Helsinki 2006

Functional Studies of Purified Transmembrane Proteases, Omptins, of Yersinia pestis and Salmonella enterica

4/2006 Kari Kreander A Study on Bacteria-Targeted Screening and in vitro Safety Assessment of Natural Products 5/2006 Gudrun Wahlström From Actin Monomers to Bundles: The Role of Twinfilin and a-Actinin in Drosophila melanogaster Development 6/2006 Jussi Joensuu Production of F4 Fimbrial Adhesin in Plants: A Model for Oral Porcine Vaccine against Enterotoxigenic Escherichia coli 7/2006 Heikki Vilen Mu in vitro Transposition Technology in Functional Genetics and Genomics: Applications on Mouse and Bacteriophages 8/2006 Jukka Pakkanen Upregulation and Functionality of Neuronal Nicotinic Acetylcholine Receptors 9/2006 Antti Leinonen Novel Mass Spectrometric Analysis Methods for Anabolic Androgenic Steroids in Sports Drug Testing 10/2006 Paulus Seitavuopio The Roughness and Imaging Characterisation of Different Pharmaceutical Surfaces 11/2006 Leena Laitinen Caco-2 Cell Cultures in the Assessment of Intestinal Absorption: Effects of Some Co-Administered Drugs and Natural Compounds in Biological Matrices 12/2006 Pirjo Wacklin Biodiversity and Phylogeny of Planktic Cyanobacteria in Temperate Freshwater Lakes 13/2006 Antti Alaranta Medication Use in Elite Athletes 14/2006 Anna-Helena Saariaho Characterization of the Molecular Components and Function of the BARE-1, Hin-Mu and Mu Transposition Machineries 15/2006 Jaana Vaitomaa The Effects of Environmental Factors on Biomass and Microcystin Production by the Freshwater Cyanobacterial Genera Microcystis and Anabaena 16/2006 Vootele Voikar Evaluation of Methods and Applications for Behavioural Profiling of Transgenic Mice 17/2006 Päivi Lindfors GDNF Family Receptors in Peripheral Target Innervation and Hormone Production 18/2006 Tarja Kariola Pathogen-Induced Defense Signaling and Signal Crosstalk in Arabidopsis 19/2006 Minna M. Jussila Molecular Biomonitoring During Rhizoremediation of Oil-Contaminated Soil 20/2006 Bamidele Raheem Developments and Microbiological Applications in African Foods: Emphasis on Nigerian Wara Cheese 21/2006 Jiri Lisal Mechanism of RNA Translocation by a Viral Packaging Motor 22/2006 Roosa Laitinen Gerbera cDNA Microarray: A Tool for Large-Scale Identification of Genes Involved in Flower Development 23/2006 Lari Lehtiö Enzymes with Radical Tendencies: The PFL Family

LEANDRO LOBO

Recent Publications in this Series:

ISSN 1795-7079 ISBN 952-10-3437-8

Functional Studies of Purified Transmembrane Proteases, Omptins, of Yersinia pestis and Salmonella enterica

LEANDRO ARAUJO LOBO Department of Biological and Environmental Sciences General Microbiology Faculty of Biosciences and Viikki Graduate School in Biosciences University of Helsinki

Dissertationes bioscientiarum molecularium Universitatis Helsingiensis in Viikki 24/2006

24/2006

Functional studies of purified transmembrane proteases, omptins, of Yersinia pestis and Salmonella enterica.

Leandro Araujo Lobo

Department of Biological and Environmental Sciences Faculty of Biosciences and Viikki Graduate School in Biosciences University of Helsinki

Academic Dissertation in General Microbiology

To be presented, with the permission of the Faculty of Biosciences of the University of Helsinki, for public criticism in the auditorium 2 at Viikki Infocenter Korona (Viikinkaari 11, Helsinki) on October 19th, 2006 at 12 o’clock noon. Helsinki 2006

Supervisor

Professor Timo Korhonen, Ph.D Department of Biological and Environmental Sciences Faculty of Biosciences University of Helsinki

Reviewers

Docent Pentti Kuusela, MD, Ph.D Department of bacteriology and Immunology Haartman Institute University of Helsinki Docent Ilkka Helander, Ph.D

Opponent

Professor Mikael Skurnik, Ph.D Department of bacteriology and Immunology Haartman Institute University of Helsinki

Cover illustration: View of Ruttopuisto (Pest park) with the old church in the back.

ISBN 952-10-3437-8 ISBN 952-10-3438-6 (PDF version, http//:ethesis.helsinki.fi) ISSN 1795-7079 Edita Prima Oy Helsinki 2006

Ao meu Avô Petrônio

Preface This work was conducted at the General Microbiology, Department of Biological and Environmental Sciences, Faculty of Biosciences in the University of Helsinki. I thank the Head of the Department, Professor Kielo Haahtela for providing a wealthy environment of creativity and outstanding working facilities. I owe my deepest gratitude to Professor Timo Korhonen, for giving me the opportunity to work at the bacterial adhesion group. Timo’s vast knowledge in microbiology and clear-cut scientific thinking were always an inspiration for me. I want to thank for your support, even in the most arduous times, and for helping me reach the end of this journey. I sincerely thank docent Pentti Kuusela and Dr. Ilkka Helander for reviewing my thesis with great expertise in such a tight schedule and for offering such insightful and valuable comments for the thesis. I am grateful to all the distant-based co-authors for their valuable collaboration to this work: Andrey Anisimov, Eva Bjur, Klaus Brandenburg, Otto Holst, Ilkka Julkunen, Yuriy Knirel, Buko Lindner, Minja Miettinen, Hanna Raukkola, Mikael Rhen I want to thank all the members of the Bacterial adhesion group for warmly welcoming me and creating a pleasant work atmosphere. Ritva and Kaarina were amazing advisors, with sharp knowledge of bacterial pathogenesis that helped and guided me during my PhD studies. My first roomies, Maini and Päivi, thanks for the pleasant conversations and enjoyable feeling in our writing room. And as I once said, if Maini could get a cent for every question I asked her, she would be a millionaire by now, thanks for being a good friend Maini! Päivi’s liveliness and good spirit made a great composition in our writing room. I also had the pleasure to share the same space with Johanna and Timo which managed to keep the same cheerful feeling. Jenni was a great lab companion and joyfully helped me innumerous times with practical issues. Many thanks to all the current and previous members of Timo’s group: Benita, Katariina, Katri, Leena, Lotta, Marjo, Raili, Riikka, Sanna, Ulla and Veera. It was a great adventure! I learned a lot from all of you and had a great time working in YMBO. I especially wish to thanks the members of the technical staff Raili, Erkki, Riitta, Pirkko and Merja for their efficient and skilled support. Thanks to Anita and Eeva for very efficient administrative assistance and for keeping me under insurance cover until the very last minute! I want to express my gratitude to my “new” and good friends here in Finland. My eternal roommate, Mana. To Anmi, Iida, Maria and Suvi, my dear friends and first Finns I ever met (thanks to Juarez!). The Holy Virgins: Hipe, Jutta, Maarit and Eeva. The guys, Moacir, Ricardo and Peter. To Satu, Rodrigo, Miltinho, Lidi, Jenni hyena, Arska, Jarkko, Anna. You all helped to keep my days shining even in the coldest winter day and made my stay in Finland unforgettable. Thanks to all my friends in Brasil, there are so many names that should be included here, galera do grupal, galera de Nilo! I hope you all know how important you are to me. You always kept me going, kept in a good mood and helped me overcome in hard times. I also have to thank you all for keeping me update of the events in Rio. Emails

describing all the barbecues, all the parties, all the trips…thanks for more than three years of torture! To my valkoista, Simona, for all your love and companion, thank you. Miliu Tave! My very special thanks go to my family, specialy to my parents Norma and Cândido and grandparents Petrônio and Jacy. You always supported me and gave me strength to continue on this journey. Special thanks to my Aunt Sonia for keeping my stash of brazilian specialties always full!

