Vibrio Cholerae Predator Interaction and Virulence

Home , Vibrio cholerae

Modulators of Vibrio cholerae predator interaction and virulence

Barbro Lindmark

Department of Molecular Biology 901 87 Umeå Umeå 2009

Till Viktor, Rosita och Ludvig

“As our circle of knowledge expands, so does the

circumference of darkness surrounding it”

-Albert Einstein

Table of Contents

ABSTRACT ...... - 1 - PAPERS IN THIS THESIS...... - 2 - INTRODUCTION ...... - 5 - CHOLERA...... - 5 - PATHOGENESIS AND VIRULENCE FACTORS ...... - 6 - Cholera toxin and toxin co-regulated pilus, the major virulence factors of O1 and O139 V. cholerae...... - 6 - Additional virulence factors ...... - 8 - Vibrio cholerae LPS ...... - 11 - Regulation of virulence factors...... - 11 - Quorum sensing...... - 12 - PROTEIN SECRETION ...... - 13 - General secretion and two-arginine (Tat) pathways- 14 - Type I secretion system (T1SS), ABC transporters ..- 14 - Type II secretion system (T2SS), secreton-dependent pathway...... - 15 - Type III secretion system (T3SS), the contact dependent pathway...... - 15 - Type IV secretion system (T4SS) ...... - 16 - Type V secretion system, autotransporters and two- partner secretion...... - 17 - Chaperone/ usher (CU) pathway...... - 18 - Type VI secretion system (T6SS) ...... - 18 - Vesicle-mediated secretion ...... - 20 - PERSISTENCE AND SURVIVAL OF ...... - 24 - Biofilm formation...... - 24 - Interactions of V. cholerae with chitin ...... - 25 - NON-O1 NON-O139 VIBRIO CHOLERAE ...... - 26 - Virulence factors...... - 27 - Serum resistance...... - 27 -

V. CHOLERAE HOST INTERACTION MODELS ...... - 28 - Mammalian models...... - 28 - Non-mammalian models...... - 29 - C. ELEGANS ...... - 30 - Innate immunity in C. elegans...... - 30 - C. elegans as a host model for bacterial infections .- 33 - AIMS OF THE THESIS...... - 35 - RESULTS AND DISCUSSION ...... - 36 - PAPER I...... - 36 - A Vibrio cholerae protease needed for killing of Caenorhabditis elegans has a role in protection from natural predator grazing ...... - 36 - PAPER II ...... - 39 - The metalloprotease PrtV from Vibrio cholerae...... - 39 - PAPER III ...... - 41 - Role of LPS in vesicle-mediated export of Vibrio cholerae PrtV protease...... - 41 - PAPER IV...... - 43 - Evaluation of the presence of virulence-associated factors in Vibrio cholerae non-O1 non-O139 isolates from clinical and environmental sources ...... - 43 - CONCLUSIONS...... - 46 - SAMMANFATTNING PÅ SVENSKA...... - 47 - ACKNOWLEDGEMENTS...... - 49 - REFERENCES ...... - 51 -

PAPER I-IV

Abstract Vibrio cholerae, the causal agent of cholera typically encodes two critical virulence factors: cholera toxin (CT), which is primarily responsible for the diarrhoeal purge, and toxin-co-regulated pilus (TCP), an essential colonisation factor. Nontoxi- genic strains expressing TCP can efficiently acquire the CT gene through lysogenic conversion with CTXΦ, a filamentous phage that encodes CT and uses TCP as a receptor. V. cholerae is a Gram-negative bacterium and a natural inhabitant of estuarine and coastal waters throughout both temperate and tropical regions of the world. In the aquatic environment, V. cholerae encounters several environmental stresses, such as change in salinity, UV stress, nutrient limitation, temperature fluctuations, viral infections and protozoan predation. To fully understand the pathogenic and virulence potential of V. cholerae, knowledge is required of its interactions with, not only human, but also environmental factors. By using the nematode Caenorhabdi- tis elegans as host model, we were able to identify a previously uncharacterised protein, the extracellular protease PrtV. PrtV was shown to be required for the killing of. elegans and also necessary for survival from grazing by the ciliate Tetrahymena pyriformis and the flagellate Cafeteria roenbergensis. The PrtV protein, which belongs to a M6 family of metallopeptidases was cloned and purified for further charac- terisations. The purified PrtV was cytotoxic against the human intestinal cell line HCT8. By using human blood plasma, fibrinogen, fibronectin and plasminogen were identified as candidate substrates for the PrtV protease. Outer membrane vesicles (OMVs) are released to the surroundings by most Gram-negative bacteria through “bulging and pinching” of the outer membrane. OMVs have been shown to contain many virulence factors important in pathogenesis. Therefore, we investigated the association of PrtV with OMVs. PrtV was not associated with OMVs from the wild type O1 strain. In contrast, in an LPS mutant lacking two sugar chains in the core oligosaccharide PrtV was found to be associated with the OMVs. The OMV-associated PrtV was shown to be proteolytically and cytotoxically active. V. cholerae strains are grouped into >200 serogroups. Only the O1 and O139 serogroups have been associated with pandemic cholera, a severe diarrhoeal disease. All other serogroups are collectively referred to as non-O1 non-O139 V. cholerae. Non-O1 non-O139 V. cholerae can cause gastroenteritis and extraintestinal infections, but unlike O1 and O139 strains of V. cholerae, little is known about the virulence gene content and their potential to become human pathogens. We analysed clinical and environmental non-O1 non-O139 isolates for their putative virulence traits. None of them carry the genes encoding CT or the TCP, but other putative virulence factors were present in these isolates. The incidence of serum resistance was found to vary considerably and was independent of encapsulation. Three strains were strongly serum-resistant, and these same strains could also kill C. elegans.

- 1 -

Papers in this thesis

Paper I

Vaitkevicius K, Lindmark B, Ou G, Song T, Toma C, Iwanaga M, Zhu J, Andersson A, Hammarstrom ML, Tuck S, Wai SN. A Vibrio cholerae protease needed for killing of Caenorhabditis elegans has a role in protection from natural predator grazing. Proc Natl Acad Sci U S A. 2006 Jun 13;103(24):9280-5

Paper II

Vaitkevicius K, Rompikuntal PK, Lindmark B, Vaitkevicius R, Song T, and Wai SN. The metalloprotease PrtV from Vibrio cholerae. FEBS J. 2008 Jun;275(12):3167-77

Paper III

Barbro Lindmark.,Pramod Rompikuntal., Joachim Reidl and Sun Nyunt Wai. Role of LPS in vesicle mediated export of Vibrio cholerae PrtV protease. (Manuscript).

Paper IV

Barbro Lindmark., Amit Pal., Pramod Kumar Rompikuntal., Mitesh Dongre., Goutam Chowdhury., Betty Collin., Ann-Sofi Rehnstam-Holm., T Ramamurthy., and Sun Nyunt Wai. Evaluation of the presence of virulence-associated factors in Vibrio cholerae non-O1 non O139 isolates from clinical and environmental sources (Submitted manuscript)

- 2 -

Contribution has also been made by the author to the following studies not included in this thesis:

Barbro Lindmark, Pramod Kumar Rompikuntal, Karolis Vaitkevicius, Tianyan Song, Yoshimitsu Mizunoe, Bernt Eric Uhlin, Patricia Guerry, and Sun Nyunt Wai Outer Outer membrane vesicle-mediated release of Cyto- lethal Distending Toxin (CDT) from Campylobacter jejuni BMC Micobiol 2009 (9) 220

Ishikawa T, Rompikuntal PK, Lindmark B, Milton DL, Wai SN.Quorum sensing regulation of the two hcp alleles in Vibrio cholerae O1 strains. PLoS One. 2009 Aug 24;4(8):e6734.

Song T, Mika F, Lindmark B, Liu Z, Schild S, Bishop A, Zhu J, Camilli A, Johansson J, Vogel J, Wai SN. A new Vibrio cholerae sRNA modulates colonization and affects release of outer membrane vesicles. Mol Microbiol. 2008 Oct;70(1):100-11

Ou G, Rompikuntal PK, Bitar A, Lindmark B, Vaitkevicius K, Wai SN, Hammarström ML Vibrio cholerae cytolysin causes an inflammatory response in human intestinal epithelial cells that is modulated by the PrtV protease PLoS One 2009 Nov 12;4,e7806

Sun Nyunt Wai, Barbro Lindmark, Tomas Söderblom, Marie Wester- mark, Jan Oscarsson, Akemi Takade, Jana Jass, Agneta Richter-Dahlfors, Yoshimitsu Mizunoe, Bernt Eric Uhlin. Vesicle-mediated export and assembly of pore-forming oligomers of the enterobacterial ClyA cytotoxin Cell. 2003 11 5: 25-35

- 3 -

Abbreviations

CT/ctx Cholera toxin

DNA Deoxyribonucleic acid

HA/P Haemagglutinin protease

LPS Lipopolysaccharide

OMP Outer membrane proteins

OMV Outer membrane vesicles

PrtV Protease of Vibrio cholerae

RNA Ribonucleic acid

Tcp Toxin co-regulated pilus

VCC/hlyA Vibrio cholerae cytolysin

VPI Vibrio pathogenicity island

- 4 -

Introduction

Cholera

Vibrio cholerae is a Gram-negative, motile, comma- shaped bacterium with a single polar flagella. It can be isolated from estuarine and aquatic environments and is well- recognised as the causative agent of the severe dehydrating diarrhoeal disease cholera, which occurs frequently as epidemics in many developing countries1, 2. Cholera is transmitted via contaminated food or water, usually with a large inoculum of bacteria. In human volunteer studies, the infectious dose was deter- mined to range from 106 to 1011 colony-forming units, depending on the inoculating conditions. Infected individuals can dis- seminate as many as 1013 organisms per day in stool, and further contaminate food or water supplies that can lead to a rapid spread among the population in areas with inadequate treatment of sewage and drinking water. The mortality for untreated cholera infection is 20-50%. The current management of cholera is immediate treatment with oral dehydration solutions (ORS), a prepackaged mixture of sugar and salts to be mixed with water and taken orally in large amounts. This solution is used throughout the world to treat diarrhoea. Severe cases also require intravenous fluid replacement. With prompt rehydration treatment, fewer than 1% of cholera patients die3.

Out of more than 208 serogroups, based on the O- antigen in the LPS structure, only serogroups O1 and O139 have caused pandemic cholera. The O1 serogroup is divided into two biotypes, classical and El Tor, depending on biochemical properties, such as Voges-Proskauer reaction, haemolysis, haemag- glutination and sensitivity to polymyxin B and phage infection4. The sixth pandemic, and probably the first five, were caused by the classical biotype of V. cholerae O1. The seventh pandemic, which is still ongoing, started in 1961 and is caused by the El Tor biotype of V. cholerae O1. In 1992 due to the emergence of a new serogroup, there was a explosive cholera outbreak in India and Bangladesh that later spread throughout southeast Asia5, 6. Genetic analysis of the new serogroup, designated O139 Bengal,

- 5 -

revealed that the 22-kb gene segment encoding the O1 El Tor antigen had been replaced by a 35-kb DNA fragment encoding the O139 side chain7 and also a capsular polysaccharide (CPS)8, 9. Both the O139 side chain and the CPS were shown to be important for serum resistance and for colonisation of mouse small intestine8. The rapid spread of the O139 strain was partly due to a population with lack of protective immunity against the O-antigen, and also because the strain is resistant to O1 lytic phages10. Later, it was shown that the O139 serogroup com- prised diverse clones of epidemic and non-epidemic strains that were derived from multiple V. cholerae strains, both O1 or non- O1 progenitors11. A recent study showed that serogroup conversion can be mediated by chitin-induced natural transformation. This can occur between two different V. cholerae serogroups living within the same biofilm on a natural chitin surface10. This mechanism of horizontal gene transfer together with conjugation and phage transduction might explain the great diversity between different V. cholerae strains12-14.

Pathogenesis and virulence factors

Cholera toxin and toxin co-regulated pilus, the major virulence factors of O1 and O139 V. cholerae

Infection due to V. cholerae starts with the digestion of contaminated food or water. After passage through the acid barrier of the stomach, V. cholerae causes cholera using its two major virulence factors: the toxin co-regulated pilus (TCP) that is important for colonisation of the intestine, and the cholera toxin (CT), which is responsible for typical symptoms of cholera the “rice water stool”. The V. cholerae genome has been completely sequenced. It consists of two chromosomes, one of approximately 3 Mb in size (chromosome I) and another of around 1 Mb in size (chromosome II). Chromosome I harbours most of the known virulence factors, including CT and TCP 15. CT is encoded by the ctxAB genes, which form part of a single- stranded DNA lysogenic filamentous bacteriophage (CTXΦ)12. The 6.9-kb CTXΦ genome contains two regions the RS2 and the core region. The RS2 region contains genes required for phage DNA replication and site-specific integration. The core region

- 6 -

contains genes required for the morphogenesis of phage parti- cles (cep, psh, pIIICTX, ace and zot) and the ctxAB genes16. The ctxAB genes do not directly participate in phage particle formation, but they might be important for the phage to provide a survival advantage to its host in the human gastrointestinal environment. The gene cluster necessary for the formation and assembly of TCP is located in region called the Vibrio pathogenicity island (VPI). The VPI region is nearly 40 kb, and is flanked on both sides by attachment sequences, and carries a putative integrase gene and a transposase gene. The structure, GC content, and codon usage suggests that acquisition of the VPI was a quite recent event. Interestingly, the CTXΦ uses TCP as a receptor for attachment and transmission, thus providing an evolutionary link between two horizontally moving ele- ments17.

CT is a classical AB toxin and an oligomeric protein composed of one A-subunit (CT-A) and five B-subunits (CT-B). The CT-A subunit is proteolytically cleaved by V. cholerae protease into two fragments, CTA1 and CTA2, which remain linked to each other by a disulfide bond. The CTA1 fragment contains the ADP-ribosylating activity and the CTA2 fragment serves to insert CTA into the CTB pentamer. The CT-B subunits bind to the monosialoganglioside GM1 receptors localised mainly in lipid rafts on the host cell surface18. The CT is endocytosed and transported to the endoplasmatic reticulum where the CTA1 fragment dissociates from the toxin complex. The CTA1 fragment is transported to the cytoplasm of the target cell where it ADP-ribosylates a GTP-binding protein that regulates adenylyl cyclase activity, causing a constitutively activated adenylyl cyclase and a marked increase of the intracellular cAMP level. The increased level of cAMP results in an imbalance in electrolyte movement in the epithelial cell. The net increase in NaCl- secretion leads to a large osmotic movement of water into the intestinal lumen. This results in the distinctive features of cholera, the painless purging of voluminous stools resembling rice water. In adults, the rate of diarrhoea may reach be 500-1000 ml/h, but in extreme cases the fluid loss can be as much as 30- 40 l/day4.

- 7 -

Additional virulence factors

CT and TCP are undoubtedly the main virulence factors of V. cholerae; however, in volunteer studies CT- strains still cause mild to severe diarrhoea. Several other putative virulence factors have been described that might contribute to the pathogenesis of V. cholerae. These include motility, the mannose- sensitive haemagglutinin (MSHA) pilus, LPS and OmpU, which might contribute to colonisation and various secreted factors such as the haemagglutinin protease (HA/P). A variety of less characterized toxins have also been suggested to contribute to cholera diarrhoea. With the exception of Vibrio cytolysin (VCC) and the multifunctional autoprocessing RTX (MARTXVc) toxin most of the genes or gene products are only present in a subset of V. cholerae strains2, 19. The roles of these accessory factors in cholera pathogenesis are not well established, but it is likely that at least some of them play a role in environmental fitness.

