRC grant MR/J001309/1. A. Burkovski. Dordrecht: Springer. http://mic.microbiologyresearch.org 9

Hacker, E. 88 Publications

C. A. Antunes and others

Mosser, T., Matic, I. & Leroy, M. (2011). Bacterium-induced internal Trost, E., Blom, J., Soares, S. C., Huang, I. H., Al-Dilaimi, A., egg hatching frequency is predictive of life span in Caenorhabditis Schro¨ der, J., Jaenicke, S., Dorella, F. A., Rocha, F. S. & other elegans populations. Appl Environ Microbiol 77, 8189–8192. authors (2012). Pangenomic study of Corynebacterium diphtheriae Ott, L., Ho¨ ller, M., Rheinlaender, J., Scha¨ ffer, T. E., Hensel, M. & that provides insights into the genomic diversity of pathogenic isolates Burkovski, A. (2010). Strain-specific differences in pili formation from cases of classical diphtheria, endocarditis, and pneumonia. and the interaction of Corynebacterium diphtheriae with host cells. J Bacteriol 194, 3199–3215. BMC Microbiol 10, 257. Ventura, M., Canchaya, C., Tauch, A., Chandra, G., Fitzgerald, G. F., Ott, L., McKenzie, A., Baltazar, M. T., Britting, S., Bischof, A., Chater, K. F. & van Sinderen, D. (2007). Genomics of Actinobacteria: Burkovski, A. & Hoskisson, P. A. (2012). Evaluation of invertebrate tracing the evolutionary history of an ancient phylum. Microbiol Mol infection models for pathogenic corynebacteria. FEMS Immunol Biol Rev 71, 495–548. Med Microbiol 65, 413–421. Viguetti, S. Z., Pacheco, L. G., Santos, L. S., Soares, S. C., Bolt, F., Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: Baldwin, A., Dowson, C. G., Rosso, M. L., Guiso, N. & other authors a Laboratory Manual, 2nd edn., Cold Sring Harbor, NY: Cold Spring (2012). Multilocus sequence types of invasive Corynebacterium Harbor Laboratory. diphtheriae isolated in the Rio de Janeiro urban area, Brazil. Santos, L. S., Antunes, C. A., Santos, C. S., Pereira, J. A. A., Epidemiol Infect 140, 617–620. Sabbadini, P. S., Luna, M., Azevedo, V., Hirata Ju´ nior, R., Wolf, A., Kra¨ mer, R. & Morbach, S. (2003). Three pathways for Burkovski, A. & other authors (2015). Corynebacterium diphtheriae trehalose metabolism in Corynebacterium glutamicum ATCC13032 putative tellurite-resistance protein (CDCE8392_0813) contributes and their significance in response to osmotic stress. Mol Microbiol to the intracellular survival in human epithelial cells and lethality of 49, 1119–1134. Caenorhabditis elegans. Mem Inst Oswaldo Cruz 110, 662–668. Yook, K. & Hodgkin, J. (2007). Mos1 mutagenesis reveals a diversity of Tauch, A. & Burkovski, A. (2015). Molecular armory or niche factors: mechanisms affecting response of Caenorhabditis elegans to the bacterial virulence determinants of Corynebacterium species. FEMS Microbiol pathogen Microbacterium nematophilum. Genetics 175, 681–697. Lett, fnv185. Zhang, Y., Lu, H. & Bargmann, C. I. (2005). Pathogenic bacteria Tropis, M., Meniche, X., Wolf, A., Gebhardt, H., Strelkov, S., Chami, M., induce aversive olfactory learning in Caenorhabditis elegans. Nature Schomburg, D., Kra¨mer, R., Morbach, S. & Daffe´ ,M.(2005).The 438, 179–184. crucial role of trehalose and structurally related oligosaccharides in Zhi,X.Y.,Li,W.J.&Stackebrandt,E.(2009).An update of the the biosynthesis and transfer of mycolic acids in Corynebacterineae. structure and 16S rRNA gene sequence-based definition of JBiolChem280, 26573–26585. higher ranks of the class Actinobacteria, with the proposal of Trost, E., Al-Dilaimi, A., Papavasiliou, P., Schneider, J., Viehoever, P., two new suborders and four new families and emended Burkovski, A., Soares, S. C., Almeida,S.S.,Dorella,F.A.&otherauthors descriptions of the existing higher taxa. Int J Syst Evol Microbiol (2011). Comparative analysis of two complete Corynebacterium ulcerans 59, 589–608. genomes and detection of candidate virulence factors. BMC Genomics 12, 383. Edited by: D. Demuth

Hacker, E. 89 References

5 References

Antunes, C. A., Sanches Dos Santos, L., Hacker, E., Köhler, S., Bösl, K., Ott, L., de Luna, M., Hirata, R., Jr., Azevedo, V. A., Mattos-Guaraldi, A. L. & Burkovski, A. (2015a). Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae. Microbiology 161, 639-647.

Antunes, C. A., Clark, L., Wanuske, M. T., Hacker, E., Ott, L., Simpson-Louredo, L., das Gracas de Luna, M., Hirata, R., Jr., Mattos-Guaraldi, A. L., Hodgkin, J. & Burkovski, A. (2015b). Caenorhabditis elegans star formation and negative chemotaxis induced by infection with corynebacteria. Microbiology, 10.1099/mic.0.000201.

Arden, S. B., Chang, W. H. & Barksdale, L. (1972). Distribution of neuraminidase and N- acetylneuraminate lyase activities among corynebacteria, mycobacteria, and nocardias. J Bacteriol 112, 1206-1212.

Barksdale, L. (1970). Corynebacterium diphtheriae and its relatives. Bacteriol Rev 34, 378- 422.

Barksdale, L., Linder, R., Sulea, I. T. & Pollice, M. (1981). Phospholipase D activity of Corynebacterium pseudotuberculosis (Corynebacterium ovis) and Corynebacterium ulcerans, a distinctive marker within the genus Corynebacterium. J Clin Microbiol 13, 335-343.