Table of contents List of original publications ...................................................................................... VII Summary....................................................................................................................VIII 1. Introduction .............................................................................................................. 1 1.1. Yersinia pestis and Salmonella enterica as bacterial pathogens ......................... 1 1.2. Surface proteolysis in bacterial infections .......................................................... 3 1.3. β-barrel outer membrane proteins ...................................................................... 5 1.4. The omptin family of bacterial proteases ........................................................... 6 1.5. Structure of omptins ........................................................................................... 7 1.6. Function of omptins ............................................................................................ 9 1.7. Lipopolysaccharide........................................................................................... 13 1.7.1. Characteristics of the LPS of Y.pestis ..................................................... 15 1.7.2. Characteristics of the LPS of S.enterica ................................................. 16 2. Aims of the study ................................................................................................... 17 3. Material and methods............................................................................................ 18 3.1. Cultivation of Y. pestis ...................................................................................... 21 3.2. Extraction of outer membrane of Y. pestis ........................................................ 21 4. Results and discussion ........................................................................................... 22 4.1. Reactivation of purified His6-Pla with E.coli K12 LPS and coating onto FMPs (I, III) ................................................................................. 22 4.2. Adhesion of His6-Pla-FMPs to ECM proteins (I) ............................................. 23 4.3. Reactivation of His6-Pla with different LPS chemotypes (II)........................... 24 4.4. Effect of phosphate and NaCl on His6-Pla activity (III) ................................... 25 4.5. Effect of mutations in the LPS-binding motif of Pla on plasminogen activation (III) ............................................................................. 25 4.6. Effect of growth temperature in Y. pestis on the activity of Pla (III)................ 26 4.7. Degradation of gelatin and collagen by Pla and PgtE (II) ................................ 27 5. Conclusion .............................................................................................................. 31 6. References ............................................................................................................... 33

List of original publications I.

Lobo, L.A. (2006) Adhesive properties of the purified plasminogen activator Pla of Yersinia pestis. FEMS Microbiol. Lett. 262:158-62

II.

Lobo, L.A., Brandenburg, K., Lindner, B., Suomalainen, M., Virkola, R., Knirel, Y.A., Anisimov A.P., Holst, O. and Korhonen, T.K. (2006) Structural features in lipopolysaccharide that affect plasminogen activation by the surface protease Pla, a major virulence factor of Yersinia Pestis. Submitted to Infect. Immun.

III.

Ramu P., Lobo L.A., Kukkonen M., Bjur, E., Suomalainen, M., Raukola, H., Miettinen, M., Julkunen, I., Holst, O., Rhen, M., Korhonen, T.K. & Lähteenmäki, K. (2006) Generation of gelatinase activity by Salmonella enterica: activation of host pro-matrix metallo proteinase-9 and degradation of gelatin. Submitted to Infect. Immun.

Additional material is presented in the text.

VII

Summary Surface proteolysis is important in migration of cells through tissue barriers. In the case of prokaryotes, surface proteolysis has been associated with invasiveness of pathogenic bacteria from the primary infection site into circulation and secondary infection sites in the host. This study addressed surface proteases of two important bacterial pathogens, Yersinia pestis which is the causative agent of the lethal systemic zoonosis, plague, and Salmonella enterica serovar Typhimurium which is an oral-faecal pathogen that annually causes millions of cases of gastoenteritis that may develop to septicaemia. Both bacterial species express an ortholog of the omptin family of transmembrane β-barrel, outer membrane proteases/adhesins. This thesis work addressed the functions of isolated plasminogen activator Pla of Y. pestis and the PgtE omptin of S. enterica. Pla and PgtE were isolated as His6-fusion proteins in denaturing conditions from recombinant Escherichia coli and activated by adding lipopolysaccharide (LPS). The structural features in LPS that enhance plasminogen activation by His6-Pla were determined, and it was found that the lack of O-specific chain, the presence of outer core oligosaccharide, the presence of phosphates in lipid A, as well as a low level of acylation in lipid A influence the enhancement of Pla activity by LPS. A conserved lipid A phosphate – binding motif in Pla and PgtE was found important for the enhancement of enzymatic activity by LPS. The results help to explain the biological significance of the genetic loss of the O-specific chain biosynthesis in Y. pestis as well as the variations in LPS structure upon entry of Y. pestis into the human host. Expression of Pla in Y. pestis is associated with adhesiveness to lamin of basement membranes. Here, isolated and LPS-activated His6-Pla was coated onto fluorescent microparticles. The coating conferred specific adhesiveness of the particles to laminin and reconstituted basement membrane, thus confirming the intrinsic adhesive characteristics of the Pla protein. The adhesiveness is thought to direct plasmin proteolysis at tissue barriers, thus increasing tissue damage and bacterial spread. Gelatinase activity has not been previously reported in enteric bacteria. Expression of PgtE in S. enterica was associated with cleavage of porcine skin gelatin, denaturated human type I collagen, as well as DQ-gelatin. Purified His6-PgtE also degraded porcine skin gelatin and human type I gelatin but did not react with DQ-gelatin, indicating that minor differences are seen in proteolysis by isolated and cell-bound PgtE. Pla was less effective in gelatin degradation. The novel gelatinase activity in S. enterica is likely to enhance bacterial dissemination during infection.

VIII

Introduction

1. Introduction 1.1. Yersinia pestis and Salmonella enterica as bacterial pathogens Y. pestis is the causative agent of plague, a zoonotic disease that affects rodents and is transmitted to humans mainly through the bite of infected fleas. Plague is one of the most feared human diseases, and it has been estimated that 200 million humans were killed by this pathogen in the course of recorded history (Reviewed in Perry & Fetherston, 1997). Globally, the World Health Organization reports 1,000 to 3,000 cases of plague every year; recent outbreaks of plague (Word Health Organization, 1990) and the reemergence of the disease in endemic areas after natural or socioeconomic aggravation, show that the plague disease is still a concern for public health and that surveillance must be maintained (Duplantier et al., 2005). Y. pestis belongs to the Enterobacteriaceae family and is the type species of the Yersinia genus, which also contains two other species pathogenic for humans: Yersinia pseudotuberculosis and Yersinia enterocolitica. DNA hybridization assays (Bercovier et al., 1980) and the recent genomic sequencing of Y. pestis and Y. pseudotuberculosis (Parkhill et al., 2001; Chain et al., 2004) have shown that Y.pestis is genetically very close to the enteropathogenic species Y. pseudotuberculosis. Indeed, sequence analyses of cytoplasmic housekeeping genes as well as genes involved in the biosynthesis of lipopolysaccharide (LPS) in Yersinia revealed that Y.pestis is a relatively recently emerged clone (ca. 1500-20000 years) of Y. pseudotuberculosis serotype O1b (Achtman et al., 1999; Skurnik et al., 2000). The high virulence and recent evolution have made Y.pestis a model

organism for rapid evolution of a bacterial pathogen (Wren et al., 2003) The reservoir of Y.pestis in nature is wild rodents. Humans are accidental hosts and have no role in its long-term survival in endemic regions (Perry & Fetherston 1997). Upon feeding on blood of infected animals, fleas acquire Y. pestis, which multiplies in the midgut (stomach) and blocks the proventriculus, a sphincterlike organ that connects the stomach with the esophagus. The blocked fleas starve and frenetically bite other rodents and, incidentally, also humans, especially when the rodent population is wiped out by the epidemic (Duplantier et al., 2005). Upon biting, the blocked fleas are unable to swallow; instead, the blood becomes contaminated with bacteria from the flea’s proventriculus and is regurgitated into the new host. The bacteria are injected into the subcutaneous tissue, where they promote local proteolysis at the infection site and migrate through the subcutaneous tissue to the lymph channels (Sebbane et al., 2006). The bacterium quickly disseminates to regional lymph nodes and multiplies causing swelling in these organs. Upon contact with immune cells in the lymph nodes, most of the bacterial cells are killed by polymorphonuclear leukocytes (PMNs) (Titball et al., 2003; Zhou et al., 2006). Some of the bacteria are internalized by tissue macrophages, where they survive, multiply and produce virulence factors (Cavanaugh et al., 1959; Straley & Harmon, 1984). Y. pestis kills the macrophage by inducing apoptosis (Sebbane et al., 2005) and is released into the extracellular environment. In this second stage, Y. pestis is able to resist phagocytosis by the PMNs. The swollen lymph nodes are