Vibrio cholerae cytolysin/haemolysin

VCC, also called haemolysin A or El Tor haemolysin, is encoded by the hlyA gene, which is a highly conserved genetic element in V. cholerae, independent of biotypes or sero- groups20. Haemolysis of sheep erythrocytes was traditionally used to distinguish between classical and El Tor biotypes of V. cholerae. Genotypically, however, it has been shown that the structural gene of hlyA has an 11-bp deletion in the classical biotype, which results in a frame-shift and a production of a truncated non-haemolytic but cytotoxic protein2. VCC is synthesised and secreted as a pro-VCC, a protein of about 80 kDa which is proteolytically activated by proteases (e.g. HA/P, trypsin, subtil- isin, papain) to a 65-kDa mature VCC after cleavage of the 15- kDa N-terminal region21. In the presence of cholesterol- and ceramide-rich membranes, the mature VCC forms heptameric oligomers, which insert into lipid membranes, generating pores with an internal diameter of 1-2 nm22-25. The VCC has been reported to cause fluid accumulation in ligated rabbit ileal loops, as well as several cytotoxic effects and extensive vacuolation in different cell types22, 26-29. Moreover, it was recently demonstrated that VCC-formed pores were capable of inducing chloride efflux from human intact sigmoid colon mucosal sheets.

- 8 -

The virtue of such an activity, it was suggested, was that VCC might be the major diarrhoeagenic factor for the non-cholera toxin producing strains, or contribute to causing diarrhoea when the toxin is present30. In our recent study, we showed that VCC alone can induce an inflammatory response in human intestinal cells characterised by increased epithelial permeability and production of IL-8 and TNF-α by human epithelial cells31.

Haemagglutinin/ protease (HA/P)

It was suggested that the bacteria need to produce mucinases in order to penetrate the mucus layer. One such mucinase could be the HA/P, also called mucinase or detachase. The secreted metalloprotease HA/P is encoded by the hapA gene. Mutants lacking a functional hapA gene show very little protease activity, suggesting that HA/P is responsible for the major proteolytic activity exerted by V. cholerae. HA/P has activity against fibronectin, ovomucin, and lactoferrin. VCC and the CT-A subunit have also been shown to be proteolytically activated by HA/P21, 32. The contribution of HA/P to pathogenesis is not fully understood, since conflicting results have been obtained with different strains, mutants, and animal models33- 35. Loss of HA/P has, in some causes, caused increased colonisation, suggesting that HA/P promotes the detachment of the bacteria from the intestine back into the environment36-38.

The PrtV protease

The PrtV protease of V. cholerae was identified in 1997 by Oigerman et al as a 102-kDa metalloprotease. The gene encoding PrtV is located on chromosome II within a putative pathogenicity island. The close proximity of the prtV gene to the hlyA gene suggested that the former gene product might be capable of damaging host cells or proteolytically activating the VCC. However, PrtV null mutants did not show any attenuation in the infant mouse model or reduction in colonisation potential compared with wild type39, 40. The haemolytic activity was also unaffected40. The PrtV protease, belongs to a M6 metalloen- dopeptidase family. The members of the M6 family all display a HEXXHXXGXXD motif. The two histidines and the aspartate are zinc ligands and the glutamate is the catalytic residue. The

- 9 -

best characterised member of this family is immune inhibitor A (InA or InhA) from Bacillus thuringiensis, a bacterium commonly used in agriculture for the control of insect pests41. InhA was first identified in B. thuringiensis as a secreted inhibitor of the humoural defence system in Hyalophora cecropia against E. coli but not B. cereus42. Purified InhA was lethal when in- jected into Callosamia promethea pupae and adult Drosophila melanogaster43. The peptidase was capable of digesting two different classes of antibacterial factors found in the insect haemolymph: cecropins and attachins. Specificity experiments of InhA towards synthetic and natural cecropins did not show a strict requirement for a specific amino acid sequence: and several peptide bonds were cleaved, but the peptidase showed se- lectivity for hydrophobic residue on the C-terminal side44. The specificity of InhA was explained in terms of the open structures of cercropins which are largely folded in random coil structures in solution. Attacins may also have an open structure since they lack disulfide bridges. In agreement with this structure requirement, the globular proteins ribonuclease and myoglobin were resistant to InhA unless the disulfide bridges were reduced or alkylated44.

Using non-mammalian host models, we found that PrtV was required for the killing of Caenorhabditis elegans, and was necessary for survival from grazing by the flagellate Cafeteria roenbergensis and the ciliate Tetrahymena pyriformis. More- over, supernatant from a V. cholerae PrtV deletion mutant caused increased interleukin-8 (IL-8) secretion in human intestinal epithelial T84 cells and increased lysis of red blood cells compared with supernatant from the parental strain31, 39. The inflammatory response in T84 cells mediated by VCC was found to be modulated the PrtV protease. The haemolytic and immuno-modulating properties of the PrtV mutant were dependent on the growth media. When grown in LB, the supernatant from the PrtV mutant was more haemolytic and caused a stronger inflammatory response compared to wild type. On the other hand, no differences between the PrtV mutant and wild type in haemolytic activity or inflammatory response could be observed using supernatants from BHI broth. We suggested that the content(s) of the BHI media might inhibit the activity or the expression of proteases from V. cholerae.

- 10 -

Vibrio cholerae LPS

LPS is one of the major cell-surface-associated polysac- charides produced by V. cholerae, including (i) lipid A, (ii) core polysaccharide (core PS), and (iii) the O-antigen polysaccharide (O-PS). The O-PS is the antigenic part of the LPS and all serogroups have a distinct structure. The lipid A region and the core PS are important for maintenance of cell membrane integrity and present essentially similar structures among different V. cholerae serogroups45. Lipid-A is the key constituent exhibiting endotoxic properties. The genes involved in core PS biosynthesis are found in the wav gene cluster. The wav gene cluster is located on chromosome I and comprises ORFs VC0223 to VC0240 in the sequenced genome of V. cholerae O1 El Tor strain N16961 46. The O1 antigen biosynthesis gene cluster (wbe) is localised upstream of VC0240, indicating that in V. cholerae most of the LPS biosynthetic genes are clustered. Fur- ther analyses of 38 different V. cholerae strains revealed the presence of five different wav gene cluster types. The El Tor wav gene cluster type is prevalent among clinical isolates of different serogroups associated with cholera and environmental O1 strains. In contrast, the other 19 isolates, belonging to non- O1 non-O139 strains not associated with cholera-like disease contained four other types of wav gene clusters. These four other types differ from each other in distinct loci. The predicted structural differences of the five wav gene cluster types are suggested to be located in the side branches of the core OS. Ten wav genes (wavA, wavB, waaC, wavC, waaA, wavI, waaF, waaL, wavL, and gmdH) were found in all 38 strains, indicating the presence of a common core-PS backbone structure46.

Regulation of virulence factors

V. cholerae has evolved with a variety of signal transduction and gene regulation systems to adapt to and respond to different conditions, such as temperature, pH, osmolarity, CO2 concentrations, bile salts and other environmental stimuli, some of which are shown in Fig 1. The most well-defined signal transduction cascade is the ToxR signalling circuit which is responsible for sensing and integrating environmental information to regulate the transcripton of the structural genes for CT,

- 11 -

TCP and other virulence factors (Fig.1)47. ToxRS and TcpPH are a pair of two interacting transmembrane proteins that bind to the toxT gene. Apart from activating toxT, ToxRS positively regulates CT and the outer membrane porin OmpU and negatively regulates expression of the porin OmpT48. ToxT is the major transcriptional activator and it directly regulates CT and other virulence factors. ToxT also amplifies its own expression47 49. The expression of the TcpPH operon is also dependent on the LysR family transcriptional regulator AphB and its partner AphA. The aphA promoter activity is repressed by HapR, a quorum-sensing regulated transcription factor50. It has been shown that mutation in either the crp gene, which encodes the cylic AMP receptor protein, or the cya gene, which is responsible for cAMP synthesis, results in expression of CT and TCP under conditions non-permissive for their production51. cAMP-CRP was found to bind in the tcpPH promoter at a site nearly identi- cal to the AphA-binding site, and therefore negatively influences the expression of tcpPH by competing with AphA for access to the promoter52. Inactivation of the crp gene caused a pleiotropic affect on the production of cholerae autoinducer 1 (CAI-1), multiple HapR-regulated genes, motility, the alternative sigma factor RpoE, outer-membrane proteins and many other genes53.

Quorum sensing

Quorum sensing is a cell-cell communication process in which bacteria monitor their population density by the production and detection of extracellular chemicals called autoinducers. As the bacterial population grows, the concentration of the autoinducers passes a threshold that leads to an alteration in gene expression54. This process allows the bacteria to communi- cate with one another to coordinate the behaviour of the entire community. Most pathogens use quorum sensing to induce the expression of virulence factors and biofilm-formation genes at high cell density. In V. cholerae, accumulation of the autoinducers AI-2 and CAI-1 activates the expression of the master quorum- sensing regulator, HapR (Fig.1). HapR represses genes important for biofilm-formation, and also attenuates virulence via repression of the transcriptional regulator, AphA, which is necessary for optimal expression of CT and TCP50, 55. HapR activates genes encoding several secreted metalloproteases and the

- 12 -

haemolysin co-regulated protein (Hcp)56. At low cell density, the autoinducer levels are low; the LuxPQ and CqsS receptors are autophosphorylated. The phosphate is transferred to LuxU, a phosphor-relay protein that phosphorylates the response regulator, LuxO. LuxO is active when it is phosphorylated. Acti- vated LuxO together with the transcription factor σ54 induces the expression of four redundant, small regulatory RNAs (Qrr1- 4) that repress expression of HapR. At high cell density the autoinducer levels are high, CqsS and LuxPQ act as phosphata- ses. As a consequence, luxO is dephosphorylated, and the Qrr1- 4 sRNAs are not expressed and HapR is therefore produced57, 58. The quorum sensing circuit of V. cholerae is illustrated in Fig. 1. Environmental signals

CAI-1 AI-2

P P R

LuxP R

p p

x x

c c

o o

T T

T T S

CqsS LuxQ S x

H x o

p o T

c T CqsA LuxS T LuxU

P tcpI tcpPH tcpA-F toxT acf ctxA-B σ54 + LuxO

AphB sRNAs + Hfq AphA ToxT

HapR VCC cAMP-Crp Biofilm Hcp, Proteases (e.g.HA/P, PrtV, LapX, Lap)

Figure 1. Model that depicting the virulence regulation in V. cholerae. Red arrows indicate phosphoryl-group transfer at low cell density in the quorum- sensing circuit of V. cholerae.

Protein secretion

Gram-negative bacteria have evolved numerous systems for protein export. Secretion requires translocation from the

- 13 -

cytoplasm across the outer as well as the inner membrane. In Gram-negative bacteria, there are now six known general classes of protein secretion systems each with considerable diversity.

General secretion and two-arginine (Tat) pathways

The general secretion (Sec) pathway and the two- arginine, or Tat translocation pathway are responsible for export of proteins into the periplasm of Gram-negative bacteria. The machinery of the Sec pathway, found ubiquitously in all living organisms, recognises a hydrophobic N-terminal leader sequence on proteins destined for secretion, and translocates proteins in an unfolded state, using ATP hydrolysis and/or a proton gradient for energy59, 60. The machinery of the Tat secretion pathway recognises a motif rich in basic amino acid residues (S-R-R-x-F-L-K) in the N-terminal region of large co- factor-containing proteins and translocates the proteins in a folded state using only a proton gradient as an energy source61.

Type I secretion system (T1SS), ABC transporters

T1SS transports various molecules, from ions and drugs, to proteins of various sizes (20 - 900 kDa). Through T1SS the protein is translocated, without any periplasmic intermedi- ate, out of the cell through an ATP-binding cassette transporter (ABC transporter), which is bound to the membrane by membrane fusion proteins (MFB). There is also an outer membrane protein associated with this mechanism, which is thought to provide a direct channel between the ABC transporter and the extracellular space. One example of a T1SS apparatus is the TolC-HlyD-HlyB complex of E. coli that secretes the haemolytic toxin HlyA. TolC is an integral membrane protein on the outer membrane while HlyD (MFP) and HlyB (ABC) occupy the periplasmic space and inner membrane, respectively62 . The RTX (repeat-in-toxin) of V. cholerae belongs to same family as HlyA, however the RTX toxin is secreted by an atypical T1SS that is composed of four components: RtxB, RtxD, TolC, and a second ATPase RtxE63.

- 14 -

Type II secretion system (T2SS), secreton-dependent pathway

T2SS involves a mechanism for exporting proteins across the two-membrane envelope in two distinct steps. Either the Sec or the Tat pathway are utilised for translocation across the inner membrane64. After cleavage of the signal peptides, the translocated protein might be processed and then transported from the periplasm to the exterior by either the main terminal branch or one of several protein-specific export systems for translocation across the outer membrane. Gram-negative type IV pili use a modified version of the T2SS for their biogenesis, and in some cases certain proteins are shared between a pilus complex and T2SS within a single bacterial species65. In V. cholerae, the T2SS is composed of 12 Eps (extracellular protein secretion) proteins and PilD66. Together, these proteins mediate the translocation of several virulence factors such as CT, HA/P, as well as chitinase, other proteases and the filamentous CTXΦ66-70.

Type III secretion system (T3SS), the contact dependent pathway

The T3SS is homologous to bacterial flagellar basal body. It is like a molecular syringe through which a bacterium (e.g. Yersinia, Salmonella, Shigella, Vibrio) can inject proteins into eukaryotic cells or the extracellular milieu. A central component of T3SSs is the needle complex, which mediates the passage of proteins through the multi-membrane envelope71, 72. The needle complex is composed of a series of basal rings, which span the inner and outer bacterial membranes, joined together by a rod and terminated with a needle-like structure that pro- trudes several nanometers from the bacterial surface. A conserved ATPase associates with the bacterial cytoplasmic base of the injectisome and energises transport. Most of the T3SS effectors have been shown to have a secretion signal within the N- terminal 20-30 amino acids that the needle complex is able to recognise. Export of the effector proteins is then triggered by contact with a eukaryotic target cell. Once the export is triggered, the T3SS effectors enter the needle complex at the base and make their way inside the needle towards the host cell. The

- 15 -

translocators are secreted first and produce a pore or a channel in the host cell membrane, through which other effectors may enter. T3SS effectors can manipulate host cells in several ways. Yersinia spp. have a dual strategy that is both antiphagocytic, owing to intracellular injection of YopE, YopH and YopT which inactivate the actin cytoskeleton, and anti-inflammatory, owing to the block in TNF production mediated by YopP73. Other pathogens, such as Shigella and Salmonella use the T3SS to reorganise the actin cytoskeleton and to induce its internalisa- tion into epithelial cells74.

After genomic sequence analysis of a non-O1 non-O139 strain (AM-19226) that appeared particularly pathogenic in experimental animals, it was suggested that this strain carries a T3SS that is related to the T3SS2 gene cluster found in a pandemic clone of Vibrio parahaemolyticus. The genes for this V. cholerae T3SS system appear to be present in many (15 out of 40) clinical and environmental non-O1, non-O139 strains, including at least one clone that is globally distributed72. The V. cholerae T3SS gene cluster was found to encode a protein (VopF) of 530 amino acids, containing one proline-rich sequence that resembles a formin homology 1 (FH1) domain and three WASP homology 2 (WH2) domains. These domains have been associated with eukaryotic actin-binding proteins72, 75. Translocation of VopF by V. cholerae altered the actin cy- toskeletal organisation of the eukaryotic host cells and was also needed for efficient mouse colonisation75.

Type IV secretion system (T4SS)

T4SSs are multiprotein complexes which transport macromolecules (proteins, DNA or DNA-protein complexes). These systems assemble as translocation channels, and often also as a surface filament or protein adhesin, at the envelopes of Gram-negative and Gram-positive bacteria. The T4SSs can be classified into four different classes according to a specific function that contributes to pathogenesis: (i) delivery of proteins or substrates to eukaryotic target cells, (ii) conjugation of chromosomal and plasmid DNA, (iii) DNA uptake and transformation, and (iv) DNA release into the extracellular milieu 76. Many organisms have homologous type IV secretion systems, including

- 16 -

the pathogens Agrobacterium tumefaciens (VirB/D4), Le- gionella pneumophila (Dot/icm), Helicobacter pylori (CAG; ComB), Pseudomonas aeruginosa (TraS/TraB), Bordetella per- tussis (Ptl) and E. coli (Tra). The best studied T4SS is the proto- typic A. tumefaciens VirB/D4 system. The VirB/D4 system is capable of delivering oncogenic T-DNA into plant cells and then cause cancerous growth or formation of crown-gall tu- mors77. The H. pylori CAG and ComB systems contain five and eight VirB/D4 homologues respectively, and the L. pneumo- philas Dot/icm system comprises only two obvious VirB/D4 homologues78. The T4SS of V. cholerae is not well-studied yet, but it is known that the non-O1 non-O139 strain V52 does not harbour any recognisable T4SS79.