Batey, R. G. (1986). Pathogenesis of caseous lymphadenitis in sheep and goats. Aust Vet J 63, 269-272.

Benevides Lde, J., Viana, M. V., Mariano, D. C., Rocha Fde, S., Bagano, P. C., Folador, E. L., Pereira, F. L., Dorella, F. A., Leal, C. A., Carvalho, A. F., Soares Sde, C., Carneiro, A., Ramos, R., Badell-Ocando, E., Guiso, N., Silva, A., Figueiredo, H., Azevedo, V. & Guimaraes, L. C. (2015). Genome sequence of Corynebacterium ulcerans strain FRC11. Genome Announc 3.

Berger, A., Huber, I., Merbecks, S. S., Ehrhard, I., Konrad, R., Hörmansdorfer, S., Hogardt, M. & Sing, A. (2011). Toxigenic Corynebacterium ulcerans in woman and cat. Emerg Infect Dis 17, 1767-1769.

Berger, A., Boschert, V., Konrad, R., Schmidt-Wieland, T., Hörmansdorfer, S., Eddicks, M. & Sing, A. (2013). Two cases of cutaneous diphtheria associated with occupational pig contact in Germany. Zoonoses Public Health 60, 539-542.

Bergin, I. L., Chien, C. C., Marini, R. P. & Fox, J. G. (2000). Isolation and characterization of Corynebacterium ulcerans from cephalic implants in macaques. Comp Med 50, 530-535.

Bernard, K. A. & Funke, G. (2012). Genus I. Corynebacterium Lehmann and Neumann 1896, 350AL emend. Bernard, Wiebe, Burdz, Reimer, Ng, Singh, Schindle and Pacheco 2010, 877. In Bergey's Manual of Systematic Bacteriology, pp. 245-289. Edited by W. B. Whitman, M. Goodfellow, P. Kämpfer, H. J. Busse, M. E. Trujillo, W. Ludwig, K. I. Suzuki & A. Parte. New York: Springer New York.

Bonmarin, I., Guiso, N., Le Fleche-Mateos, A., Patey, O., Patrick, A. D. & Levy-Bruhl, D. (2009). Diphtheria: A zoonotic disease in France? Vaccine 27, 4196-4200.

Hacker, E. 90

References

Bostock, A. D., Gilbert, F. R., Lewis, D. & Smith, D. C. (1984). Corynebacterium ulcerans infection associated with untreated milk. J Infect 9, 286-288.

Broadway, M. M., Rogers, E. A., Chang, C., Huang, I. H., Dwivedi, P., Yildirim, S., Schmitt, M. P., Das, A. & Ton-That, H. (2013). Pilus gene pool variation and the virulence of Corynebacterium diphtheriae clinical isolates during infection of a nematode. J Bacteriol 195, 3774-3783.

Burkovski, A. (2013). Cell envelope of corynebacteria: structure and influence on pathogenicity. ISRN Microbiol 2013, 935736.

Burkovski, A. (2014). Diphtheria and its etiological agents. In Corynebacterium diphtheriae and Related Toxigenic Species, pp. 1-14. Edited by A. Burkovski. Dordrecht: Springer Netherlands.

Busse, H. J. (2012). Family I. Corynebacteriaceae Lehmann and Neumann 1907AL emend. Stackebrandt, Rainey and Ward-Rainey 1997, 485 emend. Zhi, Li and Stackebrandt 2009, 593. In Bergey‘s Manual of Systematic Bacteriology, pp. 244−245. Edited by W. B. Whitman, M. Goodfellow, P. Kämpfer, H. J. Busse, M. E. Trujillo, W. Ludwig, K. I. Suzuki & A. Parte. New York: Springer New York.

Casadevall, A. & Pirofski, L. A. (2014). Microbiology: Ditch the term pathogen. Nature 516, 165-166.

Chang, C., Mandlik, A., Das, A. & Ton-That, H. (2011). Cell surface display of minor pilin adhesins in the form of a simple heterodimeric assembly in Corynebacterium diphtheriae. Mol Microbiol 79, 1236-1247.

Cianciotto, N., Rappuoli, R. & Groman, N. (1986). Detection of homology to the beta bacteriophage integration site in a wide variety of Corynebacterium spp. J Bacteriol 168, 103- 108.

Collier, R. J. (2001). Understanding the mode of action of diphtheria toxin: a perspective on progress during the 20th century. Toxicon 39, 1793-1803.

Collin, M. & Fischetti, V. A. (2004). A novel secreted endoglycosidase from Enterococcus faecalis with activity on human immunoglobulin G and ribonuclease B. J Biol Chem 279, 22558-22570.

Collins, M. D., Jones, D. & Schofield, G. M. (1982). Reclassification of 'Corynebacterium haemolyticum' (MacLean, Liebow & Rosenberg) in the genus Arcanobacterium gen.nov. as Arcanobacterium haemolyticum nom.rev., comb.nov. J Gen Microbiol 128, 1279-1281.

Colombo, A. V., Hirata, R., Jr., de Souza, C. M., Monteiro-Leal, L. H., Previato, J. O., Formiga, L. C., Andrade, A. F. & Mattos-Guaraldi, A. L. (2001). Corynebacterium diphtheriae surface proteins as adhesins to human erythrocytes. FEMS Microbiol Lett 197, 235-239.

Contzen, M., Sting, R., Blazey, B. & Rau, J. (2011). Corynebacterium ulcerans from diseased wild boars. Zoonoses Public Health 58, 479-488.

Corboz, L., Thoma, R., Braun, U. & Zbinden, R. (1996). Isolation of Corynebacterium diphtheriae subsp. belfanti from a cow with chronic active dermatitis. Schweiz Arch Tierheilkd 138, 596-599.

Hacker, E. 91

References

Corfield, T. (1992). Bacterial sialidases - roles in pathogenicity and nutrition. Glycobiology 2, 509-521.

Corti, M. A., Bloemberg, G. V., Borelli, S., Kutzner, H., Eich, G., Hoelzle, L. & Lautenschlager, S. (2012). Rare human skin infection with Corynebacterium ulcerans: transmission by a domestic cat. Infection 40, 575-578.