1

Introduction

a hallmark of this disease and are known as bubo, hence the name bubonic plague. The disease proceeds to systemic spread of the bacteria via bloodstream (septicemic plague) to the liver, spleen and lungs (pneumonic plague). In a few cases, the systemic infection is achieved by Y. pestis with no commitment of the lymph nodes. This results from direct injection of the bacteria into the blood vessels (primary septicemic plague) (Sebanne et al., 2006). The bubonic and (primary) septicemic plague both progress to the death of the host, which involves massive septicemia and hemorrhagic pneumonia. If untreated, the mortality rate for the bubonic form of the disease is 50%-60%, as the epidemic develops and the pneumonic form becomes more prevalent, the mortality rate can reach 100%. Paradoxically for the bacteria, high bacteremia leads to death of its host and hence to the end of its food supply. On the other hand, high bacteremia also is essential to ensure effective infection of new fleas and transmission of the disease (Perry & Fetherston, 1997). The life cycle of Y. pestis requires that it adapts to a broad range of environmental conditions, including temperatures o varying from 4 C (hibernating animals) to 41oC (feverish patients) (Naylor et al., 1961; Perry & Featherstone, 1997). Furthermore, Y. pestis has to cope with innate immune factors in the flea vectors as well as the innate and acquired immune defenses in the mammalian hosts. For survival of Y. pestis in the flea, two factors are essential: the haemin storage system (Hms) (Hinnebusch et al., 1996; Jones et al., 1999) and the Yersinia murine toxin (Ymt) (Hinnebusch et al., 2002). The expression of these genes is dependent on temperature, being higher at 26oC than at 37oC. Both factors are involved in the colonization and blockage of the flea. 2

Hms- strains are unable to colonize the proventriculus and promote blockage of the flea, but are still fully virulent when inoculated in mouse (Lillard Jr et al., 1999; reviewed in Zhou et al., 2005). Likewise, Ymt- strains of Y. pestis are quickly eliminated from the flea midgut and unable to block the flea (Hinnebusch et al., 2002). In the human host, the conditions are different from those in the flea, e.g. the temperature is higher and an immune response becomes effective. In the infection site, a fraction of the bacterial cells are engulfed by macrophages while others cross the underlying tissue with the help of the plasminogen activator Pla, reach the lymphatic system and colonize the regional lymph nodes (see chapter 1.6). Y. pestis can survive and replicate in macrophages (Straley & Harmon, 1984; Pujol & Bliska 2003) and is thus considered a facultative intracellular pathogen (Perry & Fetherston, 1997). Genes involved in LPS modifications and resistance to antimicrobial peptides controlled by a PhoP/PhoQ two component system are important for survival of Y. pestis inside the macrophages (Oyston et al., 2000; Grabenstein et al., 2006). It is believed that resistance to macrophage killing is a form of protection for the bacterium, allowing it to produce virulence factors and adapt to the new host. It has been hypothesized that the macrophages also function as “carriers”, taking the bacteria from the infection site to the regional lymph node (Titball et al., 2003; Pujol & Bliska, 2005) but in vivo evidence to confirm this theory is lacking. Furthermore, Sebbane et al., (2005) failed to identify intracellular Y. pestis in the marginal sinus of the lymph nodes in a rat model of plague. When released from the macrophages, the bacterium is already adapted to the new conditions in the human host and fully “armed” with

Introduction

virulence factors that confer resistance to phagocytosis. These virulence factors include: (i) the Yersinia outer membrane proteins (YOPs) virulon, a complete type three secretion system (TTSS) (Cornelis et al., 2000; Straley et al., 1993), (ii) the fraction 1 (F1) antigen (Du et al., 2002), and (iii) the pH6 antigen (Makoveichuk et al., 2003), which in concert act against phagocytosis. In contrast to the zoonotic Y. pestis, Salmonella enterica is an intracellular oral/fecal pathogen that infects millions of humans annually. Salmonellosis is a worldwide public health problem that can result in a self-limited gastroenteritis [S. enterica serovar (sv.) Typhimurium] as well as in the life-threatening disease typhoid fever [Typhi]. In more than 95% of the cases, the infection initiates as the host ingests food contaminated with Salmonella cells, which pass the gastrointestinal tract and reach the small intestine (Hohmann, 2001). S. enterica produces a number of fimbrial adhesins, autotransporters and outer membrane proteins that promote attachment to the intestinal tract (Humphries et al., 2001). Following the adhesion and colonization of the intestinal tract, the infection proceeds by bacterial invasion into epithelial cells and M-cells in Peyer’s patches lining the intestinal mucosa. The Salmonella pathogenicity island 1 (SP1) (Groisman et al., 1997) encodes a TTSS that injects effector proteins into the cytoplasm of the epithelial cells, where they promote actin rearrangement and membrane ruffling, which lead to bacterial internalization (Ohl & Miller, 2001). Upon crossing the intestinal barrier, Salmonella cells face the macrophages and dendritic cells in the underlying tissue, then again, the TTSS encoded in SP1 comes to action. Salmonella induces macropinocytosis in

resting macrophages, invades the cells and multiplies in specific vacuoles known as Salmonella containing vacuoles (SCV) (reviewed in Gorvell & Meresse, 2001). In activated macrophages, Salmonella induces apoptosis and cell death (Knodler & Finlay, 2001; Monack et al., 2004). The ability to survive and multiply in macrophages and dendritic cells is a major determinant of pathogenesis and a variety of virulence factors involved in this process have been identified in S. enterica (Groisman et al., 1997). Transcriptome analyses indicated huge differences in gene expression of S.enterica sv. Typhimurium in infected mouse macrophages compared to in vitro cultivated bacteria (Eriksson et al., 2003). A number of genes encoding surface structures are downregulated, these include genes for biosynthesis of LPS, adhesins, flagella whereas the gene encoding the outer membrane protease PgtE (see chapter below) is upregulated. The genes in the Salmonella pathogenicity island 2 (SP2) encode a second TTSS that tampers with the normal intracellular trafficking and inhibits the formation of the phagolysosome (Gorvel & Meresse, 2001). Salmonella cells are released from the phagocytes into the bloodstream and may disseminate extracellularly to infect new macrophages and establish new infection foci (Mastroeni & Shepphard, 2004). Also, migration of infected phagocytes to the liver, spleen, bone marrow and gallbladder probably facilitates dissemination of bacteria in the host (Ohl & Miller, 2001). 1.2. Surface proteolysis in bacterial infections Bacterial-promoted proteolysis has long been recognized as important during infection. An obvious function is acquisition of nutrients from the 3

Introduction

host, but bacterial proteases are also directly involved in evasion of host immune response and bacterial migration (reviewed in Travis 1995; Goguen, 1995). Prevention of host response as well as phagocytosis and complement-mediated killing is an effective strategy to enhance bacterial virulence and few surface-bound or extracellular proteases have been described in this context. Streptococcus pyogenes (group A Streptococcus, GAS) and Streptococcus agalactiae (group B Streptococcus, GBS) express the cell-wall protein C5a peptidase (SCP), a protease that inactivates the complement chemotactic factor C5a and reduces PMN recruitment (Cheng et al., 2002). The enzyme elastase produced by Pseudomonas aeruginosa interferes with the host’s immune response by degrading various plasma proteins, such as immunoglobulin, complement factors, and α-protease inhibitor (Matsumoto et al., 2004). Y. pestis expresses the plasminogen activator Pla which degrades the complement protein C3 and inhibits chemotaxis of phagocytic cells (see below) The use of proteases to promote migration and disseminate across tissue barriers is a very effective strategy employed by bacterial pathogens (Travis et al., 1995). The tissue barriers consist mainly of collagen and laminin networks in basement membranes (BMs) or extracellular matrix (ECM) (Hay, 1991). The BM is a specialized extracellular matrix that forms a continuous sheet that separates the cells from connective tissue. Its most abundant components are collagen IV, laminin, nidogen or entactin and heparin sulfate (Hay, 1991). A few bacterial species are able to attack these structures directly (Harrington, 1996). Oral pathogens such as Porphyromonas gingivalis and Fusobacterium nucleatum 4