Type V secretion system, autotransporters and two- partner secretion

Type V secretion systems (T5SSs) are used by many Gram-negative bacteria for the secretion of a large number of proteins. This family of secreted proteins includes those secreted via: autotransporters (type Va), together with the two- partner secretion system (type Vb), and the Oca family (type Vc)80. Proteins secreted via the T5SSs share some common characteristics including: (i) an N-terminal signal sequence for inner membrane export via the sec machinery, (ii) a functional passenger domain that can be surface exposed or released into the extracellular milieu, (iii) a linker region necessary for translocation through the outer membrane, and (iv) a C-terminal region involved in the formation of a transmembrane pore. In contrast to the type Va pathway, where the protein is produced as a single polypeptide, in the Vb system the passenger domain and the pore-forming β-barrel domains are translated as two separate proteins. The Oca family is somewhat a hybrid between the type Va and Vb systems81. A very large number of proteins are secreted via the T5SS. Most of the characterised proteins contribute to the virulence of animal or human pathogens including: proteases such as IgA proteases of Neisseria gonorrhoeae and Haemophilus influenza82, 83, toxins such as VacA of Helicobacter pylori84, SepA of Shigella flexneri85, ad- hesins such as Ag43 of E. coli86, Hia of Haemophilus influen- zae87, YadA of Yersinia enteroliticolitica81, UspA1 and UspA2 of

- 17 -

Moraxella catarrhalis88 and cytolysins such as ShlA/ShlB of Serratia marcescens89. The T5SS in V. cholerae is not well- studied to date.

Chaperone/ usher (CU) pathway

The CU pathway is implicated in the assembly and secretion of a superfamily of virulence-associated surface structures in Gram-negative bacteria90. The CU secretion machinery consists of only two components: a periplasmic chaperone that works in concert with an integral outer membrane protein, termed the usher. The prototypical structures assembled through the CU pathway are the adhesive type 1 and P pili of uropathogenic E. coli, which are composite structures, built from multiple subunit proteins. The components of the CU pathway are synthesised with typical N-terminal signal sequences for export across the inner membrane via the Sec system. In the periplasm, the subunit forms a stable complex with the chaperone, allowing cleavage of the signal peptide and fold- ing of the substrate. The chaperon-subunit is then targeted to the large, dimeric pore-forming usher in the outer membrane. Interaction with the usher releases the chaperone and opens the usher channel so the subunit can be exported. Repetition of the whole process allows extension of the pilus90.

Type VI secretion system (T6SS)

The T6SS was discovered by using the amoeba Dictyos- telium discoideum as a model host together with transposon mutagenesis of a non-O1 non-O139 strain of V. cholerae. This particular strain V52, belongs to the O37 serogroup that caused a large outbreak of cholera-like illness in Sudan 1968, resulting in 125 deaths. In contrast to the V. cholerae O1 El Tor strain N16961, the V52 strain shows cytotoxicity against Dictyos- telium79. The virulence displayed against amoebae by the V52 strain was found to require a gene cluster previously identified as highly conserved in several Gram-negative bacteria91. Be- cause IcmF, a known Legionella pnuemophila T4SS component, was encoded within this cluster, it was initially called IAHP (IcmF associated homologous proteins). The V. cholerae IAHP cluster was found to encode a secretion apparatus system

- 18 -

distinct from T4SS, so it was named VAS (virulence associated secretion). It was concluded that vas genes encode the “proto- typic type VI secretion system” (T6SS), and that the T6SS in a V. cholera V52 strain is responsible for the cytotoxicity to amoebae and the secretion of two proteins lacking an identifiable N- terminal signal peptide, Hcp (haemolysin-coregulated protein) and VgrG (valine-glycine repeats G)79. In silico analyses have proven the existence of one or two T6SS clusters containing 15 or more ORFs in more than 100 bacterial species92. The core components conserved among all of these T6SS are: the IcmF- and IcmH-like (or DotU-like) proteins, Hcp, VgrG, a lipoprotein and the ClpV AAA+ ATPase93. A model based on experimental evidence as well as similarities among some core components with the tail spike complex of T4 phage include: a cytoplasmic ATPase, transmembrane proteins, a hexameric translocation channel in the transmembrane region formed by Hcp and a trimeric needle-like structure on the bacterial cell surface that reaches the host cell and acts as a puncturing device formed by VgrGs 94. Overall, however, the identities and functions of T6SS effectors are still poorly understood.

Unlike the V. cholerae V52 strain, most of the O139 and O1 strains, like N16961, are permissive for Dictyostelium predation even if they possess genes encoding a T6SS. This suggested that they have a differentially regulated or non-functional T6SS79. Recently we found that the expression of one T6SS component, Hcp, in V. cholerae O1 strain A1552 was growth phase dependent and the maximum expression level was observed at OD 2,0. No Hcp was detected from an overnight cul- ture95. The expression of Hcp was positively regulated by cAMP-CRP, the alternative sigma factor RpoN, and the quorum sensing dependent regulator HapR. Several V. cholerae O1 strains, including N16961, possess a naturally occurring frame- shift mutation in hapR96. The lack of HapR might help to explain the inability of strain N16961 to survive amoeba predation even if Hcp is expressed via a heterologous promoter. On the other hand, in the non-O1/non-O139 strain V5/:04, deletion of hapR did not cause any decrease in expression or secretion of Hcp compared to the parental strain (paper IV). At this mo- ment we cannot rule out the possible existence of another T6SS regulatory pathway in strain V5/:04.

- 19 -

Vesicle-mediated secretion

In addition to the use of the multiprotein complex secretion systems listed above, Gram-negative bacteria possess another method for release of toxins or enzymes that is not observed in Gram-positive counterparts: the formation of outer membrane vesicles (OMVs)97, 98. OMVs are released to the surroundings through “constant bulging and pinching off” the outer membrane during bacterial growth. OMVs are released from bacteria in planktonic cultures, biofilm mode, liquid and solid media, as well as in natural environments. OMVs range in size from 50-250 nm and contain outer-membrane proteins, LPS, DNA, and phospholipids. As the vesicles are being released from the surface, they entrap some of the underlying perip- lasm97.

Inside

Outside

Figure 2. Localisation of bacterial toxin within and on the surface of OMVs

- 20 -

Vesicle trafficking may be involved in several cellular processes, including extracellular delivery of toxic compounds, DNA, phage proteins and periplasmic proteins, as well as producing decoys to help the pathogen evade the immune system 99. The most studied biological role of OMVs is their association with virulence factors. Several pathogenic bacteria produce OMVs that are capable of delivering virulence factors to eukaryotic targets, some of which are listed in Table 1.

Table 1. Virulence factors associated with OMVs

Species Virulence factor Reference Actinobacillus actinomy- Leukotoxin 101, 102 cetemcomitans Borrelia burgdorferi OspA, OspD 103 Campylobacter jejuni Cytolethal distending toxin 104 Escherichia coli ClyA, LT, CNF, HlyA, Shiga toxin 100, 105-109 Helicobacter pylori VacA 110 Salmonella typhi ClyA 100 Moraxella catarrhalis UspA1/A2 111 Pseudomonas Peptidoglycan hydrolase, phospholipase C, 98, 112, 113 aeruginosa haemolysin, Cif, alkaline phosphatase Shigella dysenteriae Shiga toxin 114 Shigella flexneri IpaB, IpaC, IpaD 115 Treponema denticola Proteases dentisilin 116, 117 Vibrio cholerae PrtV protease Paper III Xenorhabdus nematophi- Chitinase, bacteriocin, adhesin, pore- 118 lus forming toxin

Some of these virulence factors are enriched in the OMVs, suggesting a sorting mechanism for packaging specific factors into OMVs98. One finding that supports that theory is the activation of the pore-forming cytotoxin, ClyA, within secreted E. coli OMVs100. In the periplasm, the cystein residues in ClyA are oxidised by the Dsb pathway, renderings ClyA unable to oligomerise and form active oligomers. In contrast to the periplasmic β-lactamase, DsbA could not be detected in OMVs suggesting, that the vesicle lumen has a different redox state compared with the periplasm. As a consequence of the reducing environment in the vesicles, ClyA is reduced and can oligomer-

- 21 -

ise into pore-forming assemblies. The toxicity of ClyA in OMVs was higher than that of a preparation purified from the perip- lasm100.

It was recently demonstrated that OMVs from P. aeruginosa, E. coli and Porphyromonas gingivalis enter host cells by fusion with the lipid raft machinery119-121. OMVs from P. aeruginosa containing multiple virulence factors, including β- lactamase, alkaline phosphatase, haemolytic phospholipase C, and the CFTR inhibitory factor (Cif) enter the cytoplasm of the host cell via N-WASP-mediated actin trafficking. Virulence factors could then simultaneously be distributed to specific subcellular locations that may affect the host on multiple levels119. Moreover, OMV-mediated, long- distance delivery of virulence factors might help to explain certain observations, such as bacterial colonisation of catheters and dental pockets might contribute to systemic conditions such as cardiovascular diseases122. The OMVs are undoubtedly a vehicle for delivering virulence factors to eukaryotic targets, but they have several other roles that contribute to bacterial fitness. Since the OMVs are so ubiq- uitous and there is a high metabolic cost to produce them, it is likely that they have some fundamental physiological role. One such role might be the release of OMVs as a consequence of various stress factors exerted on the bacterium. Recently, McBroom and Kuehn suggested that production of bacterial OMVs is a fully independent, general envelope stress re- sponse123. In enteric bacteria the alternative sigma factor σE controlled one of the key pathways involved in the maintenance of cell envelope integrity124. Mutants in the σE pathway that accumulate misfolded envelope proteins showed higher vesicu- lation and survivial when exposed to rising temperatures and toxic compounds, such as polymyxinB and ethanol123. In V. cholerae, we observed that the alternative sigma factor σE controlled the expression of VrrA, a non-coding RNA (sRNA)125. VrrA was found to downregulate the peptidoglycan-binding OmpA protein level, which in turn led to increased vesicle production and thus reduced envelope stress.

Biofilm formation is recognised in providing bacteria protection from various environmental stresses, such as UV exposure, metal toxicity, acid exposure, dehydration and salinity,

- 22 -

phagocytosis and antibiotics126. One recent finding is that OMVs are a major component of the extracellular matrix of biofilms, and bacteria growing in biofilm produce more OMVs than planktonic bacteria127. Further characterisation of planktonic and biofilm-derived OMVs revealed that biofilm-derived OMVs were more proteolytic compared to planktonic OMVs, suggesting that different proteins are packaged into vesicles depending on growth mode127. It is likely that some of the protection against environmental stress that bacteria gain when growing in a biofilm is due to the presence of OMVs in the biofilm matrix, e.g., P. aeruginosa secretes β-lactamase-containing OMVs. These vesicles are permeable to β-lactams, which enter the OMVs and are inactivated by the β-lactamase128. In this way, the whole population of bacteria can survive in the presence of antibiotic even if just a subset of the population carries the antibiotic-resistance genes. Furthermore the OMVs have been shown to bind gentamicin and chorhexidine127, 129, 130 and probably also other toxic compounds that can bind to the bacterial surface and might be harmful for the bacteria. This means that the OMVs, both within and dispersed from the biofilm could act as decoys or sponges to reduce exposure to harmful agents before they affect the sessile bacteria in the biofilm.

OMVs represent the composition of the OM and therefore contain some of the most immunogenic and important antigens. They are therefore attractive candidates for vaccine development. OMV-based vaccines have been administered to control epidemics of N. meningitidis in Norway and New Zea- land131, 132. Immunisation with OMVs from P. gingivalis, S. enterica, and H. pylori has been shown to induce a stable immune respons and/or successful protection in mice132-135. OMVs released by V. cholerae have recently been added to the list of potential vaccine candidates. Mice immunised with OMVs by the intragastric, intranasal or intraperitoneal route induced a specific, long-lasting immune response against a variety of antigens present on the OMVs. The offspring of the immunised female mice were protected when challenged with V. cholerae at least three months after the last immunisation independent of immunisation route136, 137. V. cholerae OMVs was also able to in- corporate heterologous expressed periplasmic alkaline phosphatase (PhoA). Intranasal immunisation with PhoA-containing

- 23 -

OMVs induced a specific immune response in mice, which is a promising step towards designing an OMV-based vaccine137.

Persistence and survival of V. cholerae in the environment

Biofilm formation

The natural habitat for V. cholerae is the aquatic environment, and accidentally humans are infected through contaminated food or water. In the aquatic environment, bacteria encounter several environmental stresses, such as change in salinity, UV-stress, nutrient limitation, temperature fluctuations, viral infections and protozoan predation. Bacteria in aquatic environments are rarely found in the free-swimming or planktonic phase. Rather, they may form biofilms on living or non-living surfaces. Biofilm formation takes place in a sequence of steps: (i) an initial, reversible attachment of planktonic bacteria to a surface or by migration of sessile bacteria to cover an empty region of a surface, (ii) formation of microcolonies, production of exopolysaccharide (EPS) to adhere irreversibly to the substrate, (iii) the final differentiation of microcolonies, forming an EPS-encased, mature, thick and structured biofilm138.Within the biofilm the bacteria can access trapped and adsorbed nutrients, engage in favourable metabolic and genetic transactions with other members of the biofilm, and be protected from predators and toxic compounds, such as antimicrobial agents and natural oxidants139. Vibrios are an important component of marine biofilms and many studies have demonstrated the attachment of V. cholerae to environmental surfaces such as aquatic plants, zooplankton, filamentous green algae and crustaceans, such as copepods140-142. It is recognised that the copepod species of the zooplankton play a major role in the multiplication, survival and transmission of cholerae143, 144. One single copepod can carry as many as 103-105 V. cholerae cells. By binding to chitin-surfaces on zooplankton, they can acquire acid tolerance, thus getting protection from the gastric acid during stomach transit and reduce the infectious dose142. Based on that knowledge, a simple filtration procedure was developed. Colwell et al found the optimum filtering effect was attained

- 24 -

when old sari fabric was folded so that the water passed through four layers of cloth. Using that method, zooplankton, most phytoplankton, and particulates >20 μm were removed from water before use. That method yielded a 48% reduction in cholera compared with the control145.

Interactions of V. cholerae with chitin

Chitin (C8H13O5N)n is a naturally occurring, insoluble polymer composed of β-1,4-linked N-acetylglucosamine (GlcNAc) residues. Chitin is found in many places throughout the world. It is the main component of the cell walls of fungi, exoskeletons of arthropods such as crustaceans and insects and eggs from arthropods and nematodes146, 147. Interactions with chitin have likely contributed to the fitness of V. cholerae both outside and inside the human host. The ability to bind to chitin provides the bacterium with a surface for biofilm formation where they are protected from the environment. By degrading chitin by chitinases, the bacterium also acquires a continuous source of carbon, nitrogen and energy. Most vibrios do not colonise humans. It has been proposed that some virulence factors used by pathogens in the human host are derived from their role in their natural habitat. One example is TCP that is required for intestinal colonisation and cholera toxin acquisition. It was reported that TCP was needed to make a differentiated biofilm on the chitinaceous surfaces. Biofilms formed by TCP also concentrate a greater number of cells on the chitinaceous surface, which synergistically acts to secrete chitinolytic enzymes to degrade the chitin148. Another example is the secreted protein N-acetyl-D-glucosamine (GlcNAc)-binding protein A (GbpA). GbpA was identified as an attachment factor with the ability to bind both chitin surfaces and the mucus in the intestine149. A recent study showed that the level of GbpA was regulated by quorum sensing. At high cell density GbpA was degraded by the action of HA/P and PrtV proteases. This process could aid in the release of the bacteria back into the aquatic environment following the late stages of colonisation38.