Cuevas, W. A. & Songer, J. G. (1993). Arcanobacterium haemolyticum phospholipase D is genetically and functionally similar to Corynebacterium pseudotuberculosis phospholipase D. Infect Immun 61, 4310-4316.

Dias, A. A., Silva, F. C., Jr., Pereira, G. A., Souza, M. C., Camello, T. C., Damasceno, J. A., Pacheco, L. G., Miyoshi, A., Azevedo, V. A., Hirata, R., Jr., Boas, M. H. & Mattos- Guaraldi, A. L. (2010). Corynebacterium ulcerans isolated from an asymptomatic dog kept in an animal shelter in the metropolitan area of Rio de Janeiro, Brazil. Vector Borne Zoonotic Dis 10, 743-748.

Dias, A. A., Silva, F. C., Jr., Santos, L. S., Ribeiro-Carvalho, M. M., Sabbadini, P. S., Santos, C. S., Filardy, A. A., Myioshi, A., Azevedo, V. A., Hirata, R., Jr., Villas-Boas, M. H. & Mattos-Guaraldi, A. L. (2011). Strain-dependent arthritogenic potential of the zoonotic pathogen Corynebacterium ulcerans. Vet Microbiol 153, 323-331.

Domingo, M. C., Fournier, E., Masse, C., Charest, H., Bernard, K., Cote, J. C. & Tremblay, C. (2015). Draft genome sequences of two toxigenic Corynebacterium ulcerans strains. Genome Announc 3.

Dorella, F. A., Pacheco, L. G., Oliveira, S. C., Miyoshi, A. & Azevedo, V. (2006). Corynebacterium pseudotuberculosis: microbiology, biochemical properties, pathogenesis and molecular studies of virulence. Vet Res 37, 201-218.

Dubin, G. (2002). Extracellular proteases of Staphylococcus spp. Biol Chem 383, 1075-1086.

Elden, S., Coole, L., Efstratiou, A. & Doshi, N. (2007). Laboratory-confirmed case of toxigenic Corynebacterium ulcerans. Euro Surveill 12, E070329 070323.

Endo, Y., Tsurugi, K., Yutsudo, T., Takeda, Y., Ogasawara, T. & Igarashi, K. (1988). Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes. RNA N-glycosidase activity of the toxins. Eur J Biochem 171, 45-50.

Foster, G., Patterson, T., Howie, F., Simpson, V., Davison, N., Efstratiou, A. & Lai, S. (2002). Corynebacterium ulcerans in free-ranging otters. Vet Rec 150, 524.

Funke, G., von Graevenitz, A., Clarridge, J. E., 3rd & Bernard, K. A. (1997). Clinical microbiology of coryneform bacteria. Clin Microbiol Rev 10, 125-159.

Fuursted, K., Soes, L. M., Crewe, B. T., Stegger, M., Andersen, P. S. & Christensen, J. J. (2015). Non-toxigenic tox gene-bearing Corynebacterium ulcerans in a traumatic ulcer from a human case and his asymptomatic dog. Microbes Infect 17, 717-719.

Gaspar, A. H. & Ton-That, H. (2006). Assembly of distinct pilus structures on the surface of Corynebacterium diphtheriae. J Bacteriol 188, 1526-1533.

Goodfellow, M. (2012). Class I. Actinobacteria Stackebrandt, Rainey and Ward-Rainey 1997, 483. In Bergey‘s Manual of Systematic Bacteriology, pp. 34-35. Edited by W. B. Whitman, M.

Hacker, E. 92

References

Goodfellow, P. Kämpfer, H. J. Busse, M. E. Trujillo, W. Ludwig, K. I. Suzuki & A. Parte. New York: Springer New York.

Goodfellow, M. & Jones, A. L. (2012). Order Corynebacteriales. In Bergey's Manual of Systematic Bacteriology, pp. 235-243. Edited by W. B. Whitman, M. Goodfellow, P. Kämpfer, H. J. Busse, M. E. Trujillo, W. Ludwig, K. I. Suzuki & A. Parte. New York: Springer New York.

Hacker, E., Ott, L., Hasselt, K., Mattos-Guaraldi, A. L., Tauch, A. & Burkovski, A. (2015a). Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources. Microbiology 161, 1582-1591.

Hacker, E., Ott, L., Schulze-Luehrmann, J., Lührmann, A., Wiesmann, V., Wittenberg, T. & Burkovski, A. (2015b). The killing of macrophages by Corynebacterium ulcerans. Virulence, 10.1080/21505594.2015.1125068.

Hall, A. J., Cassiday, P. K., Bernard, K. A., Bolt, F., Steigerwalt, A. G., Bixler, D., Pawloski, L. C., Whitney, A. M., Iwaki, M., Baldwin, A., Dowson, C. G., Komiya, T., Takahashi, M., Hinrikson, H. P. & Tondella, M. L. (2010). Novel Corynebacterium diphtheriae in domestic cats. Emerg Infect Dis 16, 688-691.

Hart, R. J. (1984). Corynebacterium ulcerans in humans and cattle in North Devon. J Hyg (Lond) 92, 161-164.

Hatanaka, A., Tsunoda, A., Okamoto, M., Ooe, K., Nakamura, A., Miyakoshi, M., Komiya, T. & Takahashi, M. (2003). Corynebacterium ulcerans diphtheria in Japan. Emerg Infect Dis 9, 752-753.

Hett, E. C., Chao, M. C., Deng, L. L. & Rubin, E. J. (2008). A mycobacterial enzyme essential for cell division synergizes with resuscitation-promoting factor. PLoS Pathog 4, e1000001.

Hill, C. (2012). Virulence or niche factors: what's in a name? J Bacteriol 194, 5725-5727.

Hirai-Yuki, A., Komiya, T., Suzaki, Y., Ami, Y., Katsukawa, C., Takahashi, M., Yamamoto, A. & Yamada, Y. K. (2013). Isolation and characterization of toxigenic Corynebacterium ulcerans from 2 closed colonies of cynomolgus macaques (Macaca fascicularis) in Japan. Comp Med 63, 272-278.