produce collagenases that degrade tissues in the dent-epithelial junction (Potempa et al., 2000), Clostridium perfringens and Clostridium histolyticum produce an array of proteases and collagenases that cause tissue necrosis in the peritoneum and enhance bacterial access to deeper tissues (reviewed in Harrington, 1996). An alternative way to promote surface proteolysis and cell migration is to engage the host proteolytic systems. Several bacterial pathogens are able to interfere with the plasminogen (Plg) system to promote tissue degradation and cell migration in vitro (Lähteenmäki et al., 2005). Few bacterial species express a direct Plg activator activity (PA), Staphylococcus aureus and streptococci (groups A, C and G) secrete staphylokinase and streptokinase, which activate Plg through a complex interaction (Lähteenmäki et al., 2001), Y. pestis and S. enterica express the surface-bound proteases Pla and PgtE which cleave and activate Plg. A number of bacterial pathogens express Plg receptors, i.e. are able to bind Plg on their surface, and utilize the host PAs for plasmin formation, in essence thus turning themselves into proteolytic organisms using a host proteolytic system. A Plg receptor is important for virulence and tissue spread of Borrelia burgdorferi, which expresses a 70 kDa surface protein that binds Plg. In addition, the outer membrane lipoprotein OspA is able to upregulate the expression of the human Plg activator urokinase (uPA) in human monocytes (Fuchs et al., 1996). The subsequent conversion of Plg into plasmin on the bacterial surface enhances bacterial penetration through endothelial cell monolayer grown on connective tissue substrates and is though to promote bacterial migration from the infection site to the circulation and secondary organs (Coleman et al., 1999). Group

Introduction

A streptococci express the surface Plgbinding group A streptococcal M protein (PAM) which binds with high affinity to Plg. The Plg immobilized by PAM subsequently interacts with streptokinase and generates bacterium-bound plasmin activity (Ringdahl et al., 1998) that has been shown to be an important virulence factor in experimental mouse model for impetigo (Svensson et al., 2002). Plg is proteolytically activated to plasmin that degrades circulating fibrin clots, components of the ECM/BM and activates pro-matrix metalloproteinase (MMPs; gelatinases). Some pathogens inactivate regulatory proteins of the Plg system and by this way enhance uncontrolled proteolysis (Grenier, 1996; Kukkonen et al., 2001). Together, the uncontrolled plasmin activity may lead to dissolution or damage of fibrin clots or collagen and laminin networks in subepithelial tissues, which both promote cell migration (Korhonen & Kukkonen, 2004). 1.3. β-barrel outer membrane proteins In contrast with the wide range of possible folding patterns found for water-soluble proteins, integral membrane proteins in the hydrophobic environment of the lipid bilayer are much less variable and predicted to have in their membrane-embedded regions only regular secondary structure elements, i.e. α-helices and β-sheets. Due to physical constraints imposed by the nonpolar core of the lipid bilayer, the βsheets within the bacterial outer membrane (OM) often are in β-barrels (although βbarrels are not exclusively found in OM) (reviewed in Koebnik et al., 2000). β-barrel proteins also are found in the membrane of mitochondria and chloroplasts, as well as yeasts, such as Saccharomyces cerevisiae

(Wimley, 2003). The membranes of Grampositive bacteria and archaea do not appear to contain β-barrels, but prokaryotic species from several bacterial families that do not fit into classical Gram-staining classification, such as Deinococcus, Chlamydia and Cyanobacteria, Treponema and Mycobacteria, possess βbarrel transmembrane proteins (Yen et al., 2002). Several functions are described for β-barrel OMPs: nonspecific porins, TonB dependent receptors, protein secretion pores, adhesins, proteases, lipases and pore-forming toxins (reviewed in Wimley 2003), emphasizing their overall importance for bacterial cells. Schulz (2000) listed a set of principles to describe the features of known β-barrel structures. In short, all β-barrels are cylindrical, closed structures with an even number of antiparallel β strands connected at the periplasmic end by short turns and at the external barrel end by longer loops. The β-barrel surface contacting the membrane interior is mainly formed by hydrophobic residues and lined by two girdles of aromatic residues in the membrane interfaces. Folding and insertion of β-barrels in the outer membrane seems to be assisted by binding to LPS. In E. coli, OmpA is translocated across the cytoplasmic membrane to the periplasm kept in solution by binding to the chaperone Skip. The OmpA-Skip complex bind to free LPS in the periplasm and insertion occurs spontaneously (Kleinschmidt, 2003). Sequence similarity searches in the genome of Escherichia coli identified approximately 60 proteins annotated as known, “probable” or “putative” outer membrane proteins (OMPs), accounting for 1.5% of the genome (Wimley, 2003). Genomic studies using algorithms based on composition/architecture, found similar predictions for other Gram5

Introduction

negative species, 2.1% of the genome of Pseudomonas aeruginosa, 3.5% of the genome of Y. pestis and 2.5% of the genome of Haemophilus influenzae. These independent analyses indicate that 2-3% of the Gram-negative bacteria genome encode β-barrel OMPs (Wimley, 2002). One-third of these OMPs are annotated as hypothetical proteins or not annotated with a known function. 1.4. The omptin family of bacterial proteases The omptin family is composed of integral outer membrane proteins that have a similar β-barrel fold. So far, eleven members with

a high sequence similarity are included in the omptin family (Fig 1) in the MEROPS peptidase database (http://merops.sanger. ac.uk/), and the family is growing as new homologues are found in newly sequenced species. Omptins share ca. 50-70% sequence identity, mature forms of 290301 amino acids, the similar predicted βbarrel fold, and lack or a low content of cysteine (Kukkonen & Korhonen, 2004). Four members have been more extensively studied: Pla of Y. pestis, OmpT of E. coli, PgtE of S. enterica and SopA of Shigella flexneri. These omptins exhibit different substrate specificities and virulence roles. Most of the omptins genes are located in plasmids. These include the gene

Yersinia pestis Pla Erwinia pyrifoliaea Epo

Salmonella Enterica PgtE

Escherichia coli OmpT

Desulfotalea psychrophila Unassigned Mesorhizobium loti Unassigned

Escherichia coli OmpP

Shigella flexneri SopA Legionella pneumophila Vibrio fischeri Leo Unassigned

Yersinia pestis YcoA Yersinia pseudotuberculosis YcoB

Figure 1. Omptins phylogenetic tree. The protein sequences of the omptins deposited in Merops peptidase database (http://merops.sanger.ac.uk/) were compared using Clustal W (http://www. ebi.ac.uk/clustalw/index.html) and a cladogram tree was draw based on their sequence identities using the phylodraw software http://pearl.cs.pusan.ac.kr/phylodraw/).