Microarray expression profiling showed that chitin induced the expression of a 41-gene regulon (ChiS) involved in chitin colonisation, digestion, transport, and assimilation, in-

- 25 -

cluding genes predicted to encode a type IV pilus complex150. Further studies demonstrated that chitin could mediate the acquisition of genes conferring antibiotic resistance during growth of a V. cholerae strain on a crab shell fragment immersed in seawater. The transformation competence was found to require three positive regulators, HapR, RpoS, and TfoX and the type IV pilus complex.151. Chitin has also been shown to mediate serogroup conversion. Serogroup conversion occurred in a single- step exchange of large fragments of DNA between different serogroups living on the same surfaces. Crossovers were localised to regions of homology common to other V. cholerae serogroups that flank serogroup-specific encoding sequences. It was suggested that this mechanism can render the bacterium a se- lective advantage in niches of the organism that are critical for its success and that chitin induced serogroup conversion was responsible for the appearance of the O139 Bengal serogroup in 199210.

Another example of the impact of the V. cholerae interaction with chitin is its association with Chironomid (non biting midges) egg masses152, 153. Although V. cholerae has mainly been found in association with the gelatinous egg masses, attachment to the adult midge also occurs. By using GFP-tagged V. cholerae it was shown that most bacteria attached to the midge chitin. The adult midges also lais egg masses containing GFP-tagged V. cholerae in the water of external vessels, indicating that adult, non-biting midges are possible wind-borne carriers of V. cholerae154.

Non-O1 non-o139 Vibrio cholerae

Serogroups other than O1 and O139 are collectively called non-O1 non-O139 Vibrio cholerae or non-toxigenic V. cholerae. Most of the non-O1 non-O139 strains lack the CTXΦ and the V. cholerae pathogenicity island region. Still, they can adhere to and colonise the intestinal mucosal epithelium surface, causing human diarrhoea. In contrast to O1 and O139 strains, these non cholera toxin producing non-O1 non-O139 strains are commonly associated with septicaemia in patients with underlying liver disorders, otitis media, and septic wound

- 26 -

infections155-163. V. cholerae can be found not only in estuarine environments in tropical and temperate environments but also in the Bothnian Bay (latidude 65) with an average water temperature of 5C and low salinity (2 to 4‰)164. When the water temperature increases to ~20C the bacteria start to grow and there is a risk of infection from swimming outdoors. In the warm summer of 2006 in Sweden there was an increase in reports of infections caused by non-O1 non-O139 V. cholerae after bathing in the Baltic Sea. Younger people mainly got ear infections while older people got severe wound infection with con- comitant septicaemia that was fatal in three out of seven cases165, 166. Although the serogroups of non-O1 non-O139 V. cholerae rarely cause outbreaks, they represent an emerging threat and are of increasing concern in both endemic and non- endemic areas167.

Virulence factors

The majority of non-O1 non-O139 clinical and environmental isolates lack not only the virulence genes encoding CT and TCP but also the zonula occludens toxin (Zot) and accessory cholera enterotoxin (Ace)167, 168. Most of the non-O1 non-O139 strains may produce or at least possess the genes encoding other secreted factors such as, VCC, the RTX and HA/P20, 169, 170. The heat-stable toxin, NAG-ST (encoded by the gene stn/sto) has been described as an important virulence factor for non-O1 V. cholerae, but is very rare among clinical isolates171, 172.

Serum resistance

There are many case reports of septicaemia caused by non-O1 non-O139 V. cholerae161. To be able to cause septicaemia, bacteria need to invade and enter the bloodstream. Once in the bloodstream they must resist or avoid host defence mechanisms, including opsonisation, phagocytosis, intracellular killing and terminal complement lysis of bacterial cell walls. An important virulence factor in this regard is the bacterial capsule, which through its ability to resist complement-mediated bactericidal activity and phagocytosis facilitates the development of septicaemia173-175. The role of capsule in the development of dis-

- 27 -

seminated disease has clearly been demonstrated in V. vulnificus, an organism related to V. cholerae, which also causes wound infections and septicaemia176, 177. Like V. cholerae O139, a majority of non-O1 strains are able to produce a polysaccharide capsule173, 178. Still there is only one case report of bacte- raemia caused by V. cholerae O139 that also has a capsule, suggesting that capsule is not the only factor contributing to bacte- raemia179. The mechanism of serum resistance of V. cholerae non-O1 non-O139 strains has not been studied in detail. In other non-cholerae Vibrio species, it is suggested that high- molecular-weight O-antigen has a dual role in the protection against the complement in humoural fluid from the ampioxous Branchiostoma belcheri. The resistance to complement- mediated killing may result from sterically hindering of the complement components from gaining access to the cell membrane and/or avoiding initiating complement activation180.

V. cholerae host interaction models

Mammalian models

Our current knowledge about the pathogenesis of V. cholerae could not have been attained without the use of human volunteers or different animal models. Human volunteer studies have played a crucial role in evaluating host immune responses and vaccine efficacy. Due to ethical constraints and other understandable reasons, the use of human volunteers is restricted. Since Koch discovered V. cholerae in 1883, a variety of species have been tested to study cholera. Although none of the animals could be naturally infected with V. cholerae, a few of these models have provided useful information relevant to the pathogenesis of V. cholerae. Adult rabbits have been used in cholera research through surgical modification of their intestinal tracts. Preparations involve tying off segments of the small intestine before administration of test material into the ligated loops as basic technique to study secretogenic response to CT181. A variant of this model is the RITARD (removable intestinal tie- adult diarrhoea) model, which uses a temporary slip knot tie of the small bowel that is removed two hours post-challenge. This model elicits a massive and usually fatal cholera-like diarrhoea

- 28 -

in adult rabbits within 1.5 days182. Other models that have been used to measure fluid accumulation inside the mouse intestine are the sealed adult mouse model, in which the ano-rectal canal is sealed with glue, and the infant mouse model183, 184.

Non-mammalian models

Recently, model hosts, including the amoeba D. di- coideum, the insect D. melanogaster , the nematode C. elegans and the fish Danio rerio have been used as complementary system alternatives to mammalian host models for the study of bacterial pathogenesis39, 79, 185, 186. The molecular and genetic methods developed for use in these simple, non-mammalian hosts combined with their well-studied cellular biology, provide investigators with a models for analysing complex host- pathogen interactions. Another aspect is that virulence factors used by the pathogen in the human host might have been origi- nally developed to survive attack by organisms occupying the same niche. V. cholerae has recently been added to the list of microrgansims that are capable of causing lethal infections to flies and nematodes. In our recent study, we established that C. elegans can be used as an alternative model for the analysis of V. cholerae interactions with environmental hosts. V. cholerae was found to cause lethal colonisation in the intestine of the nematode independent of CT. We found that the PrtV protein has a dual role in predator grazing and immune modulation in human epithelial cells39. It has also been demonstrated that oral ingestion of V. cholerae produced an intestinally-localised, lethal infection in D. melanogaster that was dependent on CT. Furthermore, flies that lacked homologues of known human targets of CT were protected against infection. Ingestion of CT alone is enough to cause diarrhoea in human and mammalian animal models. In contrast, CT alone is not sufficient to produce lethal infection in flies. However when CT is co-administered with CT- V. cholerae, death of the fly ensued, suggesting that additional virulence factors that are essential for intoxication of the fly may not be required for intoxication of mammals187.

- 29 -

C. elegans

C. elegans is a free-living, transparent nematode (roundworm), about 1 mm in length, which lives in temperate soil environments. Most individuals are hermaphrodites, with males comprising just around 0.05% of the population. The eggs are laid by the hermaphrodite. After hatching, they go through four larval stages (L1-L4). During each stage they shed the old cuticle and synthesise a new stage-specific one. The generation time is about three days and the brood size is around 300 progeny per adult. The average life span of the nematode is approximately 2–3 weeks. Their basic anatomy includes the collagenous cuticle, mouth, pharynx, intestine and gonads (Fig.3).

Figure 3. Anatomical features of a C. elegans adult hermaphrodite. Sifri et al 2005188

Innate immunity in C. elegans

Evasion

C. elegans is a free-living hermaphroditic nematode, which is normally found in the soil where it feeds on bacteria that develops on decaying vegetable matter. In the wild the nematode encounters different kind of physical, chemical, and biological dangers. The ability to recognise and move away from harmful agents towards more advantageous environments is therefore crucial for the worm in its natural niche. In fact, C. elegans has a highly developed chemosensory system which allows the worm to perceive and respond to different chemical and physical cues so that it can either move away from harmful agents or move towards nutritious substances. C. elegans has

- 30 -

around 30 chemosensory neurons that detect odours by means of G-protein-coupled chemoreceptors. The ability to detect odours makes it possible for the worm to distinguish between different food bacteria and escape from pathogenic bacteria189.

Physical barrier

C. elegans is enclosed within a collagenous exoskeleton called the cuticle, which is the first defence against any pathogen it encounters. The cuticle is highly impermeable and few bacteria can bind to or degrade the cuticle. Two exceptions are Y. pestis and Y. enterocolitica, which attaches to the cuticle and form biofilms which cover the mouth and thus prevent feed- ing190. Access beyond the cuticle can be gained through the mouth, anus, vulva, or sensory openings. Microbes that are in- gested have to pass the pharyngeal grinder, which is very efficient at disrupting microbes and therefore prevents the entry of bacteria into the intestine. In our studies, V. cholerae could pass through the pharyngeal grinder and successfully colonise the intestine (paper I, Fig. 2). If the bacteria have managed to survive the passage through the grinder to the intestine, rhythmic peristalsis results in defecation every 45 seconds, which serves to remove intact bacteria from the digestive tract. Therefore, it has been suggested that bacteria need to adhere to specific receptors in the intestine to establish a persistent infection185.

Effector molecules of C. elegans innate immunity

The innate immunity of C. elegans does not contain any equivalent of phagocytic cells or receptors required for the en- gulfment of the invading pathogen191. For that reason C. elegans needs to rely on its production of effector molecules that de- stroy and/or inhibit the growth of pathogen, or protect the nematode against their effects. The C. elegans genome contains multiple gene families for this role, including homologues with well-known roles in host defence192. There are 10 homologues of lysozymes and at least three of them are induced upon infection by S. marcescens. The C. elegans lysozymes are suggested to act together with amoebapore-like proteins to degrade peptidoglycan. The amoebapore-like proteins belong to the saposin family. The saposin family contains pore-forming peptides and lipid-

- 31 -

degrading enzymes implicated in the innate immunity of mammals. There are 20 genes encoding saposin-like proteins present in the C. elegans genome. A number of neuropeptide- like (nlp) genes encoding peptides with an YGGYG motif have been shown to have antibacterial activity. Some of them are induced in vitro by certain fungal pathogens or bacteria192. Six ABF peptides have been identified in C. elegans based on homology to the ASABF peptides (Ascaris suum antibacterial factor) ABF is a distinct class of antimicrobial peptides with homology to insect defensins. The ABF-1 and ABF-2 peptides are constitutively expressed in the pharynx of C. elegans and are suggested to protect the pharyngeal tissue against infection. Recombinant ABF-2 has broad activity against Gram-positive and Gram-negative bacteria and yeast193.

Signalling pathways regulating immune response in C. elegans

At least four signalling pathways have been shown to be involved in innate immunity in C. elegans. The best characterised is the MAPK signalling pathway, which regulates many, but not all, candidate microbial peptides and affects survival when the nematodes are exposed to many pathogenic spp.194, 195. The DAF-2/DAF16 pathway also called the insulin-signalling pathway, regulates antimicrobial genes, such as those coding for lysozyme (lys-7) and saposins (spp-1 and spp-12) and is crucial for regulation of aging. The DBL-1 pathway is homologous to the mammalian TGF-β signalling cascade and one target is a lysozyme (lys-8) gene that is upregulated by bacterial infection. In addition to its role in innate immunity, the DBL-1 pathway controls body size, hypodermal and intestinal polyploidy, and the structure of the male tail. The only Toll-like receptor in C. elegans (TOL-1) has been shown to mediate the recognition and avoidance of certain pathogens. Until recently, it was suggested that the C. elegans Toll signalling pathway was dysfunctional or did not play a significant role in immune defence194. However, it was recently demonstrated that TOL-1 is required to prevent S. enterica invasion to the pharynx. TOL-1 was also shown to be required for correct expression of the defensin ABF-2 and a heat shock protein, both of which are expressed in the pharynx196. Most of the pathways mentioned above are differentially regu-

- 32 -

lated by certain pathogens or certain groups of pathogens. Some of these pathways seem to function independently, in parallel or together, to regulate various developmental, behavioural processes and general stress responses.

C. elegans as a host model for bacterial infections

In the laboratory, C. elegans is normally cultivated on a solid agar medium (NGM) in Petri dishes, feeding on a lawn of E. coli. In infection experiments, the E. coli lawn is displaced by the pathogen of interest and L4 larvae are put on the lawn to initiate infection. The pathogenic effects can be divided into two broad categories. The “fast killing” process is mediated by dif- fusible toxins and exposed worms die within hours without any direct exposure of live pathogens to the nematode. The “slow killing” is a consequence of infection, and the presence of live microbes within the nematode or attached to the cuticle is cor- related with disease. Infected worms die over a number of days rather than hours. The transparency of the nematode allows direct observations of the infection process in C. elegans by dif- ferential contrast microscopy or fluroescence microscopy in case the pathogen expresses GFP protein. In many cases, there is a visible distension of the intestine, but whether this is due to a physiological response of the nematode or simply to physical pressure exerted by the growing pathogens is currently not known.

While much can be learned about bacterial host interactions using C. elegans as an invertebrate host model, the model has its strengths and weaknesses. One drawback of using the nematode as a model to study bacterial physiology and pathogenesis is the fact that C. elegans is grown at temperatures below 25C, whereas some virulence genes involved in mammalian pathogenesis are only expressed at 37C. Furthermore, the immune responses to infection are relatively dissimilar between invertebrates and mammals. C. elegans does not possess macrophage-like cells, thus limiting the utility of C. elegans as a model for the study of intracellular pathogenesis191. Neverthe- less, no animal model completely simulates all features of human disease. C. elegans is a convenient animal model since the

- 33 -

genome is fully sequenced and different phenotypes and mutants are easily available. The strains can be kept as frozen stocks for long periods of time. In addition, the size and short generation time of C. elegans enables studies that require a large number of animals, which are unfeasible with mammals due to ethical constraints, cost and space requirements. An increasing number of bacterial pathogens have been shown to kill or injure the worm. In some cases, known virulence factors in mammalian models are also required for maximum virulence in C. elegans e.g., genera of Pseudomonas, Burkholderia, Yersinia, Salmonella, Serratia, Staphylococcus and Enterococ- cus185. In a recent study, a high-throughput screening method was developed, in which C. elegans killing was assessed in a semi-automated liquid assay. Screening a transposon library of more than 2,000 P. aeruginosa mutants identified 12 highly attenuated mutants. One mutant was affected in the cheB2 gene, which encodes a component of a chemotaxis pathway. The cheB2 mutant was further characterised and was found to be severely impaired for virulence in a mouse model197. It is possible that the C. elegans killing assay can be used as a high- throughput screen to identify novel types of antimicrobial compounds that repress virulence functions of a pathogen without affecting the growth of the pathogen or the nematode.