Hirata, R., Jr., Souza, S. M., Rocha-de-Souza, C. M., Andrade, A. F., Monteiro-Leal, L. H., Formiga, L. C. & Mattos-Guaraldi, A. L. (2004). Patterns of adherence to HEp-2 cells and actin polymerisation by toxigenic Corynebacterium diphtheriae strains. Microb Pathog 36, 125- 130.

Hodgkin, J., Kuwabara, P. E. & Corneliussen, B. (2000). A novel bacterial pathogen, Microbacterium nematophilum, induces morphological change in the nematode C. elegans. Curr Biol 10, 1615-1618.

Hodgkin, J., Felix, M. A., Clark, L. C., Stroud, D. & Gravato-Nobre, M. J. (2013). Two Leucobacter strains exert complementary virulence on Caenorhabditis including death by worm-star formation. Curr Biol 23, 2157-2161.

Hodgson, A. L., Krywult, J., Corner, L. A., Rothel, J. S. & Radford, A. J. (1992). Rational attenuation of Corynebacterium pseudotuberculosis: potential cheesy gland vaccine and live delivery vehicle. Infect Immun 60, 2900-2905.

Höfler, W. (1991). Cutaneous diphtheria. Int J Dermatol 30, 845-847.

Hacker, E. 93

References

Hogg, R. A., Wessels, J., Hart, J., Efstratiou, A., De Zoysa, A., Mann, G., Allen, T. & Pritchard, G. C. (2009). Possible zoonotic transmission of toxigenic Corynebacterium ulcerans from companion animals in a human case of fatal diphtheria. Vet Rec 165, 691-692.

Holmes, R. K. (2000). Biology and molecular epidemiology of diphtheria toxin and the tox gene. J Infect Dis 181 (Suppl 1), 156-167.

Hommez, J., Devriese, L. A., Vaneechoutte, M., Riegel, P., Butaye, P. & Haesebrouck, F. (1999). Identification of nonlipophilic corynebacteria isolated from dairy cows with mastitis. J Clin Microbiol 37, 954-957.

Indrigo, J., Hunter, R. L., Jr. & Actor, J. K. (2003). Cord factor trehalose 6,6'-dimycolate (TDM) mediates trafficking events during mycobacterial infection of murine macrophages. Microbiology 149, 2049-2059.

Katsukawa, C., Kawahara, R., Inoue, K., Ishii, A., Yamagishi, H., Kida, K., Nishino, S., Nagahama, S., Komiya, T., Iwaki, M. & Takahashi, M. (2009). Toxigenic Corynebacterium ulcerans isolated from the domestic dog for the first time in Japan. Jpn J Infect Dis 62, 171- 172.

Katsukawa, C., Komiya, T., Yamagishi, H., Ishii, A., Nishino, S., Nagahama, S., Iwaki, M., Yamamoto, A. & Takahashi, M. (2012). Prevalence of Corynebacterium ulcerans in dogs in Osaka, Japan. J Med Microbiol 61, 266-273.

Kim, S., Oh, D. B., Kwon, O. & Kang, H. A. (2010). Identification and functional characterization of the NanH extracellular sialidase from Corynebacterium diphtheriae. J Biochem 147, 523-533.

Kolodkina, V., Denisevich, T. & Titov, L. (2011). Identification of Corynebacterium diphtheriae gene involved in adherence to epithelial cells. Infect Genet Evol 11, 518-521.

König, C., Meinel, D. M., Margos, G., Konrad, R. & Sing, A. (2014). Multilocus sequence typing of Corynebacterium ulcerans provides evidence for zoonotic transmission and for increased prevalence of certain sequence types among toxigenic srains. J Clin Microbiol 52, 4318-4324.

Konrad, R., Hörmansdorfer, S. & Sing, A. (2015). Possible human-to-human transmission of toxigenic Corynebacterium ulcerans. Clin Microbiol Infect 21, 768-771.

Lartigue, M. F., Monnet, X., Le Fleche, A., Grimont, P. A., Benet, J. J., Durrbach, A., Fabre, M. & Nordmann, P. (2005). Corynebacterium ulcerans in an immunocompromised patient with diphtheria and her dog. J Clin Microbiol 43, 999-1001.

Leggett, B. A., De Zoysa, A., Abbott, Y. E., Leonard, N., Markey, B. & Efstratiou, A. (2010). Toxigenic Corynebacterium diphtheriae isolated from a wound in a horse. Vet Rec 166, 656- 657.

Lehmann, K. B. & Neumann, R. (1896). Atlas und Grundriss der Bakteriologie und Lehrbuch der speziellen bakteriologischen Diagnostik (Erstausgabe): J. F. Lehmann, München.

Lingwood, C. A. (1993). Verotoxins and their glycolipid receptors. Adv Lipid Res 25, 189-211.

Louie, G. V., Yang, W., Bowman, M. E. & Choe, S. (1997). Crystal structure of the complex of diphtheria toxin with an extracellular fragment of its receptor. Mol Cell 1, 67-78.

Hacker, E. 94

References

Lucas, E. A., Billington, S. J., Carlson, P., McGee, D. J. & Jost, B. H. (2010). Phospholipase D promotes Arcanobacterium haemolyticum adhesion via lipid raft remodeling and host cell death following bacterial invasion. BMC Microbiol 10, 270.

Madigan, M. T., Martinko, J. M., Dunlap, P. V. & Clark, D. P. (2009). Brock Biology of Microorganisms, 12 edn. San Fransisco: Pearson.

Mandlik, A., Swierczynski, A., Das, A. & Ton-That, H. (2007). Corynebacterium diphtheriae employs specific minor pilins to target human pharyngeal epithelial cells. Mol Microbiol 64, 111-124.

Marsh, E. K. & May, R. C. (2012). Caenorhabditis elegans, a model organism for investigating immunity. Appl Environ Microbiol 78, 2075-2081.

Mattos-Guaraldi, A. L., Formiga, L. C. & Andrade, A. F. (1998). Trans-sialidase activity for sialic acid incorporation on Corynebacterium diphtheriae. FEMS Microbiol Lett 168, 167-172.