6

Introduction

coding for Epo and for Pla (McDonough & Falkow, 1989; McGhee et al., 2002; Sodeinde & Goguen, 1989). The later is located in the 9.5kb plasmid pPCP1 along with the genes for the bacteriocin pesticin and the pesticin immunity protein (Sodeinde & Goguen 1989), whereas pgtE is a chromosomal gene (Guina et al., 2000). Pla, Epo and PgtE are in predicted structure more closely related to each other than to the other omptins and form a subfamily (Fig 1). It has been proposed that PgtE is a common ancestor of Pla (Sodeinde & Guoguen 1989) and pla, pgtE, and epo have been partners in horizontal gene transfer between Y. pestis, S. enterica,and E. pyrifoliaea (Kukkonen & Korhonen 2004). Recent analyses have shown that a relatively small number of mutations in the surface-exposed loops cause major changes in omptin functions (Kukkonen et al. 2004). The omptin family is a good example of how virulence factors can adapt to new functions and different lifestyle of bacterial pathogens or commensals. 1.5. Structure of omptins The structure of OmpT was solved by crystallography to a resolution of 2.6Å (Vandeputte-Rutten et al., 2001). OmpT is an antiparallel β-barrel formed by ten strands in a vase-like shape that span the outer membrane (Fig 2). OmpT has a height of 70Å. The β-sheets have on the average 23 residues and a tilt of 37-45o in relation to the barrel axis and are connected by five mobile extracellular loops (L1 to L5) and four short periplasmic turns (T1 to T4). The barrel extends ~40Å to the extracellular side of the outer membrane and the loops are positioned right above the LPS core region (Ferguson 1998 et al.; Kramer et

al., 2002), this feature is important for the omptins’ function because they interact with LPS (see chapter 1.7). Similarly to other β-barrels, two girdles of aromatic residues border the hydrophobic area of the protein embedded in the lipid bilayer and position the β-barrel in the membrane. The top of the protein has a circular diameter of ~32 Å which becomes elliptical in the central part to dimensions of ~13 Å x ~26 Å (Fig 2). Initial findings suggested that the omptins should be classified as serine proteases (Sugimura et al., 1988). OmpT is weakly inhibited by a few serine protease inhibitors (serpins), and studies using site-directed mutagenesis of conserved serine and histidine residues showed that substitution of Ser99 and His212 nearly completely abolish the protease activity of OmpT measured against chromogenic substrates (Kramer et al., 2000). This amino acid pair, together with an aspartic or a glutamic acid residue, could form a classical serine protease catalytic triad. The resolution of the crystal structure of OmpT, however, revealed that the large distance between the Ser99 and the His212-Asp210 residues and the lack of a nucleophilic residue closer to the HisAsp couple make the proposed catalytic triad unfeasible (Vandeputte-Rutten et al., 2001) (Fig 2D, E and F). The same study suggested that the catalytic site is actually formed by an Asp-Asp and an Asp-His dyad, which is conserved in all members of the family (Fig 2B). This is supported by the findings that substitution of the aspartate residues at positions 83, 85 or 210 reduced the activity of OmpT to 0,01% (Kramer et al., 2001). Mutation of the corresponding residues in the aspartic catalytic dyad of Pla of Yersinia pestis lead to a similar effect, abrogating

7

Introduction

the protease activity measured with Plg, the biologically important polypeptide substrate (Kukkonen et al., 2001). These findings led to the reclassification of the family as aspartatic proteases (Kramer et al., 2001), where the nucleophilic attack is made by a water molecule stabilized in the

A

B L3

L2

L5 K262

L1

Asp-Asp couple and activated by the AspHis couple (Dunn et al., 2002). The proper in vitro reactivation of purified OmpT, Pla and PgtE (Kramer et al., 2002; Kukkonen et al., 2004) is dependent on the presence of LPS. The effect of LPS on the in vitro folding of bacterial outer membrane proteins has C

L3 L2

L1

L4

L3

L2 L4

L5

L5

L1

L4

K226

R171

R138

R175

R171

R138

R138

OM

D

L1

E

L3

L2 D86

D206

D84

H208 L4 L5

L1

F

L3

L2

D83

D85

H212 D210

L2

L1

L3

D84

D86

H208 D206

L4

L4 L5

L5

Figure 2. Omptins model. The crystal structure of OmpT (B and E) (PDB accession number: 1I78) was used as a template to prepare the models of Pla (A and D) and PgtE (C and F). The models were drawn with the Swiss-Pdb viewer software (Guex and Peitsch, 1997; http://www. expasy.org/spdbv/). The loops (L1-L5) and residues important for LPS binding, auto-processing and the catalytic site are marked. The models show the side view (A to C) and the top view (D to F) of the omptin β-barrel. The black lines represent the position of the polar head groups of the lipids in the outer membrane (OM). A. The conserved LPS-binding motif residues R138 and R171 and the auto-processing site K262 in Pla are shown. B. The LPS binding motif R138, R175 and K226 in OmpT are shown. C. The LPS-binding motif R138 and R171 in PgtE are shown. D to F. The conserved catalytic residues are marked.

8

Introduction

been studied with PhoE of E. coli. When mixed with LPS micelles in the presence of Triton X-100 in vitro, PhoE presents differential migration in SDS-PAGE and resistance to trypsin, both characteristics of the folded protein (de Cock & Tomasen, 1996). Subsequently, de Cock et al., (1999) demonstrated with a range of LPS types, including smooth-type LPS and rough mutants with a core oligosaccharide of different length (Ra to Re) (Hoslt et al., 1993) that charges in the inner core region of LPS contribute to the high folding efficiency of PhoE in LPS/TritonX-100 mixtures. The crystallographic resolution of the folded structure of the iron transporter FhuA of E. coli in complex with LPS (Ferguson et al., 1998) added more insight in the role of LPS in OMP folding by characterizing an LPS-binding cluster of four basic amino acids (Lys306, Lys351, Arg382 and Lys428) (Ferguson et al., 2000). This three dimensional LPSbinding motif is partially present in the structure of OmpT in the residues Arg138, Arg175 and Lys226 (Vandeputte-Ruten et al., 2001) as well as in other members of the omptin family such as Pla (Arg138 and Arg171) and PgtE (Arg 138 and Arg171) (Kukkonen et al., 2004). Mutation of the two arginines in the LPS-binding motif of PgtE abolished the ability of the protease to activate plasminogen (Kukkonen et al., 2004). The interaction of the LPS-binding motif with LPS is not fully understood, and the omptin-binding regions in the LPS molecule remain to be determined. In the FhuA–LPS complex, charged residues in the LPS-binding motif interact with the negatively charged phosphates of the inner core and the diglucosamine backbone of lipid A (Ferguson et al., 1998; Ferguson et al., 2000). Ferguson et al. (2000) identified the LPS-binding motif characterized in FhuA

in four proteins that mediate LPS-induced immune response: the bactericidal/ permeability-increasing protein (BPI), lactoferrin, lysozyme and the Limulus antibacterial and anti-LPS factor (LALF), thus demonstrating that the LPS-binding motif is present in cells of both prokaryotic and eukaryotic origin. The Toll-like receptor 4 (TLR4) is the mammalian receptor for LPS, and together with the protein MD2 and CD14 forms a complex responsible for recognition of and the innate immune response to LPS (reviewed in Miller, 2005). An accessory protein known as the LPS binding protein (LBP) that belongs to the same family of proteins as BPI, is needed to convert oligomeric micelles of LPS into monomers for delivery to the LPS recognition complex (Zweiger et al., 2006). A cluster of positively charged amino-acids formed by one Arg and two Lys residues, were showed by mutagenesis to be responsible for the LPS-binding activity of LBP (Lamping et al., 1996). A similar pattern of basic aminoacids involved in LPS binding is found in the MD2 protein (Visintin et al., 2003). MD2 binds directly to LPS with high affinity, and then to TLR4 initiating the signaling cascade that leads to an immune response. Mutation in basic aminoacid clusters in MD2 demonstrated that electrostatic interactions also are important in LPS binding (Gruber, 2004). 1.6. Function of omptins Omptins have been identified in several Gram-negative bacterial species that infect humans or plants. Some of these species are aggressive pathogens (Y. pestis, S. enterica, S. flexneri, Legionella pneumophila), while some are opportunists (E. coli), and some are essentially saprotrophic (Vibrio fischeri); the functions of these omptins 9

Introduction

seem to be adapted to the life style of their host species. Minor variations at critical sites in surface loops and the interaction with other surface components may account for the divergence of functions displayed in the family (Kukkonen & Korhonen, 2004). Using synthetic tetrapeptides immobilized on a cellulose membrane support, Dekker et al. 2001 determined that OmpT preferentially cleaves between two basic amino acids in positions P1 and P1’ followed by and alanine in position P2’ but with no clear preference for any aminoacid in position P2 (Fig. 3). A negatively charged residue in position P2 or P2’ reduces the activity of OmpT towards polypeptide substrates. Controversially, the same study (Dekker et al., 2001) found that OmpT also cleaves tetrapeptides which have Ile, His, Ala, Phe, Pro, Leu, Met, Gln, Asn or Val in position P1’, whereas most of the OmpT natural substrates are cleaved between dibasic residues, except for plasminogen that is cleaved, although slowly, between Arg and Val (McCarter et al., 2004). Okuno et al. (2002) used recombinant proteins as substrate to study the effect of residues further away from the cleavage site and found that basic residues in positions P4 and P6 positively affect the activity of OmpT. McCarter et al.,