- 34 -

Aims of the thesis

To explore whether the nematode C. elegans can be used as an invertebrate infection model in the study of V. cholerae predator interactions To identify bacterial factor(s) important for the survival against predation by invertebrates To characterise the PrtV protein, which was found to be required for lethal infection of C. elegans and also important for survival against protozoan grazing To investigate PrtV in association with OMVs from LPS- defective V. cholerae To evaluate the presence of virulence-associated factors in Vibrio cholerae non-O1 non-O139 isolates from clinical and environmental sources

- 35 -

Results and discussion

Paper I

A Vibrio cholerae protease needed for killing of Caenorhabditis elegans has a role in protection from natural predator grazing

Various mammalian animal species have been used to study cholera and none of them have been found to be naturally infected by V. cholerae. It has therefore been stated that humans and primates are the only natural hosts for V. cholerae. We consider it presumptive to believe that primates are the only natural hosts for a marine organism such as V. cholerae. It is likely that the virulence mechanisms of V. cholerae have evolved through interactions with environmental factors. To fully understand the virulence potential of V. cholerae, it is therefore important to study the bacterium’s interaction with invertebrates. The C. elegans infection model has provided in- sights into virulence mechanisms of several human pathogens 185. We envision that this model could be helpful in identifying V. cholerae factors that may contribute to survival and persistence in the environment and therefore be of importance in pathogenesis.

We tested whether different V. cholerae strains could infect C. elegans. Lifespan assays revealed that regardless of serogroup, some clinical and environmental V. cholerae isolates could cause lethal colonisation of the pharynx and intestine of the nematode. We found that the killing efficiency was dependent on quorum sensing. The nematode survived only four days in the presence of a luxO mutant compared to a median survival of 15 days when feeding on a hapR mutant or the non- pathogenic E. coli OP50. A ctx mutant was fully capable of shortening C. elegans lifespan, showing that cholera toxin is not responsible for V. cholerae-mediated worm killing. The major extracellular proteases of a luxO mutant are HA/P, leucine aminopeptidase (Lap), a leucine aminopeptidase-related protein (LapX), and PrtV. By making in-frame deletions of the genes encoding the secreted proteases, we identified PrtV as the

- 36 -

only protease responsible for V. cholerae-mediated worm killing (Fig. 4).

Figure 4 Killing of C. elegans by V. cholerae protease-deficient mutants

The mechanism(s) by which PrtV mediates worm killing is currently not known. The PrtV protein belongs to a M6 family of peptidases and has 37% identity with InhA from B. thuringiensis, which has been shown to have activity against antibacterial factors found in the insect haemolymph. Other homologues to the PrtV protein are found in various environmental bacterial species including other Vibrios, Bacillus, Clostridium, Geobacillus, Aeromonas, Shewanella, and several other marine bacteria. We therefore hypothesised that PrtV might protect V. cholerae from antibacterial effectors produced by host/predators in the aquatic environment or degrade some other physiologically important substrates. To test this hypothesis, we co-cultured protozoan grazing predators with V. cholerae strains and their derivatives harbouring deletions in the genes encoding PrtV, LuxO or HapR. In agreement with our hypothesis, we found that PrtV and the quorum sensing cascade were necessary for survival from grazing by the flagellate Cafe- teria roenbergensis and the ciliate Tetrahymena pyriformis. Moreover, we found that the human T84 cell line produced more IL-8 when exposed to supernatant lacking PrtV compared to supernatant from the parental strain. The inflammatory re-

- 37 -

sponse in T84 cells was recently found to be mediated by VCC and was found to be modulated the PrtV protease31. The haemolytic and immuno-modulating properties of the PrtV mutant were dependent on the growth media. When grown in LB, the supernatant from the PrtV mutant was more haemolytic and caused a stronger inflammatory response compared to wild type. On the other hand, no differences between the PrtV mutant and wild type in haemolytic activity or inflammatory response could be observed using supernatants from BHI broth. We suggested that the content(s) of the BHI media might inhibit the activity or the expression of proteases from V. cholerae.

In this work, we established that C. elegans and protozoan like C. roenbergensis and T. pyriformis can be used as host models to analyse factor(s) contributing to the environmental fitness of V. cholerae. Due to the ecological importance of grazing predators, resistance against predation must have a great impact on the persistence and spreading of a pathogen such as V. cholerae. The PrtV protease was identified as a factor required for lethal infection of C. elegans and for survival against protozoan predation.

- 38 -

Paper II

The metalloprotease PrtV from Vibrio cholerae

To further characterise the PrtV protein, we cloned and overexpressed the PrtV protease in a V. cholerae strain deficient in other abundantly secreted proteases including HA/P, Lap and LapX. The strain was grown under optimised conditions and the secreted PrtV protein was purified from the culture supernatant. PrtV is synthesised from a 918-bp ORF and under- goes several cleavage steps to form the active proteolytic peptides of 81 kDa and 73 kDa. Both forms lack the first 105 N- terminal amino acids of the full-length protein and the 73-kDa form lacks one out of two polycystic kidney disease (PKD) domains (Fig.5). One or more PKD domains are found in several other bacterial proteins including chitinases, collagenases, proteases and other secreted or surface-exposed proteins (http://smart.embl-heidelberg.de/). The function of the PKD domains has therefore been suggested to be involved in protein- protein interaction or protein-carbohydrate inter-actions. The 81-kDa and 73-kDa forms of PrtV are relatively stable at low temperature and in the presence of Ca2+, but further degradation by autoproteolysis occurs, resulting in an approximately 55-kDa product, which after heat denaturing separates into two migrating bands of 18 kDa and 37 kDa on SDS-PAGE.

HEYGHDLGLPD M6.002 peptidase PKD PKD 1 918 81 kDa 106 834 73 kDa 106 763 37 kDa 18 kDa 106 434 587 749

Figure 5. Schematic representation of the identified PrtV forms.

The 55-kDa form of PrtV, which is proteolytically active and stable in the absence of Ca2+ was used for further charac-

- 39 -

terisation. PrtV activity was completely inhibited by metal ion chelators EDTA and 1,10-phenanthroline, confirming that PrtV is a metalloprotease. Proteolytic activity was enhanced by addition of the divalent metal ions Ca2+, Str2, and Mg2+. Monovalent metal ions, Li+, Na+, and K+ also increased the proteolytic activity, but higher concentrations were required for optimal activity compared to the divalent metal ions (Fig. 3, paper II). Purified PrtV induced cell rounding and detachment of HCT8 cells. By using human blood plasma, we could identify potential substrates for the protease. The preferred substrates were plasminogen, fibronectin, and fibrinogen (Fig.6, paper II). PrtV has the PKD domains, which are found in several collagenases and chitinases. Since the cuticle of C. elegans consists of collagen, we thought that interaction or degradation of the cuticle might be involved in the lethality caused by PrtV, however no activity against synthetic collagenase could be observed. Chitin is present in the C. elegans eggshells and pharynx as well as in some extracellular threads of some flagellates198, 199. Because of that, we considered that the PrtV somehow could be involved in binding and degradation of chitin and thus facilitate colonisation of the predators; however, a recent study showed that PrtV and HA/P participate in the targeted degradation of the chitin attachment factor GbpA38. The ability of V. cholerae to detach from chitin does not explain the lethal effect against C. elegans since a PrtV mutant still expresses HA/P, which also promotes detachment from chitin. So the question of the mechanism(s) by which PrtV prevents predator grazing is still open, and will be the subject of future studies.

- 40 -

Paper III

Role of LPS in vesicle-mediated export of Vibrio cholerae PrtV protease

Several observations in Gram-negative bacteria have revealed the presence of toxins and other virulence-associated factors in association with OMVs200. We have observed that V. cholerae produces at least six times more OMVs compared with the E. coli strain MC1061, but there are few reports of proteins specifically associated with V. cholerae OMVs. In our lab, we have analysed the protein content of OMVs derived from V. cholerae O1 El Tor strains A1552 and C6706. Immunoblot analysis using antisera against CT, Tcp, VCC, PrtV, OmpT and OmpU as well as N-terminal sequencing of proteins extracted from 2-D gels revealed only outer membrane proteins as being present in the OMVs. It has been shown that CT can associate with OMVs from E. coli and structures in the core LPS appear to be important for this association. Therefore we analysed the V. cholerae O1 El Tor strain P27459 and its derivative harbouring a chromosomal insertion in the waaL gene, which encodes an O-antigen ligase and is therefore lacking the O-antigen. There were no differences in growth rate or amount of OMVs produced between the waaL mutant and the parental strain. SDS- PAGE and immunoblot analysis as well as immunogold label- ling of OMVs showed that the PrtV protein was associated with OMVs from the waaL mutant but not to wild type OMVs. Im- muno-EM showed that PrtV localise at the surface of the OMVs and not the inside, indicating that PrtV associates the OMVs after it is secreted to the exterior (Fig. 4B, paper III). The PrtV- containing OMVs showed a cytotoxic effect against HeLa cells that was similar to that of the purified PrtV protein. Further- more, the OMVs from the waaL mutant were capable of digesting gelatine, showing that the OMV-associated PrtV is active. We have also observed that active PrtV is associated with OMVs from a wavB mutant (Fig.6). The wavB gene most likely encodes a ß-1,4-glucosyl transferase. It is predicted that LPS from the wavB mutants contains the O-antigen but lacks two sugar chains in the core LPS46. In this work, we have demonstrated that the LPS of V. cholerae mediates association of -biological

- 41 -

active PrtV with OMVs. But still there remain questions to be answered, e.g., why can’t we detect specific protein associated with OMVs from wild type O1 V. cholerae? Is it possible that during human infection proteins secreted from V. cholerae associate with LPS on the commensals e.g., E. coli, thus making them vehicles for the delivery of toxins and also targets for the immune defence? Is V. cholerae’s production of OMVs just a simple and rapid method to change the surface, when it encounters a challenging environment?

tV r 9 p 5 B  4 v B 7 a v 2 a w P w

Figure 6. Zymogram on an SDS-PAGE gel containing gelatin as a substrate for OMVs derived from P27459, wavB::pGP704 and wavB::pGP704prtV as indicated. The Coomassie stained SDS-PAGE below serves as a loading control.

- 42 -

Paper IV

Evaluation of the presence of virulence-associated factors in Vibrio cholerae non-O1 non-O139 isolates from clinical and environmental sources

Bacterial pathogens like V. cholerae continue to develop in nature, and the continued emergence of new environmentally prevalent pathogens is an indicator of the need for continued research towards the understanding of host defence and pathogen virulence. Therefore, we evaluated the presence of virulence-associated factors in non-O1 non-O139 V. cholerae isolated from patients or the environment. The presence of three major virulence genes, toxR, tcpA and ctxA, as well as expression of other putative virulence factors such as OmpU and haemolysin (HlyA) was compared among these strains. None of the strains carried the genes for ctxA and tcpA. All strains pos- sessed the gene encoding the regulatory protein ToxR. In addition, we have also tested the presence of Type III secretion system (T3SS) locus and the important components of Type VI secretion system (T6SS) genes in these V. cholerae strains. Poly- merase chain reaction analyses revealed the presence of T6SS secretion system genes in all tested non-O1 non-O139 strains. T3SS genes were only present in the clinical isolates. Initial PCR screening using primers for detection of ompU, which have been used in several studies revealed that ompU was only present in one isolate 201. Western blot analysis, however showed that OmpU was present in all strains, although the strain V:6/04 has a smaller molecular size OmpU protein. A BLAST search (http://www.ncbi.nlm.nih.gov/tools/cobalt/cobalt.cgi) at the protein level revealed that many of the up-to-date sequenced V. cholerae strains are missing some amino-acids in the regions where the primer was designed. By designing oli- gonucleotide primer pairs specific for more conserved regions of the OmpU protein, we were able to detect ompU in all strains by PCR. It was reported earlier that non-pathogenic or environmental V. cholerae strains express a truncated form of OmpT202. In agreement with this report, in our analysed non-O1 non-O139 strains, we find that OmpT was either not detectable or expressed in a truncated form. Furthermore, the molecular

- 43 -

size of the OmpW protein also seems to vary between the strains. Outer membrane proteins (OMPs) play important roles in bacterial pathogenesis by enhancing the adaptability of pathogens to various environments. Whether the differentially sized outer membrane proteins differ in their contribution to pathogenesis or the environmental fitness of V. cholerae is not known. We have found significant variability amongst non-O1 non-O139 isolates with regard to serum resistance: three strains are highly serum resistant as indicated in (Fig. 7, Fig. 2 paper IV). The ability of some pathogenic strains to resist complement killing clearly resides in their surface structure. Encapsulation could not explain the serum resistance of the non-O1 non-O139 V. cholerae, since some serum-sensitive strains also produce capsule (Fig. 3 paper IV). Interestingly, we found that the three clinical isolates, which had the highest serum resistance, all belong to the O5 serogroup. High-molecular-weight O-antigen is suggested to confer protection against complement in other non-cholerae Vibrio species180. However, two environmental strains also of the O5 serogroup were not as serum resistant as the clinical strains, suggesting that the O-antigen is not solely responsible for protection against complement (Fig. 7). It can not be excluded that there might be some differences in other parts of the LPS or that the strains exhibit some differences in OMP profiles, which affect the ability to resist complement- mediated killing. Future studies, including chemical analysis of purified LPS as well as more detailed studies of OMP profiles, might clarify the contribution of LPS and OMPs to serum resistance of the O5 strains.

We also used the C. elegans infection model to evaluate the ability of the non-O1 non-O139 isolates to avoid predator grazing. The strains that were most potent in resisting serum killing were also the most potent nematode killers. A hapR mutant was clearly attenuated in the worm killing assay, consistent with our previous data, which showed that the HapR- regulated PrtV protease was required for the killing of C. elegans (Paper I). HapR was not required for serum resistance, showing that the mechanism of complement protection is not quorum sensing dependent.

- 44 -

120

100

% Bacterial survival 20

2 5 0 5 1 1 0 0 9 3 R :4 69 06 3 3 3 1 9 - 8 1 1 1 V1 V5 p L:1 L L:5 6 0 0 3 3 L M -19 - g g V a .c .c 33 7 7 6 P A P G A A h V 1 1 A 93Ag13 3 D CSK n I M14 M18781M19425M M13396 N 9 9 Nag 5 . .n V K KI 1 KI 17036KI 17044 NAGV17NAG V B B

Fig.7 Serum killing assay with human serum. The results are expressed as percent survival in normal human serum compared to that in heat-inactivated serum. Bars with stripes indicate strains belonging to the O5 serogrup. Strains that could efficiently kill C. elegans are marked with

Taken together, our study shows that the non-O1 non- O139 V. cholerae isolates are very heterogeneous. There was significant variability amongst the isolates with regard to serum resistance, the ability to kill C. elegans, surface structure and protein composition as well as secreted protein profiles. HapR might be involved in pathogenesis of non-O1 non-O139 V. cholerae by regulating, directly or indirectly, the levels of proteases, haemolysin, and Hcp. To conclude any potential pathogenicity pattern, more isolates per infection type would be necessary. Nevertheless, in the environment V. cholerae continues to evolve and it is likely that most strains have the potential to cause human disease. Therefore, it is important to improve our understanding of V. cholerae, since no one can predict which pathogenic variants of V. cholerae will emerge in the future (cover page).

- 45 -

Conclusions

C. elegans and protozoan organisms like T. pyriformis and C. roenbergensis can be used as host models to analyse factor(s) contributing to the environmental fitness of V. cholerae. The quorum sensing regulated PrtV protease is a factor required for lethal infection of C. elegans and also required for survival against protozoan predation. The PrtV protein is a calcium-stabilised metalloprotease. Purified PrtV protein can degrade fibronectin, fibrinogen, and plasminogen and shows cytotoxicity against HCT8 cells. OMV-associated PrtV protein is biologically active and association of the PrtV with OMVs is depend- ant on LPS-structure. Some non-O1 non-O139 V. cholerae, which efficiently kill C. elegans are highly resistant against human serum. Encapsulation is not a main factor for serum resistance in V. cholerae non-O1 non-O139 strains. The quorum sensing regulator HapR might be involved in pathogenesis of non-O1 non-O139 V. cholerae by regulating, directly or indirectly, the levels of proteases, haemolysin, and Hcp.