Mattos-Guaraldi, A. L., Sampaio, J. L., Santos, C. S., Pimenta, F. P., Pereira, G. A., Pacheco, L. G., Miyoshi, A., Azevedo, V., Moreira, L. O., Gutierrez, F. L., Costa, J. L., Costa-Filho, R., Damasco, P. V., Camello, T. C. & Hirata, R., Jr. (2008). First detection of Corynebacterium ulcerans producing a diphtheria-like toxin in a case of human with pulmonary infection in the Rio de Janeiro metropolitan area, Brazil. Mem Inst Oswaldo Cruz 103, 396- 400.

McKean, S. C., Davies, J. K. & Moore, R. J. (2007). Expression of phospholipase D, the major virulence factor of Corynebacterium pseudotuberculosis, is regulated by multiple environmental factors and plays a role in macrophage death. Microbiology 153, 2203-2211.

McNamara, P. J., Bradley, G. A. & Songer, J. G. (1994). Targeted mutagenesis of the phospholipase D gene results in decreased virulence of Corynebacterium pseudotuberculosis. Mol Microbiol 12, 921-930.

McNamara, P. J., Cuevas, W. A. & Songer, J. G. (1995). Toxic phospholipases D of Corynebacterium pseudotuberculosis, C. ulcerans and Arcanobacterium haemolyticum: cloning and sequence homology. Gene 156, 113-118.

Meinel, D. M., Margos, G., Konrad, R., Krebs, S., Blum, H. & Sing, A. (2014). Next generation sequencing analysis of nine Corynebacterium ulcerans isolates reveals zoonotic transmission and a novel putative diphtheria toxin-encoding pathogenicity island. Genome Med 6, 113.

Meinel, D. M., Konrad, R., Berger, A., König, C., Schmidt-Wieland, T., Hogardt, M., Bischoff, H., Ackermann, N., Hörmansdorfer, S., Krebs, S., Blum, H., Margos, G. & Sing, A. (2015). Zoonotic transmission of toxigenic Corynebacterium ulcerans strain, Germany, 2012. Emerg Infect Dis 21, 356-358.

Mitamura, T., Higashiyama, S., Taniguchi, N., Klagsbrun, M. & Mekada, E. (1995). Diphtheria toxin binds to the epidermal growth factor (EGF)-like domain of human heparin- binding EGF-like growth factor/diphtheria toxin receptor and inhibits specifically its mitogenic activity. J Biol Chem 270, 1015-1019.

Mizutani, S., Nakazono, N. & Sugino, Y. (1999). The so-called chromosomal verotoxin genes are actually carried by defective prophages. DNA Res 6, 141-143.

Hacker, E. 95

References

Mochizuki, Y., Saeki, H., Iwaki, M., Takagi, H., Shibayama, K., Amao, H. & Yamamoto, A. (2015). A novel experimental platform for toxigenic and nontoxigenic Corynebacterium ulcerans infection in mice. Pathog Dis, 10.1093/femspd/ftv109.

Morris, W. E., Uzal, F. A. & Cipolla, A. L. (2005). Pyogranulomatous meningoencephalitis in a goat due to Corynebacterium ulcerans. Vet Rec 156, 317-318.

Naglich, J. G., Metherall, J. E., Russell, D. W. & Eidels, L. (1992). Expression cloning of a diphtheria toxin receptor: identity with a heparin-binding EGF-like growth factor precursor. Cell 69, 1051-1061.

Nappi, A. J. & Christensen, B. M. (2005). Melanogenesis and associated cytotoxic reactions: applications to insect innate immunity. Insect Biochem Mol Biol 35, 443-459.

Olson, M. E., Goemans, I., Bolingbroke, D. & Lundberg, S. (1988). Gangrenous dermatitis caused by Corynebacterium ulcerans in Richardson ground squirrels. J Am Vet Med Assoc 193, 367-368.

Ott, L., Höller, M., Gerlach, R. G., Hensel, M., Rheinlaender, J., Schaffer, T. E. & Burkovski, A. (2010). Corynebacterium diphtheriae invasion-associated protein (DIP1281) is involved in cell surface organization, adhesion and internalization in epithelial cells. BMC Microbiol 10, 2.

Ott, L., McKenzie, A., Baltazar, M. T., Britting, S., Bischof, A., Burkovski, A. & Hoskisson, P. A. (2012). Evaluation of invertebrate infection models for pathogenic corynebacteria. FEMS Immunol Med Microbiol 65, 413-421.

Pappenheimer, A. M., Jr. (1977). Diphtheria toxin. Annu Rev Biochem 46, 69-94.

Putong, N., Agustin, G., Pasubillo, M., Miyagi, K. & Dimaano, E. M. (2011). Diphtheria-like illness due to Corynebacterium ulcerans infection. Trop Med Health 39, 1-2.

Riegel, P., Ruimy, R., de Briel, D., Prévost, G., Jehl, F., Christen, R. & Monteil, H. (1995). Taxonomy of Corynebacterium diphtheriae and related taxa, with recognition of Corynebacterium ulcerans sp. nov. nom. rev. FEMS Microbiol Lett 126, 271-276.

Roggentin, P., Schauer, R., Hoyer, L. L. & Vimr, E. R. (1993). The sialidase superfamily and its spread by horizontal gene transfer. Mol Microbiol 9, 915-921.

Rückert, C., Eimer, J., Winkler, A. & Tauch, A. (2015). Complete genome sequence of the type strain Corynebacterium epidermidicanis DSM 45586, isolated from the skin of a dog suffering from pruritus. Genome Announc 3.

Sabbadini, P. S., Genovez, M. R., Silva, C. F., Adelino, T. L., Santos, C. S., Pereira, G. A., Nagao, P. E., Dias, A. A., Mattos-Guaraldi, A. L. & Hirata R., J. (2010). Fibrinogen binds to nontoxigenic and toxigenic Corynebacterium diphtheriae strains. Mem Inst Oswaldo Cruz 105, 706-711.