(2004) also demonstrated that amino acids residues further away from the cleavage site, especially in positions P2’, P3 and P4, are important in substrate recognition. A study with larger substrates displayed in bacteriophage and phage screening methods confirmed the overwhelming preference for OmpT cleavage between two basic residues in position P1 and P1’, with a small tolerance for Val and Gly in P1’ (McCarter et al., 2004). This is relevant in the case of Pla, where the main physiological substrate, Plg, is cleaved at an Arg-Val bond. The requirement of basic amino acids in position P1 and P1’ as well as lack of acidic residues in positions P2 and P2’ are consistent with the facts that omptins have a negatively charged catalytic groove (Vandeputte-Rutten et al., 2001). In relation to the role in virulence, the best studied omptin is Pla of Y. pestis. Pla is pivotal in the spread of Y. pestis from the subcutaneous site of infection to the regional lymph nodes and critical in the development of bubonic plague. In experimental mouse infections, strains devoid of the pPCP1 plasmid have a million-fold increase in the LD50 after subcutaneous injection (Brubaker et al., 1965). Isogenic pPCP1- mutants of Y. pestis were also less efficient in reaching

Cleavage site

P2 Pn…P3

No clear preference

P1

P1’

P2’

Lys/Arg

Lys/Arg* Val/Gly

Ala

P3’…Pn’

Figure 3. Amino acid specificity in OmpT cleavage site. P1 and P1’ are the amino acid residues immediately prior to and following the scissile bond and so forth (P2…Pn, P2’…Pn’). (*) In position P1’ Lys and Arg are the favourite substrates, while Val and Gly are well tolerated. Nomenclature according to Schechter & Bergen (1967)

10

Introduction

the local lymph nodes, liver and spleen of infected mice. This attenuation is not observable when the bacteria are injected intravenously, highlighting the importance of Pla in bacterial migration through subcutaneous tissue and to the lymphatic system (Brubaker et al., 1965; Sodeinde et al., 1992; Welkos et al., 1997). Further genetic analysis of pPCP1 showed that Pla is responsible for the effect of pPCP1 on virulence (Sodeinde et al., 1992; Welkos et al., 1997). A recent study showed that mice infected by fleas contaminated with Pla- Y. pestis strains failed to develop bubonic plague, but one third of the infected mice developed terminal plague through primary septicemia. In contrast, eight out of ten mice infected with the Pla+ strain progressed to terminal plague, from which six mice presented the bubonic plague form and two the septicemic form (Sebbane et al., 2006). Y. pestis colonizes the flea midgut and forms a biofilm that blocks their normal blood feeding (Hinnebusch et al., 2002). The blocked fleas’ repeated attempts to feed occasionally result in direct injection of bacteria into blood vessels, leading to primary septicemic plague with Pla- strains also. While the same phenomenon was observed with the Pla+ strain, its normal infection route was via the lymph nodes (Sebbane et al., 2006). The primary septicemia then is not dependent on Pla, which may account for the persistence of virulent Y. pestis strains devoid of Pla (Filippov et al., 1990) On the other hand, these results highlight the importance of Pla in pathogenesis of bubonic plague. The major attribute of Pla is the capacity to activate circulating human Plg into plasmin. Plg is a proenzyme found in high concentration (2μM) in human plasma, and also in other body fluids and in the extracellular matrices (Stephens

& Vaheri, 1993). Plg is converted into plasmin by a proteolytic cleavage at the single site Arg505-Val506 (Robbins et al., 1967). Plasmin is a broad-spectrum serine protease that degrades fibrin and plays a central role in the process of fibrinolysis (Collen & Lijnen, 1991). Plasmin also degrades non-collagenous components of the extracellular matrix (ECM) and activates procollagenases (Saksela, 1985) and participates in processes of tissue remodeling and cell migration (Vassalli et al., 1992). Metastatic tumor cells are known to bind plasmin on their surface in order to facilitate their dissemination through tissue barriers (Mignatti & Rifkin, 1993) In normal conditions, the mechanism of fibrinolysis is tightly regulated. The physiological activators of Plg are the tissue-type Plg activator (tPA) and urokinase (uPA) which are also present in the human plasma, but in a 100.000 fold lower concentration than Plg (Rijken & Sakharov, 2001). During fibrinolysis, fibrin clots bind Plg and tPA thus increasing the Plg to plasmin conversion rate; the formed plasmin cleaves a polypeptide chain preferentially after a lysine residue. Carboxyterminal lysine residues are thus generated in the fibrin network to provide additional binding sites for Plg and to accelerate the degradation process (Rijken & Sakharov, 2001). Binding to fibrin also causes a conformational change in the Plg molecule exposing the cleavage site and enhancing activation by tPA (Mangel et al., 1990). Plasmin activity is controlled by serpins, mainly α2-antiplasmin (α2AP) and, to a lesser degree, α2-macroglobulin (α2M), while tPA and uPA are inhibited by Plg activator inhibitor-1 (PAI-1) and Plg activator inhibitor-2 (PAI-2). Plasmin, on a fibrin surface, is relatively protected from inactivation by α2AP; in solution, 11

Introduction

however, α2AP binds to plasmin and reduces its half-life to 0,1s (reviewed in Booth, 1999). Pla also interferes with the regulation of the fibrinolytic system by inactivating α2AP (Kukkonen et al., 2001). These two features of Pla, in concert with its adhesive characteristics (see below), promote uncontrolled proteolysis as well as damage of tissue barriers at the infection site (Lähteenmäki et al., 1998). Another proteolytic target of Pla is the serum complement protein C3 (Sodeinde et al., 1992) that has an essential role in bacterial opsonization, in the formation of the C5a fragment involved in chemotaxis of phagocytic cells and in the formation of the membrane attack complex, also known as MAC (Roitt & Delves, Essential immunology, eleventh edition). Pla cleaves the C3 fragment, and histological examination of subcutaneous infection sites in mice showed that lesions containing Pla-positive strains contained less inflammatory cells than did the lesions with Pla-negative bacteria (Sodeinde et al., 1992). Pla-negative isogenic strains showed resistance to high concentration (90%) of human serum (Sodeinde et al., 1992), and Welkos et al. (1997) could not find any differences in the inflammatory response to Pla positive and negative isogenic strains. These results show that the major role of Pla in plague pathogenesis is not serum resistance, and the importance of the Pla-mediated cleavage of complement remains to be elucidated. Pla also degrades Yersinia outer membrane proteins (YOPs) (Sodeinde et al., 1988) which are important virulence factors and confer resistance of Yersinia to phagocytic killing (Viboud & Bliska, 2005). The degradation of YOPs has been demonstrated in vitro (Sodeinde et al., 1988), Y. pestis cells, however,

12

are capable of synthesizing intact YOPs in vivo in guinea pigs (Skurnik et al., 1985) and deliver YOPs into mammalian cells (Skrzypek et al., 1998) and the significance of YOPs degradation remains to be established. Pla is implicated in bacterial adhesion to components of the ECM and the BM (Lähteenmäki et al., 1998). Plg, tPA as well as other components of the Plg system are present in BM (McGuire & Seeds, 1989). BMs are involved in various cellular functions such as adhesion, migration, differentiation, proliferation and apoptosis (Aumailley & Gayraud, 1998). Recombinant E. coli expressing Pla adhere to Matrigel (Lähteenmäki et al., 1998), a biologically active BM-like preparation extracted from EngelbrethHolm-Swarm mouse sarcoma (Kleinman et al., 1982). The bacteria also adhere to ECM preparation from the human mucoepidermoid carcinoma cell line NCI-H292 (Lähteenmäki et al., 2001). Wild type Y. pestis binds very efficiently to laminin and Matrigel while a Y. pestis strain cured of the pPCP1 plasmid adheres poorly. Both strains were found to bind to laminin, collagen type IV and Matrigel with low affinity, suggesting that adhesin(s) other than Pla may be involved in these processes (Lähteenmäki et al., 1998). An earlier study had reported that Pla mediates bacterial adhesion to mouse collagen type IV (Kienle et al., 1992), but this could not be repeated with human collagen IV (Lähteenmäki et al., 1998). Taken together, the adhesiveness to BM and the Plg activation by Pla create localized proteolysis that promotes damage of tissue barriers and bacterial migration (Kukkonen & Korhonen, 2004), and such damage has indeed been experimentally demonstrated in vitro (Lähteenmäki et al., 2005).