- 46 -

Sammanfattning på svenska

Vibrio cholerae - den Gram-negativa bakterien som orskar kolera - uttrycker två kritiska virulensfaktorer: koleratoxinet (CT), som är den huvudskliga orsaken till diarreén och toxin-coregulated pilus (TCP), som är väsentligt för kolonisering. Stammar som uttrycker TCP men saknar CT- genen kan på effektivt sätt förvärva den genom lysogenisk konvertering med hjälp av CTX Φ, en fintrådig bakteriofag som innehåller CT-genen och använder TCP som receptor. V. cholerae är en naturlig invånare i flodmynningar och kustvatten i tempererade och tropiska områden över hela världen. I sin akvatiska livsmiljö måste V. cholerae kunna bemöta flera miljöbetingade stressfaktorer som förändringar i salthalt och pH, UV-stress, begränsningar av näringsämnen, temperatur- växlingar, virusinfektioner och protozoapredation. För att fullt ut kunna förstå virulenspotentialen hos V. cholerae, krävs kunskap om dess samverkan med inte bara människan utan även olika faktorer i miljön.

Med nematoden Caenorhabditis elegans som värdmodell kunde vi identifiera ett tidigare okaraktäriserat protein, det extracellulära proteaset PrtV. PrtV visar sig vara nödvändigt för att döda C. elegans och också nödvändigt för att överleva att bli uppäten av ciliaten Tetrahymena pyriformis och flagellaten Caféteria roenbergis.

PrtV-proteinet, som tillhör M6 metallopeptidas- familjen, renades fram för ytterligare karaktärisering och vi fann att proteinet var cytotoxisk mot den mänskliga cell-linjen HCT8. PrtV visar sig kunna degradera fibrinogen, fibronektin och plasminogen i blodplasma.

De flesta Gram-negativa bakterier släpper ifrån sig yttermembran-vesiklar (OMVs) genom att knoppa av små blåsor av yttermembranet. OMVs har visat sig innehålla många virulensfaktorer som är viktiga för bakteriell patogenes. Vi undersökte därför om PrtV kunde associera till OMVs och fann att så inte var fallet i vildtypsstammen. PrtV kan däremot, i biologiskt aktiv form, associera till OMVs från en LPS mutant.

- 47 -

V. cholerae stammar kan indelas i över 200 serogrupper. Endast serogrupperna O1 och 0139 sammankopplas med den allvarliga diarreésjukdomen kolera. De andra serogrupperna kallas kollektivt för icke-O1/icke-O139 V. cholerae. Icke-O1/icke-0139 V. cholerae kan orsaka gastroenterit och andra infektioner, men till skillnad från V. cholerae från O1 och O139 stammarna vet man inte så mycket om deras sjukdomsframkallande potential. Vi analyserade sju kliniska och två miljöisolat för att undersöka deras virulensegenskaper. Ingen av isolaten hade generna som kodar för CT eller TCP. Förmågan att motstå serum varierade betydligt mellan stammarna och är oberoende av inkapsling. De stammar som var mest serum-resistenta kunde också orsaka dödlig infektion hos C. elegans.

- 48 -

Acknowledgements

All good things must come to an end, and so my PhD studies. Many people have contributed in a variety of ways to the production of this thesis over the last years. Collaborators, co-authors, students, friends, former and present people that have worked in the Department of Molecular Biology, THANK YOU! In particular (at the risk of offending the others), I'd like to say a special thank to my supervisor Sun Nyunt Wai for her never ending enthusiasm, guidance, great support, and kind- ness throughout my PhD studies. She has also allowed me the freedom to follow my own research interests in the course of my PhD studies, which I greatly appreciate. Thanks Sun! Without you, I might have chosen pest instead of cholera. I also would like to thank my co-supervisor Bernt Eric Uhlin for construc- tive suggestions, for arranging meetings, workshops, and nice garden parties. Thanks also to my cholera companions: Kaloris, -you abandoned me for some Listeria guys, WHY? Tianyan, for de- licious gioza and for being a nice person, Pramod, for doing that, always with a smile on your face, Takahiko for keeping the lab in silence, Sridhar and Amit –see you in Kolkata, Mitesh, for helping me with word formatting and students, Tetsuya and Ryoma – korekaramo ganbatte kudasai, Maria for taking over after me?, Jurate for bringing Kaloris here, Monica, my lab-management guru, Annika, for nice book re- views, Stina, for always saying something funny, Connie, for always saying the truth, Ting for saying hello, Berit, for gar- dening tips, Hyun-Sook, for decoration tips, Claudia, Anna, and Carlos, the department feels so much emptier without you. Thanks also to past and present members of our Journal club, for our scientific Friday morning sessions. There are many other people at the department who have made my life a lot easier. Debbie as the head of the infras- ructure organisation, the secretaries, media, Jhonny Marek, Marinella, Gunilla, Siv, , THANK YOU! During my research I have also benefited from some much broader discussions in the lunch-room, and “bunkern/beer corner” about science, life, the universe and eve-

- 49 -

rything. These discussions have often been lubricated with plenty of coffee or beer, and fed with copious amounts of choco- lates and cakes. I am indebted to Herr Professor Sven Bergström and his Borrelia group, Marie alias Bindefeld for party arrangements, Ingela, -I need your formula for the Glögg à la Molbiol, Betty, for help with the English language, Ignaz, Sandra, Elin, Patrik, Pär, Korven, Jenny, Maria, Lelleman, Sara, Mikael, Erik, Olena, Mark, Hannah, Anders, Tomas, Login, Lena, Kemal, Linda, Vicky, Reine, Sofia, Jågren (holy Jesus of mother), Jonääs, for beer corner management, Edmund, for being one of a kind, Sakura, Teresa, and many others…for making the department such a fun place to work. I would not have been able to conduct this research without the financial assistance of grants from the Swedish Re- search Council, the Swedish Foundation for International Co- operation in Research and Higher Education (STINT), and the Faculty of Medicine, Umeå University. This work was in part performed within the Umeå Centre for Microbial Research (UCMR). Of course, there has been more than work these last years, and I would like to thank those who have kept me away from science life and helped to make my real life so enjoyable. Så ett stort tack till: Monica utan dig hade jag väl fortfarande skyfflat dynga på farmen i Dalkarlså, eller klostridier på NUS, Eva och Elisabet, jag saknar våra Morkulleblues-sessioner, Ann-Sofi, för bl.a. alla svampexcursioner och Mikael för att han är som han är, Annika för att du är en god vän och alla andra ”Bygdevänner”. Ett stor tack till David, för rese- och festarrangemang och skötsel av Zlabbarn med mera, Marie för goa efterrätter, Anders för lift när Zlabbarn strejkar, blotgänget, the leftovers incorporated, tjejgänget och alla ni andra som har förgyllt min fritid. Ett stort tack till mina föräldrar för att ni alltid finns där för mig och för att ni skapat mig till den jag är idag! David, Berit och Jonas med familjer, jag uppskattar verkligen erat stöd. Till slut vill jag tacka mina små guldklimpar Viktor, Rosita och Ludvig, utan er hjälp och tålamod hade det här aldrig gått. Jag älskar er!

- 50 -

References

1. Faruque, S. M., Albert, M. J. & Mekalanos, J. J. Epidemiology, ge- netics, and ecology of toxigenic Vibrio cholerae. Microbiol Mol Biol Rev 62, 1301-14 (1998). 2. Kaper, J. B., Morris, J. G., Jr. & Levine, M. M. Cholera. Clin Micro- biol Rev 8, 48-86 (1995). 3. Desjeux, J., Briend, A. & Butzner, J. Oral rehydration solution in the year 2000: pathophysiology, efficacy and effectiveness. Baillieres Clin Gastroenterol. 11, 509-527 (1997). 4. Sack, D. A., Sack, R. B., Nair, G. B. & Siddique, A. K. Cholera. Lancet 363, 223-33 (2004). 5. Bhattacharya, M. K. et al. Outbreak of Vibrio cholerae non-O1 in India and Bangladesh. Lancet 341, 1346-7 (1993). 6. Ramamurthy, T. et al. Emergence of novel strain of Vibrio cholerae with epidemic potential in southern and eastern India. Lancet 341, 703-4 (1993). 7. Comstock, L. E., Johnson, J. A., Michalski, J. M., Morris, J. G., Jr. & Kaper, J. B. Cloning and sequence of a region encoding a surface polysaccharide of Vibrio cholerae O139 and characterization of the insertion site in the chromosome of Vibrio cholerae O1. Mol Micro- biol 19, 815-26 (1996). 8. Waldor, M. K., Colwell, R. & Mekalanos, J. J. The Vibrio cholerae O139 serogroup antigen includes an O-antigen capsule and lipopolysaccharide virulence determinants. Proc Natl Acad Sci U S A 91, 11388-92 (1994). 9. Bik, E. M., Bunschoten, A. E., Willems, R. J., Chang, A. C. & Mooi, F. R. Genetic organization and functional analysis of the otn DNA essential for cell-wall polysaccharide synthesis in Vibrio cholerae O139. Mol Microbiol 20, 799-811 (1996). 10. Blokesch, M. & Schoolnik, G. K. Serogroup conversion of Vibrio cholerae in aquatic reservoirs. PLoS Pathog 3, e81 (2007). 11. Faruque, S. M. et al. The O139 serogroup of Vibrio cholerae comprises diverse clones of epidemic and nonepidemic strains derived from multiple V. cholerae O1 or non-O1 progenitors. J Infect Dis 182, 1161-8 (2000). 12. Waldor, M. K. & Mekalanos, J. J. Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272, 1910-4 (1996). 13. Choi, S., Dunams, D. & Jiang, S. C. Transfer of cholera toxin genes from O1 to non-O1/O139 strains by vibriophages from California coastal waters. J Appl Microbiol (2009). 14. Boyd, E. F. & Waldor, M. K. Alternative mechanism of cholera toxin acquisition by Vibrio cholerae: generalized transduction of CTXPhi by bacteriophage CP-T1. Infect Immun 67, 5898-905 (1999).

- 51 -

15. Heidelberg, J. F. et al. DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature 406, 477-83 (2000). 16. Kimsey, H. H. & Waldor, M. K. CTXphi immunity: application in the development of cholera vaccines. Proc Natl Acad Sci U S A 95, 7035- 9 (1998). 17. Kovach, M. E., Shaffer, M. D. & Peterson, K. M. A putative integrase gene defines the distal end of a large cluster of ToxR-regulated colonization genes in Vibrio cholerae. Microbiology 142 (Pt 8), 2165-74 (1996). 18. Sanchez, J. & Holmgren, J. Cholera toxin structure, gene regulation and pathophysiological and immunological aspects. Cell Mol Life Sci 65, 1347-60 (2008). 19. Reidl, J. & Klose, K. E. Vibrio cholerae and cholera: out of the water and into the host. FEMS Microbiol Rev 26, 125-139 (2002). 20. Kurazono, H. et al. Distribution of genes encoding cholera toxin, zonula occludens toxin, accessory cholera toxin, and El Tor hemolysin in Vibrio cholerae of diverse origins. Microb Pathog 18, 231-5 (1995). 21. Nagamune, K. et al. In vitro proteolytic processing and activation of the recombinant precursor of El Tor cytolysin/hemolysin (pro-HlyA) of Vibrio cholerae by soluble hemagglutinin/protease of V. cholerae, trypsin, and other proteases. Infect Immun 64, 4655-8 (1996). 22. Zitzer, A., Wassenaar, T. M., Walev, I. & Bhakdi, S. Potent membrane-permeabilizing and cytocidal action of Vibrio cholerae cytolysin on human intestinal cells. Infect Immun 65, 1293-8 (1997). 23. Harris, J. R. et al. Interaction of the Vibrio cholerae cytolysin (VCC) with cholesterol, some cholesterol esters, and cholesterol derivatives: a TEM study. J Struct Biol 139, 122-35 (2002). 24. Krasilnikov, O. V., Muratkhodjaev, J. N. & Zitzer, A. O. The mode of action of Vibrio cholerae cytolysin. The influences on both erythrocytes and planar lipid bilayers. Biochim Biophys Acta 1111, 7-16 (1992). 25. Yuldasheva, L. N. et al. Lumen geometry of ion channels formed by Vibrio cholerae EL Tor cytolysin elucidated by nonelectrolyte exclu- sion. Biochim Biophys Acta 1512, 53-63 (2001). 26. Moschioni, M. et al. The Vibrio cholerae haemolysin anion channel is required for cell vacuolation and death. Cell Microbiol 4, 397-409 (2002). 27. Coelho, A., Andrade, J. R., Vicente, A. C. & Dirita, V. J. Cytotoxic cell vacuolating activity from Vibrio cholerae hemolysin. Infect Immun 68, 1700-5 (2000). 28. Saka, H. A. et al. Vibrio cholerae cytolysin is essential for high enterotoxicity and apoptosis induction produced by a cholera toxin gene-negative V. cholerae non-O1, non-O139 strain. Microb Pathog 44, 118-28 (2008).

- 52 -

29. Ichinose, Y. et al. Enterotoxicity of El Tor-like hemolysin of non-O1 Vibrio cholerae. Infect Immun 55, 1090-3 (1987). 30. Debellis, L. et al. The Vibrio cholerae cytolysin promotes chloride secretion from intact human intestinal mucosa. PLoS One 4, e5074 (2009). 31. Ou, G. et al. Vibrio cholerae Cytolysin Causes an Inflammatory Re- sponse in Human Intestinal Epithelial Cells That Is Modulated by the PrtV Protease. PLoS One 4, e7806 (2009). 32. Booth, B. A., Boesman-Finkelstein, M. & Finkelstein, R. A. Vibrio cholerae hemagglutinin/protease nicks cholera enterotoxin. Infect Immun 45, 558-60 (1984). 33. Finkelstein, R. A., Boesman-Finkelstein, M., Chang, Y. & Hase, C. C. Vibrio cholerae hemagglutinin/protease, colonial variation, virulence, and detachment. Infect Immun 60, 472-478 (1992). 34. Hase, C. C. & Finkelstein, R. A. Cloning and nucleotide sequence of the Vibrio cholerae hemagglutinin/protease (HA/protease) gene and construction of an HA/protease-negative strain. J Bacteriol 173, 3311-7 (1991). 35. Wu, Z., Milton, D., Nybom, P., Sjo, A. & Magnusson, K. E. Vibrio cholerae hemagglutinin/protease (HA/protease) causes morpho- logical changes in cultured epithelial cells and perturbs their paracellular barrier function. Microb Pathog 21, 111-23 (1996). 36. Benitez, J. A. et al. Adherence of Vibrio cholerae to cultured differentiated human intestinal cells: an in vitro colonization model. In- fect Immun 65, 3474-7 (1997). 37. Olivier, V., Queen, J. & Satchell, K. J. Successful small intestine colonization of adult mice by Vibrio cholerae requires ketamine an- esthesia and accessory toxins. PLoS One 4, e7352 (2009). 38. Jude, B. A., Martinez, R. M., Skorupski, K. & Taylor, R. K. Levels of the Secreted Vibrio cholerae Attachment Factor, GbpA, are Modu- lated by Quorum Sensing Induced Proteolysis. J Bacteriol (2009). 39. Vaitkevicius, K. et al. A Vibrio cholerae protease needed for killing of Caenorhabditis elegans has a role in protection from natural predator grazing. Proc Natl Acad Sci U S A 103, 9280-5 (2006). 40. Ogierman, M. A. et al. Characterization of the Vibrio cholerae El Tor lipase operon lipAB and a protease gene downstream of the hly region. J Bacteriol 179, 7072-80 (1997). 41. Rawlings, N. D., Morton, F. R., Kok, C. Y., Kong, J. & Barrett, A. J. MEROPS: the peptidase database. Nucleic Acids Res 36, D320-5 (2008). 42. Edlund, T., Siden, I. & Boman, H. G. Evidence for two immune in- hibitors from Bacillus thuringiensis interfering with the humoral de- fense system of saturniid pupae. Infect Immun 14, 934-41 (1976). 43. Siden, I., Dalhammar, G., Telander, B., Boman, H. G. & Somerville, H. Virulence factors in Bacillus thuringiensis: purification and prop-