Sabbadini, P. S., Assis, M. C., Trost, E., Gomes, D. L., Moreira, L. O., Dos Santos, C. S., Pereira, G. A., Nagao, P. E., Azevedo, V. A., Hirata R., J., Dos Santos, A. L., Tauch, A. & Mattos-Guaraldi, A. L. (2012). Corynebacterium diphtheriae 67-72p hemagglutinin, characterized as the protein DIP0733, contributes to invasion and induction of apoptosis in HEp-2 cells. Microb Pathog 52, 165-176.

Hacker, E. 96

References

Sandvig, K. & van Deurs, B. (1994). Endocytosis and intracellular sorting of ricin and Shiga toxin. FEBS Lett 346, 99-102.

Sangal, V. & Hoskisson, P. A. (2014). Corynephages: infections of the infectors. In Diphtheria and its Etiological Agents, pp. 67-82. Edited by A. Burkovski. Dordrecht: Springer Netherlands.

Sangal, V., Nieminen, L., Weinhardt, B., Raeside, J., Tucker, N. P., Florea, C. D., Pollock, K. G. & Hoskisson, P. A. (2014). Diphtheria-like disease caused by toxigenic Corynebacterium ulcerans strain. Emerg Infect Dis 20, 1257-1258.

Santos, L. S., Antunes, C. A., Santos, C. S., Pereira, J. A., Sabbadini, P. S., Luna, M., Azevedo, V., Hirata Junior, R., Burkovski, A., Asad, L. M. & Mattos-Guaraldi, A. L. (2015). Corynebacterium diphtheriae putative tellurite-resistance protein (CDCE8392_0813) contributes to the intracellular survival in human epithelial cells and lethality of Caenorhabditis elegans. Mem Inst Oswaldo Cruz 110, 662-668.

Schuhegger, R., Schoerner, C., Dlugaiczyk, J., Lichtenfeld, I., Trouillier, A., Zeller- Peronnet, V., Busch, U., Berger, A., Kugler, R., Hörmansdorfer, S. & Sing, A. (2009). Pigs as source for toxigenic Corynebacterium ulcerans. Emerg Infect Dis 15, 1314-1315.

Sekizuka, T., Yamamoto, A., Komiya, T., Kenri, T., Takeuchi, F., Shibayama, K., Takahashi, M., Kuroda, M. & Iwaki, M. (2012). Corynebacterium ulcerans 0102 carries the gene encoding diphtheria toxin on a prophage different from the C. diphtheriae NCTC 13129 prophage. BMC Microbiol 12, 72.

Senior, N. J., Bagnall, M. C., Champion, O. L., Reynolds, S. E., La Ragione, R. M., Woodward, M. J., Salguero, F. J. & Titball, R. W. (2011). Galleria mellonella as an infection model for Campylobacter jejuni virulence. J Med Microbiol 60, 661-669.

Seto, Y., Komiya, T., Iwaki, M., Kohda, T., Mukamoto, M., Takahashi, M. & Kozaki, S. (2008). Properties of corynephage attachment site and molecular epidemiology of Corynebacterium ulcerans isolated from humans and animals in Japan. Jpn J Infect Dis 61, 116-122.

Sifri, C. D., Begun, J. & Ausubel, F. M. (2005). The worm has turned--microbial virulence modeled in Caenorhabditis elegans. Trends Microbiol 13, 119-127.

Silva, A. S., Baraúna, R. A., de Sá, P. C., das Gracas, D. A., Carneiro, A. R., Thouvenin, M., Azevedo, V., Badell, E., Guiso, N., da Silva, A. L. & Ramos, R. T. (2014). Draft genome sequence of Corynebacterium ulcerans FRC58, isolated from the bronchitic aspiration of a patient in France. Genome Announc 2.

Simpson-Louredo, L., Ramos, J. N., Peixoto, R. S., Santos, L. S., Antunes, C. A., Ladeira, E. M., Santos, C. S., Vieira, V. V., Boas, M. H., Hirata, R., Jr. & Mattos-Guaraldi, A. L. (2014). Corynebacterium ulcerans isolates from humans and dogs: fibrinogen, fibronectin and collagen-binding, antimicrobial and PFGE profiles. Antonie Van Leeuwenhoek 105, 343-352.

Sing, A., Hogardt, M., Bierschenk, S. & Heesemann, J. (2003). Detection of differences in the nucleotide and amino acid sequences of diphtheria toxin from Corynebacterium diphtheriae and Corynebacterium ulcerans causing extrapharyngeal infections. J Clin Microbiol 41, 4848- 4851.

Sing, A., Bierschenk, S. & Heesemann, J. (2005). Classical diphtheria caused by Corynebacterium ulcerans in Germany: amino acid sequence differences between diphtheria toxins from Corynebacterium diphtheriae and C. ulcerans. Clin Infect Dis 40, 325-326.

Hacker, E. 97

References

Sykes, J. E., Mapes, S., Lindsay, L. L., Samitz, E. & Byrne, B. A. (2010). Corynebacterium ulcerans bronchopneumonia in a dog. J Vet Intern Med 24, 973-976.

Tauch, A. & Sandbote, J. (2014). The Family Corynebacteriaceae. In The Prokaryotes, pp. 239-277. Edited by E. Rosenberg, E. DeLong, S. Lory, E. Stackebrandt & F. Thompson: Springer Berlin Heidelberg.

Tauch, A. & Burkovski, A. (2015). Molecular armory or niche factors: virulence determinants of Corynebacterium species. FEMS Microbiol Lett 362.

Taylor, D. J., Efstratiou, A. & Reilly, W. J. (2002). Diphtheria toxin production by Corynebacterium ulcerans from cats. Vet Rec 150, 355.

Tejedor, M. T., Martin, J. L., Lupiola, P. & Gutierrez, C. (2000). Caseous lymphadenitis caused by Corynebacterium ulcerans in the dromedary camel. Can Vet J 41, 126-127.

Tenor, J. L., McCormick, B. A., Ausubel, F. M. & Aballay, A. (2004). Caenorhabditis elegans-based screen identifies Salmonella virulence factors required for conserved host- pathogen interactions. Curr Biol 14, 1018-1024.