Introduction

Besides the adhesive function, a role of Pla in bacterial invasion into eukaryotic cells has also been observed. Cowan et al. (2000) reported that Y. pestis invades HeLa cells, derived from a cervical cancer, in the absence of fetal calf serum. The invasion was associated with the presence of the pPCP1 plasmid. Further studies found that Pla mediates invasion of recombinant E. coli to human endothelial cells (HUVECs, human umbilical vein endothelial cells) and the endothelial-like cell type ECV304 (Lähteenmäki et al., 2001). Pla may provide an intracellular route for bacterial dissemination across the endothelial barrier of lymphatic vessels or the lung epithelium. The second member of the omptin family investigated in this thesis work is PgtE of S. enterica (Guina et al., 2000). PgtE shares 75% amino acid sequence identity with Pla, but their activities are different. Kukkonen et al., 2004 showed that PgtE activates plg and this activity is dependent on the absence of O-specific chain (also known as the O-antigen). S. enterica cells grown in laboratory medium express smooth LPS type (see chapter 1.7.2), and PgtE apparently is inactive. This accounts for the fact that earlier studies with wild type isolates of S. enterica presented no or very low activity towards Plg (Sodeinde & Goguen, 1998). The expression of PgtE and the shortening of the O-specific chain in LPS are both controlled by the PhoP/PhoQ system (reviewed in Groisman, 2001). The pgtE gene is one of the most highly transcribed genes of S. enterica sv. Typhimurium within murine macrophages (Eriksson et al., 2003). Accordingly, S. enterica sv. Typhimurium cells isolated from SCV of mouse macrophages have a high PgtE activity and efficiently inactivate α2AP and activate Plg (Lähteenmäki et

al., 2005a). Together, these two studies suggest that PgtE on the surface of S. enterica in macrophages is fully active and could contribute to bacterial spread and survival in vivo (Lähteenmäki et al., 2005a). S. enterica sv Typhimurium also has Plg receptors, and the bound Plg can be activated by tPA, and the bound plasmin degraded basement membrane in vitro (Lähteenmäki et al., 1995). PgtE was not expressed in these test conditions but, in principle, can enhance the efficiency of the process and increase damage of tissue barriers. PgtE also confers resistance to cationic antimicrobial peptides (CAMPs), Guina et al (2000) showed that PgtE cleaves an α-helical antimicrobial peptide and strains that had the pgtE gene deleted are more sensitive to CAMPs. They suggested that the proteolytic activity of PgtE within macrophages protects the bacteria by degrading CAMPs, thus lowering the concentration of these molecules in the SCVs and improving survival of intracellular bacteria (Guina et al., 2000). Recently it has been demonstrated that CAMPs also directly activate the PhoP/PhoQ system (Bader et al., 2005) and consequently upregulate the expression of PgtE. PgtE is a poorer adhesin to laminin and BMs than Pla and does not enhance bacterial invasion into human endotheliallike cells (Kukkonen & Korhonen, 2004) 1.7. Lipopolysaccharide Lipopolysaccharides (LPS) are found in the cell envelope of Gram-negative bacteria and are the major components of the outer leaflet of their outer membrane. Typically, an LPS molecule consists of three structural domains: the lipid A, which holds the endotoxic biological activity; a non-repeating oligosaccharide core; and a 13

Introduction

polysaccharide, the O-specific chain, also known as O-antigen (Raetz & Whitfield, 2002). Lipid A is the hydrophobic anchor for the incorporation of LPS in the outer membrane and is composed of a glucosamine disaccharide (β-Dglucosaminyl-(1→6)-D-glucosamine), which is conserved in most Gram-negative species, (for a review Seltmann & Holst, 2001). with long-chain 3-hydroxy fatty acids substitutions in positions 2, 3, 2’ and 3’ and with phosphates in positions 1 and 4’. The acylation pattern, including the type, number, length and position of the fatty acid chains bound in lipid A vary according to the bacterial species (Seltmann & Holst, 2001). The phosphates can be modified by the attachment of amino compounds, including 2-aminoethanol and 4-amino-4-deoxy-L-arabinose (Ara4N), and this modification is related to increased resistance to cationic antimicrobial peptides (Nummila et al., 1995; Raetz & Whitfield, 2002). Lipid A is responsible for the majority of the immunomodulatory activities of LPS. Many different eukaryotic cell types recognize LPS in subnanogram quantities and respond almost immediately by producing cytokines, chemokines, and cellular adhesion molecules (Darveau, 1998). High bacteremia with a Gramnegative pathogen often leads to endotoxic shock, a major clinical problem caused by the host innate inflammatory response to the LPS released in the bloodstream (reviewed in Sharma & Dellinger, 2006). The core oligosaccharide connects the lipid A to the O-specific chain and can be further divided in the inner core (IC) and the outer core (OC). The IC in Enterobacteriaceae has a conserved backbone structure composed of two

14

monosaccharides, 3-deoxy-D-mannooct-2-ulosonic acid (Kdo) or, in some cases, its 3-hydroxy derivative Dglycero-D-talo-octulosonic acid (Ko) and L-glycero-D-manno-heptopyranose (Hep) (Frirdich & Whitfield, 2005). The basic structure can be modified through the addition of rhamnose (Rha), galactose (Gal), glucosamine (GlcN), Nacetylglucosamine (GlcNAc), Kdo, Ko, phosphate and phosphorylethanolamine (PEtN) residues (Frirdich & Whitfield, 2005). The OC exhibits more variation and is composed of more ordinary hexoses and hexosamines, in contrast with the less common sugars such as Kdo and Hep. The outer core is also the docking site for the O-specific chain (Wilkinson, 1996). The O-specific chain is composed of repeating oligosaccharidic units, and its structure is remarkably diverse, more than 60 monosaccharides and 30 different noncarbohydrate components have been identified in different bacterial species (Raetz & Whitfield, 2002). The repeating units are the key to the serological behavior of the bacteria, and the chemical diversity of the O-specific chain confers specificity to an organism (serotype) (Seltmann & Holst, 2001). The O-specific chain of LPS is considered important for bacterial pathogenesis in humans, since it blocks the access of complement proteins to the bacterial surface and hampers the formation of the MAC, and bacteria lacking O-specific chain are generally killed by serum complement (Rautemaa & Meri, 1999). LPS molecules containing the O-specific chain are termed smooth LPS (S-LPS) while the ones lacking O-antigen are known as rough LPS (R-LPS).