- 53 -

erties of a protein inhibitor of immunity in insects. J Gen Microbiol 114, 45-52 (1979). 44. Dalhammar, G. & Steiner, H. Characterization of inhibitor A, a protease from Bacillus thuringiensis which degrades attacins and cecropins, two classes of antibacterial proteins in insects. Eur J Bio- chem 139, 247-52 (1984). 45. Chatterjee, S. N. & Chaudhuri, K. Lipopolysaccharides of Vibrio cholerae. I. Physical and chemical characterization. Biochim Biophys Acta 1639, 65-79 (2003). 46. Nesper, J. et al. Comparative and genetic analyses of the putative Vibrio cholerae lipopolysaccharide core oligosaccharide biosynthesis (wav) gene cluster. Infect Immun 70, 2419-33 (2002). 47. DiRita, V. J., Parsot, C., Jander, G. & Mekalanos, J. J. Regulatory cascade controls virulence in Vibrio cholerae. Proc Natl Acad Sci U S A 88, 5403-7 (1991). 48. Provenzano, D. & Klose, K. E. Altered expression of the ToxR- regulated porins OmpU and OmpT diminishes Vibrio cholerae bile resistance, virulence factor expression, and intestinal colonization. Proc Natl Acad Sci U S A 97, 10220-4 (2000). 49. Miller, V. L. & Mekalanos, J. J. Synthesis of cholera toxin is positively regulated at the transcriptional level by toxR. Proc Natl Acad Sci U S A 81, 3471-5 (1984). 50. Kovacikova, G., Lin, W. & Skorupski, K. The virulence activator AphA links quorum sensing to pathogenesis and physiology in Vi- brio cholerae by repressing the expression of a penicillin amidase gene on the small chromosome. J Bacteriol 185, 4825-4836 (2003). 51. Skorupski, K. & Taylor, R. K. Cyclic AMP and its receptor protein negatively regulate the coordinate expression of cholera toxin and toxin-coregulated pilus in Vibrio cholerae. Proc Natl Acad Sci U S A 94, 265-270 (1997). 52. Kovacikova, G. & Skorupski, K. Overlapping binding sites for the virulence gene regulators AphA, AphB and cAMP-CRP at the Vibrio cholerae tcpPH promoter. Mol Microbiol 41, 393-407 (2001). 53. Liang, W., Pascual-Montano, A., Silva, A. J. & Benitez, J. A. The cyclic AMP receptor protein modulates quorum sensing, motility and multiple genes that affect intestinal colonization in Vibrio cholerae. Microbiology 153, 2964-75 (2007). 54. Waters, C. M. & Bassler, B. L. Quorum sensing: cell-to-cell communication in bacteria. Annu Rev Cell Dev Biol 21, 319-46 (2005). 55. Hammer, B. K. & Bassler, B. L. Quorum sensing controls biofilm formation in Vibrio cholerae. Mol Microbiol 50, 101-104 (2003). 56. Zhu, J. et al. Quorum-sensing regulators control virulence gene expression in Vibrio cholerae. Proc Natl Acad Sci U S A 99, 3129-34 (2002).

- 54 -

57. Lenz, D. H. et al. The small RNA chaperone Hfq and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio cholerae. Cell 118, 69-82 (2004). 58. Miller, M. B., Skorupski, K., Lenz, D. H., Taylor, R. K. & Bassler, B. L. Parallel quorum sensing systems converge to regulate virulence in Vibrio cholerae. Cell 110, 303-314 (2002). 59. Cao, T. B. & Saier, M. H., Jr. The general protein secretory pathway: phylogenetic analyses leading to evolutionary conclusions. Biochim Biophys Acta 1609, 115-25 (2003). 60. Papanikou, E., Karamanou, S. & Economou, A. Bacterial protein secretion through the translocase nanomachine. Nat Rev Microbiol 5, 839-51 (2007). 61. Berks, B. C., Palmer, T. & Sargent, F. Protein targeting by the bacterial twin-arginine translocation (Tat) pathway. Curr Opin Microbiol 8, 174-81 (2005). 62. Holland, I. B., Schmitt, L. & Young, J. Type 1 protein secretion in bacteria, the ABC-transporter dependent pathway (review). Mol Membr Biol 22, 29-39 (2005). 63. Boardman, B. K. & Satchell, K. J. Vibrio cholerae strains with muta- tions in an atypical type I secretion system accumulate RTX toxin in- tracellularly. J Bacteriol 186, 8137-43 (2004). 64. Voulhoux, R. et al. Involvement of the twin-arginine translocation system in protein secretion via the type II pathway. Embo J 20, 6735-41 (2001). 65. Peabody, C. R. et al. Type II protein secretion and its relationship to bacterial type IV pili and archaeal flagella. Microbiology 149, 3051- 72 (2003). 66. Sandkvist, M. et al. General secretion pathway (eps) genes required for toxin secretion and outer membrane biogenesis in Vibrio cholerae. J Bacteriol 179, 6994-7003 (1997). 67. Cianciotto, N. P. Type II secretion: a protein secretion system for all seasons. Trends Microbiol 13, 581-8 (2005). 68. Scott, M. E., Dossani, Z. Y. & Sandkvist, M. Directed polar secretion of protease from single cells of Vibrio cholerae via the type II secretion pathway. Proc Natl Acad Sci U S A 98, 13978-83 (2001). 69. Connell, T. D., Metzger, D. J., Lynch, J. & Folster, J. P. Endochiti- nase is transported to the extracellular milieu by the eps-encoded general secretory pathway of Vibrio cholerae. J Bacteriol 180, 5591- 5600 (1998). 70. Davis, B. M. et al. Convergence of the secretory pathways for cholera toxin and the filamentous phage, CTXphi. Science 288, 333-5 (2000). 71. Galan, J. E. & Wolf-Watz, H. Protein delivery into eukaryotic cells by type III secretion machines. Nature 444, 567-73 (2006).

- 55 -

72. Dziejman, M. et al. Genomic characterization of non-O1, non-O139 Vibrio cholerae reveals genes for a type III secretion system. Proc Natl Acad Sci U S A 102, 3465-70 (2005). 73. Cornelis, G. R. The Yersinia Ysc-Yop 'type III' weaponry. Nat Rev Mol Cell Biol 3, 742-52 (2002). 74. Cossart, P. & Sansonetti, P. J. Bacterial invasion: the paradigms of enteroinvasive pathogens. Science 304, 242-8 (2004). 75. Tam, V. C., Serruto, D., Dziejman, M., Brieher, W. & Mekalanos, J. J. A type III secretion system in Vibrio cholerae translocates a formin/spire hybrid-like actin nucleator to promote intestinal colonization. Cell Host Microbe 1, 95-107 (2007). 76. Backert, S. & Meyer, T. F. Type IV secretion systems and their effectors in bacterial pathogenesis. Curr Opin Microbiol 9, 207-17 (2006). 77. Christie, P. J., Atmakuri, K., Krishnamoorthy, V., Jakubowski, S. & Cascales, E. Biogenesis, architecture, and function of bacterial type IV secretion systems. Annu Rev Microbiol 59, 451-85 (2005). 78. Baron, C. in Bacterial Secreted Proteins: Secretory Mechanisms and Role in Pathogenesis (ed. K, W.) 117-137 (Caister Academic Press, Hethersett, Norwich UK, 2009). 79. Pukatzki, S. et al. Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system. Proc Natl Acad Sci U S A 103, 1528-33 (2006). 80. Desvaux, M., Parham, N. J. & Henderson, I. R. Type V protein secretion: simplicity gone awry? Curr Issues Mol Biol 6, 111-24 (2004). 81. Roggenkamp, A. et al. Molecular analysis of transport and oli- gomerization of the Yersinia enterocolitica adhesin YadA. J Bacteriol 185, 3735-44 (2003). 82. Pohlner, J., Halter, R. & Meyer, T. F. Neisseria gonorrhoeae IgA protease. Secretion and implications for pathogenesis. Antonie Van Leeuwenhoek 53, 479-84 (1987). 83. Poulsen, K., Brandt, J., Hjorth, J. P., Thogersen, H. C. & Kilian, M. Cloning and sequencing of the immunoglobulin A1 protease gene (iga) of Haemophilus influenzae serotype b. Infect Immun 57, 3097- 105 (1989). 84. Fischer, W., Buhrdorf, R., Gerland, E. & Haas, R. Outer membrane targeting of passenger proteins by the vacuolating cytotoxin autotransporter of Helicobacter pylori. Infect Immun 69, 6769-75 (2001). 85. Benjelloun-Touimi, Z., Sansonetti, P. J. & Parsot, C. SepA, the major extracellular protein of Shigella flexneri: autonomous secretion and involvement in tissue invasion. Mol Microbiol 17, 123-35 (1995). 86. Henderson, I. R. & Owen, P. The major phase-variable outer membrane protein of Escherichia coli structurally resembles the immunoglobulin A1 protease class of exported protein and is regulated by a novel mechanism involving Dam and oxyR. J Bacteriol 181, 2132- 41 (1999).

- 56 -

87. St Geme, J. W., 3rd & Cutter, D. The Haemophilus influenzae Hia adhesin is an autotransporter protein that remains uncleaved at the C terminus and fully cell associated. J Bacteriol 182, 6005-13 (2000). 88. Cope, L. D. et al. Characterization of the Moraxella catarrhalis uspA1 and uspA2 genes and their encoded products. J Bacteriol 181, 4026-34 (1999). 89. Schiebel, E., Schwarz, H. & Braun, V. Subcellular location and unique secretion of the hemolysin of Serratia marcescens. J Biol Chem 264, 16311-20 (1989). 90. Thanassi, D. G., Stathopoulos, C., Karkal, A. & Li, H. Protein secretion in the absence of ATP: the autotransporter, two-partner secretion and chaperone/usher pathways of gram-negative bacteria (review). Mol Membr Biol 22, 63-72 (2005). 91. Das, S. & Chaudhuri, K. Identification of a unique IAHP (IcmF associated homologous proteins) cluster in Vibrio cholerae and other proteobacteria through in silico analysis. In Silico Biol 3, 287-300 (2003). 92. Bingle, L. E., Bailey, C. M. & Pallen, M. J. Type VI secretion: a be- ginner's guide. Curr Opin Microbiol 11, 3-8 (2008). 93. Cascales, E. The type VI secretion toolkit. EMBO Rep 9, 735-41 (2008). 94. Shrivastava, S. & Mande, S. S. Identification and functional characterization of gene components of Type VI Secretion system in bacterial genomes. PLoS One 3, e2955 (2008). 95. Ishikawa, T., Rompikuntal, P. K., Lindmark, B., Milton, D. L. & Wai, S. N. Quorum sensing regulation of the two hcp alleles in Vibrio cholerae O1 strains. PLoS One 4, e6734 (2009). 96. Joelsson, A., Liu, Z. & Zhu, J. Genetic and phenotypic diversity of quorum-sensing systems in clinical and environmental isolates of Vibrio cholerae. Infect Immun 74, 1141-7 (2006). 97. Beveridge, T. J. Structures of gram-negative cell walls and their derived membrane vesicles. J Bacteriol 181, 4725-33 (1999). 98. Kadurugamuwa, J. L. & Beveridge, T. J. Virulence factors are released from Pseudomonas aeruginosa in association with membrane vesicles during normal growth and exposure to gentamicin: a novel mechanism of enzyme secretion. J Bacteriol 177, 3998-4008 (1995). 99. Mashburn-Warren, L. M. & Whiteley, M. Special delivery: vesicle trafficking in prokaryotes. Mol Microbiol 61, 839-46 (2006). 100. Wai, S. N. et al. Vesicle-mediated export and assembly of pore- forming oligomers of the enterobacterial ClyA cytotoxin. Cell 115, 25- 35 (2003). 101. Kato, S., Kowashi, Y. & Demuth, D. R. Outer membrane-like vesicles secreted by Actinobacillus actinomycetemcomitans are enriched in leukotoxin. Microb Pathog 32, 1-13 (2002).

- 57 -

102. Nowotny, A. et al. Release of toxic microvesicles by Actinobacillus actinomycetemcomitans. Infect Immun 37, 151-4 (1982). 103. Shoberg, R. J. & Thomas, D. D. Borrelia burgdorferi vesicle production occurs via a mechanism independent of immunoglobulin M involvement. Infect Immun 63, 4857-61 (1995). 104. Lindmark, B. et al. Outer membrane vesicle-mediated release of cytolethal distending toxin (CDT) from Campylobacter jejuni. BMC Microbiol 9, 220 (2009). 105. Wai, S. N., Takade, A. & Amako, K. The release of outer membrane vesicles from the strains of enterotoxigenic Escherichia coli. Micro- biol. Immunol. 39, 451-6 (1995). 106. Balsalobre, C. et al. Release of the type I secreted alpha-haemolysin via outer membrane vesicles from Escherichia coli. Mol. Microbiol. 59, 99-112 (2006). 107. Kouokam, J. C. et al. Active cytotoxic necrotizing factor 1 associated with outer membrane vesicles from uropathogenic Escherichia coli. Infect. Immun. 74, 2022-30 (2006). 108. Horstman, A. L. & Kuehn, M. J. Enterotoxigenic Escherichia coli secretes active heat-labile enterotoxin via outer membrane vesicles. J Biol Chem 275, 12489-12496 (2000). 109. Kolling, G. L. & Matthews, K. R. Export of virulence genes and Shiga toxin by membrane vesicles of Escherichia coli O157:H7. Appl Envi- ron Microbiol 65, 1843-8 (1999). 110. Fiocca, R. et al. Release of Helicobacter pylori vacuolating cytotoxin by both a specific secretion pathway and budding of outer membrane vesicles. Uptake of released toxin and vesicles by gastric epithelium. J Pathol 188, 220-6 (1999). 111. Tan, T. T., Morgelin, M., Forsgren, A. & Riesbeck, K. Haemophilus influenzae survival during complement-mediated attacks is pro- moted by Moraxella catarrhalis outer membrane vesicles. J Infect Dis 195, 1661-70 (2007). 112. Li, Z., Clarke, A. J. & Beveridge, T. J. A major autolysin of Pseudo- monas aeruginosa: subcellular distribution, potential role in cell growth and division and secretion in surface membrane vesicles. J Bacteriol 178, 2479-88 (1996). 113. MacEachran, D. P. et al. The Pseudomonas aeruginosa secreted protein PA2934 decreases apical membrane expression of the cystic fibrosis transmembrane conductance regulator. Infect Immun 75, 3902-12 (2007). 114. Dutta, S. et al. Release of Shiga toxin by membrane vesicles in Shig- ella dysenteriae serotype 1 strains and in vitro effects of antimicro- bials on toxin production and release. Microbiol Immunol 48, 965-9 (2004). 115. Kadurugamuwa, J. L. & Beveridge, T. J. Delivery of the non- membrane-permeative antibiotic gentamicin into mammalian cells

- 58 -

by using Shigella flexneri membrane vesicles. Antimicrob Agents Chemother 42, 1476-83 (1998). 116. Rosen, G., Naor, R., Rahamim, E., Yishai, R. & Sela, M. N. Proteases of Treponema denticola outer sheath and extracellular vesicles. In- fect Immun 63, 3973-9 (1995). 117. Chi, B., Qi, M. & Kuramitsu, H. K. Role of dentilisin in Treponema denticola epithelial cell layer penetration. Res Microbiol 154, 637-43 (2003). 118. Khandelwal, P. & Banerjee-Bhatnagar, N. Insecticidal activity associated with the outer membrane vesicles of Xenorhabdus nemato- philus. Appl Environ Microbiol 69, 2032-7 (2003). 119. Bomberger, J. M. et al. Long-distance delivery of bacterial virulence factors by Pseudomonas aeruginosa outer membrane vesicles. PLoS Pathog 5, e1000382 (2009). 120. Furuta, N. et al. Porphyromonas gingivalis outer membrane vesicles enter human epithelial cells via an endocytic pathway and are sorted to lysosomal compartments. Infect Immun 77, 4187-96 (2009). 121. Kesty, N. C., Mason, K. M., Reedy, M., Miller, S. E. & Kuehn, M. J. Enterotoxigenic Escherichia coli vesicles target toxin delivery into mammalian cells. Embo J 23, 4538-49 (2004). 122. Meurman, J. H., Sanz, M. & Janket, S. J. Oral health, atherosclero- sis, and cardiovascular disease. Crit Rev Oral Biol Med 15, 403-13 (2004). 123. McBroom, A. J. & Kuehn, M. J. Release of outer membrane vesicles by Gram-negative bacteria is a novel envelope stress response. Mol Microbiol 63, 545-58 (2007). 124. De Las Penas, A., Connolly, L. & Gross, C. A. The sigmaE-mediated response to extracytoplasmic stress in Escherichia coli is transduced by RseA and RseB, two negative regulators of sigmaE. Mol Microbiol 24, 373-85 (1997). 125. Song, T. et al. A new Vibrio cholerae sRNA modulates colonization and affects release of outer membrane vesicles. Mol Microbiol 70, 100-11 (2008). 126. Hall-Stoodley, L., Costerton, J. W. & Stoodley, P. Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Micro- biol 2, 95-108 (2004). 127. Schooling, S. R. & Beveridge, T. J. Membrane vesicles: an overlooked component of the matrices of biofilms. J Bacteriol 188, 5945-57 (2006). 128. Ciofu, O., Beveridge, T. J., Kadurugamuwa, J., Walther-Rasmussen, J. & Hoiby, N. Chromosomal beta-lactamase is packaged into membrane vesicles and secreted from Pseudomonas aeruginosa. J An- timicrob Chemother 45, 9-13 (2000). 129. Kadurugamuwa, J. L., Clarke, A. J. & Beveridge, T. J. Surface action of gentamicin on Pseudomonas aeruginosa. J Bacteriol 175, 5798- 805 (1993).