Terwisscha van Scheltinga, A. C., Hennig, M. & Dijkstra, B. W. (1996). The 1.8 A resolution structure of hevamine, a plant chitinase/lysozyme, and analysis of the conserved sequence and structure motifs of glycosyl hydrolase family 18. J Mol Biol 262, 243-257.

Ton-That, H. & Schneewind, O. (2003). Assembly of pili on the surface of Corynebacterium diphtheriae. Mol Microbiol 50, 1429-1438.

Trost, E., Al-Dilaimi, A., Papavasiliou, P., Schneider, J., Viehoever, P., Burkovski, A., Soares, S. C., Almeida, S. S., Dorella, F. A., Miyoshi, A., Azevedo, V., Schneider, M. P., Silva, A., Santos, C. S., Santos, L. S., Sabbadini, P., Dias, A. A., Hirata, R., Jr., Mattos- Guaraldi, A. L. & Tauch, A. (2011). Comparative analysis of two complete Corynebacterium ulcerans genomes and detection of candidate virulence factors. BMC Genomics 12, 383. van Meeteren, L. A., Frederiks, F., Giepmans, B. N., Pedrosa, M. F., Billington, S. J., Jost, B. H., Tambourgi, D. V. & Moolenaar, W. H. (2004). Spider and bacterial sphingomyelinases D target cellular lysophosphatidic acid receptors by hydrolyzing lysophosphatidylcholine. J Biol Chem 279, 10833-10836.

Varki, A. (1993). Biological roles of oligosaccharides: all of the theories are correct. Glycobiology 3, 97-130.

Varol, B., Özerman Edis, B. & Bektas, M. (2014). Toxin Structure, Delivery and Action. In Corynebacterium diphtheriae and Related Toxigenic Species, pp. 83-94. Edited by A. Burkovski. Dordrecht: Springer Netherlands.

Ventura, M., Canchaya, C., Tauch, A., Chandra, G., Fitzgerald, G. F., Chater, K. F. & van Sinderen, D. (2007). Genomics of Actinobacteria: tracing the evolutionary history of an ancient phylum. Microbiol Mol Biol Rev 71, 495-548.

Viana, M. V., de Jesus Benevides, L., Batista Mariano, D. C., de Souza Rocha, F., Bagano Vilas Boas, P. C., Folador, E. L., Pereira, F. L., Alves Dorella, F., Gomes Leal, C. A., Fiorini de Carvalho, A., Silva, A., de Castro Soares, S., Pereira Figueiredo, H. C., Azevedo, V. & Guimaraes, L. C. (2014). Genome sequence of Corynebacterium ulcerans strain 210932. Genome Announc 2.

Hacker, E. 98

References

Vimr, E. & Lichtensteiger, C. (2002). To sialylate, or not to sialylate: that is the question. Trends Microbiol 10, 254-257.

Vimr, E. R. (1994). Microbial sialidases: does bigger always mean better? Trends Microbiol 2, 271-277.

Vimr, E. R., Kalivoda, K. A., Deszo, E. L. & Steenbergen, S. M. (2004). Diversity of microbial sialic acid metabolism. Microbiol Mol Biol Rev 68, 132-153.

Wagner, K. S., White, J. M., Crowcroft, N. S., De Martin, S., Mann, G. & Efstratiou, A. (2010). Diphtheria in the United Kingdom, 1986-2008: the increasing role of Corynebacterium ulcerans. Epidemiol Infect 138, 1519-1530.

Walker, J., Jackson, H. J., Eggleton, D. G., Meeusen, E. N., Wilson, M. J. & Brandon, M. R. (1994). Identification of a novel antigen from Corynebacterium pseudotuberculosis that protects sheep against caseous lymphadenitis. Infect Immun 62, 2562-2567.

Wilson, M. J., Brandon, M. R. & Walker, J. (1995). Molecular and biochemical characterization of a protective 40-kilodalton antigen from Corynebacterium pseudotuberculosis. Infect Immun 63, 206-211.

Yoshimura, Y., Tachikawa, N., Komiya, T. & Yamamoto, A. (2014). A case report and epidemiological investigation of axillary lymph node abscess caused by Corynebacterium ulcerans in an HIV-1-positive patient. Epidemiol Infect 142, 1541-1544.

Hacker, E. 99

Curriculum vitae

6 Curriculum vitae

Elena Hacker

Geboren am 27.12.1986 in Erlangen

Akademische Ausbildung

09/2015 – 03/2016 Koordination des integrierten Graduiertenkollegs des SFB796

06/2011 – 03/2016 Promotion Friedrich-Alexander-Universität Erlangen-Nürnberg Lehrstuhl für Mikrobiologie Betreuer: Prof. Dr. Andreas Burkovski Thema: „Characterization of virulence of Corynebacterium ulcerans“/ “Charakterisierung der Virulenzeigenschaften von Corynebacterium ulcerans“ Assoziiertes Mitglied des integrierten Graduiertenkollegs des SFB796

10/2009 – 03/2011 Masterstudium Molecular Life Science Friedrich-Alexander-Universität Erlangen-Nürnberg Studienschwerpunkte: Drug Discovery, Lebensmittelchemie, Molekularbiologie

Masterarbeit am Lehrstuhl für Mikrobiologie Betreuer: Prof. Dr. Andreas Burkovski Thema: “Characterization of virulence of Corynebacterium ulcerans”,

31.03.2011 Abschluss: Master of Science

10/2006 – 10/2009 Bachelorstudium Molecular Life Science Friedrich-Alexander-Universität Erlangen-Nürnberg Studienschwerpunkte: organische und anorganische Chemie, pharmazeutische Chemie, Molekularbiologie

Bachelorarbeit am Lehrstuhl für Mikrobiologie Betreuer: Prof. Dr. Wolfgang Hillen Thema: “Charakterisierung der Induktion von Derivaten des Transkriptionsregulators TetR ”

15.10.2009 Abschluss: Bachelor of Science

06/2006 Abitur Albert-Schweitzer-Gymnasium, Erlangen

Hacker, E. 100

Curriculum vitae

Fortbildungen/ Workshops/Präsentationen

05/2015 Bioinformatik Workshop

12/2014 Workshop in Fluoreszenzmikroskopie und Bildanalyse

10/2014 Statistikkurs

02/2014 GMP Schulung

2013 Workshops in Scientific writing und Presenting professionally

2011 – 2015 Teilnahme an verschiedenen Tagungen mit Posterpräsentationen oder Vorträgen (englisch)

Qualifikationen

Methoden Arbeiten im S2 Labor mit humanpathogenen Bakterien Zellkultur (Infektionsassays, Antikörperfärbungen, Durchflusszytometrie) Mikroskopie (Konfokal- und Fluoreszenzmikroskopie), ELISA RNA/DNA-Techniken (PCR, RNA-/DNA-Isolation, Southern Blot, Dot Blot, Klonierungen) Verantwortlichkeiten Planung und Betreuung des Mastermoduls „Pathogenitätsmechanismen in Gram-positiven Bakterien“ Betreuung von Studenten in Master- und Bachelorarbeiten EDV Kenntnisse MS Office Photoshop, ImageJ, Kaluza Sprachkenntnisse Deutsch (Muttersprache) Englisch (verhandlungssicher) Spanisch (Grundlagen), Französisch (Grundlagen)

Publikationen

Hacker, E., Ott, L., Hasselt, K., Mattos-Guaraldi, A. L., Tauch, A. & Burkovski, A. (2015). Colonization of human epithelial cell lines by Corynebacterium ulcerans from human and animal sources. Microbiology 161, 1582-1591. Hacker, E., Ott, L., Schulze-Luehrmann, J., Lührmann, A., Wiesmann, V., Wittenberg, T., & Burkovski, A. (2015). The killing of macrophages by Corynebacterium ulcerans. Virulence, 2015 Dec 2, doi: 10.1080/21505594.2015.1125068. Antunes, C. A., Sanches Dos Santos, L., Hacker, E., Köhler, S., Bösl, K., Ott, L., de Luna, M., Hirata, R., Jr., Azevedo, V. A., Mattos-Guaraldi, A. L. & Burkovski, A. (2015). Characterization of DIP0733, a multi-functional virulence factor of Corynebacterium diphtheriae. Microbiology 161, 639-647. Antunes, C. A., Clark, L., Wanuske, M. T., Hacker, E., Ott, L., Simpson-Louredo L., , das Gracas de Luna M., Hirata, R., Jr., Mattos-Guaraldi, A. L., Hodgkin J., & Burkovski, A. (2015). Caenorhabditis elegans star formation and negative chemotaxis induced by infection with corynebacteria. Microbiology, 2015 Oct 20, doi: 10.1099/mic.0.000201.

Hacker, E. 101

Acknowledgment/Danksagung

7 Acknowledgment/Danksagung

Mein größter Dank geht an Prof. Dr. Andreas Burkovski für die Bereitstellung des interessanten und vielseitigen Themas, die immer offene Tür, die vielen guten Tipps und Vorschläge und natürlich für die Begutachtung dieser Arbeit. Außerdem möchte ich mich ganz herzlich für die insgesamt lockere und angenehme Arbeitsatmosphäre bei ihm bedanken. Ich hätte mir keinen besseren Chef vorstellen können! Weiterhin möchte ich apl. Prof. Dr. Andreas Tauch von der Universität Bielefeld für die Erstellung des Zweitgutachtens und die Bereitstellung der C. ulcerans-Isolate danken. Danke an PD Dr. Anja Lührmann für das Mentoring im Rahmen des SFB796 und die vielen hilfreichen Tipps für meine Arbeit. Danke an Dr. Ralf Palmisano und an das Team vom OICE für die Hilfe bei allem was die Fluoreszenzmikroskopie betroffen hat. Ein ganz großes Dankeschön geht auch an alle ehemaligen und aktuellen Mitarbeiter des Mibi-Lehrstuhls für die schöne Zeit, die ich hier verbracht habe! Besonders hervorheben möchte ich unsere Sekretärin Frau Wehr, danke für die vielen Lacher, Gerald, der immer irgendwie ALLES gemanagt hat und für lustige neue Namenskreationen sorgte, Manu und Markus, die einen großen Beitrag dazu geleistet haben, dass alle immer reibungslos ihre Experimente durchführen konnten und Susi und Klaus für die netten Pausengespräche. Ein noch größeres Dankeschön an das gesamte Burki-Labor, ich habe in den letzten Jahren hier so viele nette Menschen kennengelernt und Freundschaften geschlossen, dass es den Rahmen sprengen würde, alle aufzuzählen. Es war eine wunderbare Zeit! Ganz besonders möchte ich Lisa danken, die mich in die ganze Thematik eingeführt hat und in vielen Angelegenheiten eine große Hilfe war, außerdem meiner Lieblings-Masterstudentin Julia E. für die schöne Zeit im Labor und Marie für die ganzen netten Aufmunterungen und die vielen, vielen Gespräche in teilweise etwas schwierigen Zeiten. Danke auch an Julia B. für die nette Zeit im gemeinsamen Büro! A very special thanks to all the international people in the lab, particularly to the Brazilian girls Camila, Dayana and Renata. Camila, thank you for the wonderful time we had together working in the lab, for the delicious Caipirinhas, the awesome brigadeiros, and for becoming a really good friend! You made life easy!!!! Thank you Ian, for some final English language corrections of this thesis. Zum Glück gibt es auch ein außer-universitäres Leben, deswegen natürlich auch danke an alle meine Freunde, die dieses mit viel Spaß gefüllt haben und auf die immer Verlass ist. Danke Domi für die Hilfe bei der Formatierung, ohne dich wär die Arbeit jetzt nicht so schön! Zu guter Letzt möchte ich mich bei meinen wunderbaren Eltern und meiner supertollen Schwester für einfach alles bedanken. Ich bin froh, dass ich euch hab, ihr seid die besten!!

Hacker, E. 102