- 59 -

130. Grenier, D., Bertrand, J. & Mayrand, D. Porphyromonas gingivalis outer membrane vesicles promote bacterial resistance to chlorhexi- dine. Oral Microbiol Immunol 10, 319-20 (1995). 131. Fredriksen, J. H. et al. Production, characterization and control of MenB-vaccine "Folkehelsa": an outer membrane vesicle vaccine against group B meningococcal disease. NIPH Ann 14, 67-79; discussion 79-80 (1991). 132. Girard, M. P., Preziosi, M. P., Aguado, M. T. & Kieny, M. P. A review of vaccine research and development: meningococcal disease. Vac- cine 24, 4692-700 (2006). 133. Alaniz, R. C., Deatherage, B. L., Lara, J. C. & Cookson, B. T. Mem- brane vesicles are immunogenic facsimiles of Salmonella typhi- murium that potently activate dendritic cells, prime B and T cell responses, and stimulate protective immunity in vivo. J Immunol 179, 7692-701 (2007). 134. Kesavalu, L., Ebersole, J. L., Machen, R. L. & Holt, S. C. Porphyro- monas gingivalis virulence in mice: induction of immunity to bacterial components. Infect Immun 60, 1455-64 (1992). 135. Keenan, J. I., Rijpkema, S. G., Durrani, Z. & Roake, J. A. Differences in immunogenicity and protection in mice and guinea pigs following intranasal immunization with Helicobacter pylori outer membrane antigens. FEMS Immunol Med Microbiol 36, 199-205 (2003). 136. Schild, S., Nelson, E. J. & Camilli, A. Immunization with Vibrio cholerae outer membrane vesicles induces protective immunity in mice. Infect Immun 76, 4554-63 (2008). 137. Schild, S., Nelson, E. J., Bishop, A. L. & Camilli, A. Characterization of Vibrio cholerae outer membrane vesicles as a candidate vaccine for cholera. Infect Immun 77, 472-84 (2009). 138. Stoodley, P., Sauer, K., Davies, D. G. & Costerton, J. W. Biofilms as complex differentiated communities. Annu Rev Microbiol 56, 187- 209 (2002). 139. Donlan, R. M. & Costerton, J. W. Biofilms: survival mechanisms of clinically relevant microorganisms. Clin Microbiol Rev 15, 167-93 (2002). 140. Huq, A. et al. Coexistence of Vibrio cholerae O1 and O139 Bengal in plankton in Bangladesh. Lancet 345, 1249 (1995). 141. Huq, A. et al. Ecological relationships between Vibrio cholerae and planktonic crustacean copepods. Appl Environ Microbiol 45, 275-83 (1983). 142. Nalin, D. R., Daya, V., Reid, A., Levine, M. M. & Cisneros, L. Adsorp- tion and growth of Vibrio cholerae on chitin. Infect Immun 25, 768- 70 (1979). 143. Rawlings, T. K., Ruiz, G. M. & Colwell, R. R. Association of Vibrio cholerae O1 El Tor and O139 Bengal with the Copepods Acartia tonsa and Eurytemora affinis. Appl Environ Microbiol 73, 7926-33 (2007).

- 60 -

144. Colwell, R. R. Global climate and infectious disease: the cholera paradigm. Science 274, 2025-31 (1996). 145. Colwell, R. R. et al. Reduction of cholera in Bangladeshi villages by simple filtration. Proc Natl Acad Sci U S A 100, 1051-5 (2003). 146. Bird, A. & Bird, J. The structure of nematodes. (New York: Academic Press, 1991). 147. Gooday, G. W. in Advances in microbial ecology (ed. Marshall, K. C.) 387-430 (Plenum Press, New York, N.Y, 1990). 148. Reguera, G. & Kolter, R. Virulence and the environment: a novel role for Vibrio cholerae toxin-coregulated pili in biofilm formation on chitin. J Bacteriol 187, 3551-5 (2005). 149. Kirn, T. J., Jude, B. A. & Taylor, R. K. A colonization factor links Vibrio cholerae environmental survival and human infection. Nature 438, 863-6 (2005). 150. Meibom, K. L. et al. The Vibrio cholerae chitin utilization program. Proc Natl Acad Sci U S A 101, 2524-9 (2004). 151. Meibom, K. L., Blokesch, M., Dolganov, N. A., Wu, C. Y. & Schoolnik, G. K. Chitin induces natural competence in Vibrio cholerae. Science 310, 1824-7 (2005). 152. Broza, M. & Halpern, M. Pathogen reservoirs. Chironomid egg masses and Vibrio cholerae. Nature 412, 40 (2001). 153. Halpern, M., Broza, Y. B., Mittler, S., Arakawa, E. & Broza, M. Chi- ronomid egg masses as a natural reservoir of Vibrio cholerae non-O1 and non-O139 in freshwater habitats. Microb Ecol 47, 341-9 (2004). 154. Broza, M., Gancz, H., Halpern, M. & Kashi, Y. Adult non-biting midges: possible windborne carriers of Vibrio cholerae non-O1 non- O139. Environ Microbiol 7, 576-85 (2005). 155. Mulder, G. D., Ries, T. M., Beaver, T. R. & Cover, T. Nontoxigenic Vibrio cholerae wound infection after exposure to contaminated lake water. J Infect Dis 159, 809-11 (1989). 156. Magnusson, M. R. & Pegg, S. P. Vibrio cholerae non-O1 primary septicaemia following a large thermal burn. Burns 22, 44-7 (1996). 157. Cover, T. L., Dunn, B. E., Ellison, R. T., 3rd & Blaser, M. J. Vibrio cholerae wound infection acquired in Colorado. J Infect Dis 160, 1083 (1989). 158. Handrick, W. et al. [Otitis caused by Vibrio cholerae non-01/non- 0139 strains acquired in Germany]. MMW Fortschr Med 146, 38-9 (2004). 159. El-Hiday, A. H., Khan, F. Y., Al Maslamani, M. & El Shafie, S. Bac- teremia and spontaneous bacterial peritonitis due to Vibrio cholerae (non-O1 non-O139) in liver cirrhosis. Indian J Gastroenterol 25, 107 (2006). 160. Restrepo, D., Huprikar, S. S., VanHorn, K. & Bottone, E. J. O1 and non-O1 Vibrio cholerae bacteremia produced by hemolytic strains. Diagn Microbiol Infect Dis 54, 145-8 (2006).

- 61 -

161. Strumbelj, I. et al. A case of Vibrio cholerae non-O1, non-O139 septicaemia in Slovenia, imported from Tunisia, July 2005. Euro Sur- veill 10, E051020 6 (2005). 162. Thibaut, K., Van de Heyning, P. & Pattyn, S. R. Isolation of non-01 V. cholerae from ear tracts. Eur J Epidemiol 2, 316-7 (1986). 163. Gordon, M. A. et al. Three cases of bacteremia caused by Vibrio cholerae O1 in Blantyre, Malawi. Emerg Infect Dis 7, 1059-61 (2001). 164. Eiler, A., Johansson, M. & Bertilsson, S. Environmental influences on Vibrio populations in northern temperate and boreal coastal waters (Baltic and Skagerrak Seas). Appl Environ Microbiol 72, 6004- 11 (2006). 165. Schönning, C., Carlander, A., Hansen, A. & Thors, C. Hur vanlig är badsårsfeber. Smittskydd 3 (2008). 166. Andersson Y & K, E. Wound infections due to Vibrio cholerae in Sweden after swimming in the Baltic Sea, summer 2006. Euro Sur- veill. 11, 3013 (2006). 167. Faruque, S. M. et al. Genetic diversity and virulence potential of environmental Vibrio cholerae population in a cholera-endemic area. Proc Natl Acad Sci U S A 101, 2123-8 (2004). 168. Nair, G. B. et al. Toxin profiles of Vibrio cholerae non-O1 from environmental sources in Calcutta, India. Appl Environ Microbiol 54, 3180-2 (1988). 169. Chow, K. H., Ng, T. K., Yuen, K. Y. & Yam, W. C. Detection of RTX toxin gene in Vibrio cholerae by PCR. J Clin Microbiol 39, 2594-7 (2001). 170. Ghosh, A. et al. Enterotoxigenicity of mature 45-kilodalton and processed 35-kilodalton forms of hemagglutinin protease purified from a cholera toxin gene-negative Vibrio cholerae non-O1, non- O139 strain. Infect Immun 74, 2937-46 (2006). 171. Arita, M., Takeda, T., Honda, T. & Miwatani, T. Purification and characterization of Vibrio cholerae non-O1 heat-stable enterotoxin. Infect Immun 52, 45-9 (1986). 172. Pal, A. et al. Reassessment of the prevalence of heat-stable enterotoxin (NAG-ST) among environmental Vibrio cholerae non-O1 strains isolated from Calcutta, India, by using a NAG-ST DNA probe. Appl Environ Microbiol 58, 2485-9 (1992). 173. Johnson, J. A., Panigrahi, P. & Morris, J. G., Jr. Non-O1 Vibrio cholerae NRT36S produces a polysaccharide capsule that determines colony morphology, serum resistance, and virulence in mice. Infect Immun 60, 864-9 (1992). 174. Roantree, R. J. & Rantz, L. A. A Study Of The Relationship Of The Normal Bactericidal Activity Of Human Serum To Bacterial Infec- tion. J Clin Invest 39, 72-81 (1960). 175. Podschun, R. & Ullmann, U. Klebsiella capsular type K7 in relation to toxicity, susceptibility to phagocytosis and resistance to serum. J Med Microbiol 36, 250-4 (1992).

- 62 -

176. Carruthers, M. M. & Kabat, W. J. Vibrio vulnificus (lactose-positive vibrio) and Vibrio parahaemolyticus differ in their susceptibilities to human serum. Infect Immun 32, 964-6 (1981). 177. Jacques, M., Gottschalk, M., Foiry, B. & Higgins, R. Ultrastructural study of surface components of Streptococcus suis. J Bacteriol 172, 2833-8 (1990). 178. Johnson, J. A. et al. Vibrio cholerae O139 synonym bengal is closely related to Vibrio cholerae El Tor but has important differences. In- fect Immun 62, 2108-10 (1994). 179. Jesudason, M. V., Cherian, A. M. & John, T. J. Blood stream invasion by Vibrio cholerae O139. Lancet 342, 431 (1993). 180. Li, Z., Zhang, S., Wang, C. & Pang, Q. Complement-mediated killing of Vibrio species by the humoral fluids of amphioxus Branchiostoma belcheri: implications for a dual role of O-antigens in the resistance to bactericidal activity. Fish Shellfish Immunol 24, 215-22 (2008). 181. De, S. N. & Chatterje, D. N. An experimental study of the mechanism of action of Vibriod cholerae on the intestinal mucous membrane. J Pathol Bacteriol 66, 559-62 (1953). 182. Spira, W. M., Sack, R. B. & Froehlich, J. L. Simple adult rabbit model for Vibrio cholerae and enterotoxigenic Escherichia coli diarrhea. Infect Immun 32, 739-47 (1981). 183. Richardson, S. H., Giles, J. C. & Kruger, K. S. Sealed adult mice: new model for enterotoxin evaluation. Infect Immun 43, 482-6 (1984). 184. Baselski, V., Briggs, R. & Parker, C. Intestinal fluid accumulation induced by oral challenge with Vibrio cholerae or cholera toxin in infant mice. Infect Immun 15, 704-12 (1977). 185. Sifri, C. D., Begun, J. & Ausubel, F. M. The worm has turned-- microbial virulence modeled in Caenorhabditis elegans. Trends Mi- crobiol 13, 119-27 (2005). 186. Vodovar, N., Acosta, C., Lemaitre, B. & Boccard, F. Drosophila: a polyvalent model to decipher host-pathogen interactions. Trends Microbiol 12, 235-42 (2004). 187. Blow, N. S. et al. Vibrio cholerae infection of Drosophila melanogaster mimics the human disease cholera. PLoS Pathog 1, e8 (2005). 188. Altun, Z. F. & Hall, D. H. (2005). 189. Bargmann, C. I. Chemosensation in C. elegans. WormBook, 1-29 (2006). 190. Darby, C., Hsu, J. W., Ghori, N. & Falkow, S. Caenorhabditis elegans: plague bacteria biofilm blocks food intake. Nature 417, 243- 4 (2002). 191. Ewbank, J. J. Tackling both sides of the host-pathogen equation with Caenorhabditis elegans. Microbes Infect 4, 247-56 (2002). 192. Kurz, C. L. & Tan, M. W. Regulation of aging and innate immunity in C. elegans. Aging Cell 3, 185-93 (2004).

- 63 -

193. Kato, Y. et al. abf-1 and abf-2, ASABF-type antimicrobial peptide genes in Caenorhabditis elegans. Biochem J 361, 221-30 (2002). 194. Ewbank, J. J. Signaling in the immune response. WormBook, 1-12 (2006). 195. Alper, S., McBride, S. J., Lackford, B., Freedman, J. H. & Schwartz, D. A. Specificity and complexity of the Caenorhabditis elegans innate immune response. Mol Cell Biol 27, 5544-53 (2007). 196. Tenor, J. L. & Aballay, A. A conserved Toll-like receptor is required for Caenorhabditis elegans innate immunity. EMBO Rep 9, 103-9 (2008). 197. Garvis, S. et al. Caenorhabditis elegans semi-automated liquid screen reveals a specialized role for the chemotaxis gene cheB2 in Pseudomonas aeruginosa virulence. PLoS Pathog 5, e1000540 (2009). 198. Peperzak, L., Colijn, F., Vrieling, E. G., Gieskes, W. W. C. & Peeters, J. C. H. Observations of flagellates in colonies of Phaeocystis glo- bosa (Prymnesiophyceae); a hypothesis for their position in the life cycle. Journal of Plankton Research 22, 2181-2203 (2000). 199. Zhang, Y., Foster, J. M., Nelson, L. S., Ma, D. & Carlow, C. K. The chitin synthase genes chs-1 and chs-2 are essential for C. elegans development and responsible for chitin deposition in the eggshell and pharynx, respectively. Dev Biol 285, 330-9 (2005). 200. Kuehn, M. J. & Kesty, N. C. Bacterial outer membrane vesicles and the host-pathogen interaction. Genes Dev. 19, 2645-55 (2005). 201. Singh, D. V. et al. Molecular analysis of Vibrio cholerae O1, O139, non-O1, and non-O139 strains: clonal relationships between clinical and environmental isolates. Appl Environ Microbiol 67, 910-21 (2001). 202. Li, C. C., Merrell, D. S., Camilli, A. & Kaper, J. B. ToxR interferes with CRP-dependent transcriptional activation of ompT in Vibrio cholerae. Mol Microbiol 43, 1577-1589 (2002).

- 64 -