Introduction

1.7.1. Characteristics of the LPS of Y.pestis The lack of O-antigen expression in Y. pestis results from mutations in five genes of the gene cluster encoding the LPS biosynthetic pathway, thus Y. pestis has a rough type (R-LPS) (Skurnik et al., 2000). The expression of a heterologous O-antigen from Y. enterocolitica in Y. pestis had no effect in virulence after intravenous challenge in mice, neither an effect in resistance to serum complementmediated killing was seen (Oyston et al., 2002). It was postulated that Y. pestis developed other virulence mechanisms to compensate the absence of the O-antigen, such as the expression of the F1 capsule, the C3-degrading activity of Pla and a less endotoxic LPS, (see chapter 1.1 and next paragraph). Recently, it was demonstrated that Y. pestis binds the C4b binding protein (C4bp) from human, rabbit, guinea pig, rat, and mouse sera (Ngampasutadol et al., 2005), which confers resistance to complement killing. Many phenotypic properties are regulated by temperature in Y. pestis, and the transition point, i.e. up- or downregulation of the genes in question, is around 26oC (Straley & Perry, 1995). Significant structural variations can be found in the LPS of Y pestis in response to external factors. A shift in temperature from 21oC/27oC to 37oC, which represents the transmission from the flea to the human host, induces alterations in the number and type of acyl chains in the lipid A (Knirel et al., 2005; Kawahara et al., 2002; Rebeil et al., 2004). The predominant form of lipid A in cells grown at 37oC is tetraacylated with smaller amounts of the pentaacylated form. At lower (27oC and 25oC) temperatures, the picture is more complicated, with a mixture of tetra-, penta- and hexaacylated

forms (Kawahara et al., 2002; Knirel et al., 2005b). In cells grown at 21oC, the main lipid A type found is hexaacylated (Rebeil et al., 2004). The acylation level of lipid A affects the aggregation and the biological activity of LPS (Seltmann & Holst, 2001). The ability to stimulate the production of TNF-α in human macrophage cell lines can be used to evaluate the endotoxicity of LPS (Yao et al., 1997). LPS isolated from Y. pestis cells grown at 21oC (Rebeil et al., 2004) and 27oC (Kawahara et al., 2002) stimulates the secretion of TNF-α, whereas LPS from bacteria grown at 37oC is a poor stimulator. The expression of LPS with low immunostimulatory characteristics is an obvious advantage for Y. pestis in the mammal hosts. The core oligosaccharide of the LPS is also affected by the growth temperature, at 37oC Y. pestis ssp pestis has Kdo and Hep as the core terminal residues, while at 25oC a mixture containing all four possible combinations of residues is found (Kdo/Ko and Hep/Gal). Recently it was demonstrated that mutant Y. pestis strains which completely lack the outer core are still viable (Tan & Darby, 2005), but the effect of this mutation in the virulence of Y. pestis has not been reported yet and direct evidence linking changes in the sugar core of LPS to virulence are still lacking. It is interesting, however, that the LPS of Y. pestis ssp caucasica, which exhibits reduced virulence in guinea pigs and humans, lacks Hep residues (Knirel et al., 2005b). The amount of Ara4N in the lipid A is higher in cells grown at 27oC and 25oC than at 37oC (Kawahara 2002; Knirel 2005b). The modification of the LPS of Y. pestis with Ara4N appears to be controlled by the PhoP-PhoQ system. A phoP negative strain of Y. pestis failed to incorporate Ara4N in LPS at both 21oC and 37oC 15

Introduction

(Rebeil et al., 2004). Indeed, microarray analyses of the Y. pestis transcriptome revealed that the genes for biosynthesis and addition of Ara4N to lipid A are regulated by the PhoP-PhoQ system (Zhou et al., 2005). More recently, it was described that expression of the biosynthetic genes pbgP and ugd responsible for the addition of Ara4N to the LPS of Y. pestis can also be controlled by the PmrA/PmrB system which, independently of the PhoP-PhoQ system, is activated by high concentration of Fe3+, indicating that these genes are regulated by two separated promoters (Winfield et al., 2005). A Y. pestis phoP mutant was highly sensitive to polymyxin B and to cationic antimicrobial peptides that are abundant within macrophage phagosomes, an environment depleted of Mg2+ (Hitchen et al., 2002), and presented a reduced ability to survive in J774 macrophage cell cultures (Oyston et al., 2000). Furthermore, the modification of lipid A with Ara4N may help the bacteria cope with low Mg2+ stress by making the Mg2+ present in the LPS available to other compartments in the bacterial cell (Groisman, 2001). These data suggests that LPS modifications mediated by the PhoP-PhoQ system are important for the bacterial survival in the intracellular environment of macrophages and may help the bacteria deal with the host defenses of the flea (Rebeil et al., 2004). 1.7.2. Characteristics of the LPS of S.enterica Salmonella has an LPS typical of the family Enterobacteriaceae presenting an O-specific chain that confers major

16

antigenic variability to the cell surface. S. enterica exhibits 46 serogroups that differ in the O-antigen (Popoff et al., 1997). As mentioned earlier in this text, expression of O-antigen is related to resistance to phagocytosis and serum-mediated killing (Skurnik & Toivanen, 1993). In mouse infection models, Typhimurium mutants expressing LPS devoid of the O-antigen display a severe reduction in virulence (Thomsen et al., 2003). On the other hand, Typhimurium downregulates genes for synthesis of LPS and reduces the length of the O-specific chain inside macrophage-like cells (Eriksson et al., 2003, Lähteenmäki et al., 2005a), which add to the high activity of PgtE observed in cells from SCVs (Lähteenmäki et al., 2005a). Mutant strains devoid of the Oantigen induced production of lower amounts of nitric oxide and were able to grow faster in these macrophage-like cells as compared with wild-type bacteria (Bjur et al., 2006) S. enterica has a PhoP-PhoQ twocomponent system that regulates structural modifications of lipid A in a similar fashion as in Y. pestis (Guo et al., 1997). Activation of the PhoP-PhoQ system under low Mg2+ concentration or by antimicrobial peptides leads to increased acylation and addition of Ara4N to LPS (Guo et al., 1997; Bader et al., 2005). These modifications in LPS are associated with resistance to antimicrobial cationic peptides and likely to occur in the intracellular milieu. It then appears that both S. enterica and Y. pestis modify the composition of their LPS as the bacteria moves from an extracellular to an intracellular environment or from the flea to the mammalian host.

Aims of the Study

2. Aims of the study When this work begun, the virulence role of Pla of Y. pestis had been well established and putative virulence-associatedfunctions for PgtE of S. enterica had been identified. Pla and PgtE are transmembrane proteases/adhesins, and their functions had been studied in recombinant bacteria. An aim of this work was to develop functional studies with purified omptin proteins and

to ascertain the observed functions using the omptin proteins. The importance of LPS for Pla function had been observed (Kukkonen et al., 2004), and the first study of OmpT-LPS interaction had been reported (Kramer et al., 2002). A major aim in this work was to determine the structural aspects in LPS that affect Pla function.

17

Materials and Methods

3. Material and methods Bacterial strains and plasmids used in this study are listed in table 1 and 2. The methods used are listed in table 3 and are described in the original articles. The methods not found in the articles are described bellow. Table 1 - Bacterial strains used in this study

Bacterial strains

Characteristics

Used in

Reference

Y. pestis KIM D27

pPCP1+ pgm pYV+

Paper II

Une & Brubaker,

derivative of KIM-10 Y. pestis KIM D34

pPCP1- pgm pYV+

1984 Paper II

derivative of KIM-10

Une & Brubaker, 1984

S. enterica 14028

Smooth LPS

Paper III

ATCC

S. enterica 14028R

Rough LPS derivative of

Paper III

Wick et al., 1994

pgtE- derivative of 14028 Paper III

Kukkonen et al.,

14028 S. enterica 14028-1

2004 S. enterica 14028R-1

Rough LPS derivative of

Paper III

14028-1 (pgtE-) E. coli XL1 Blue MRF

Lähteenmäki et al., 2005a

R

endA1 gyrA96(nal ) thi-

Paper II, III

Stratagene

Paper I, II and

Studier &

III

Moffatt, 1986

1 recA1 relA1 lac glnV44 F'[ ::Tn10 proAB+ lacIq Δ(lacZ)M15] hsdR17(rKmK+) E. coli BL21

F– ompT gal dcm lon -

-

hsdSB(rB mB ) λ(DE3)

18

Materials and Methods

Table 2 - Plasmids used in this study

Plasmids

Characteristics

Used in

Reference

pSE380

Trc promoter, LacO

Paper I, II and

Invitrogen

operator, lacI

III

pla in Pse380

Paper III

pMRK1

Kukkonen et al., 2001

pMRK3

pgtE in Pse380

Paper III

Kukkonen et al., 2004

pMRK31

pgtE D206A in Pse380

Paper III

Kukkonen et al., 2004

pMRK34

pgtE R138E R171E in

Paper III

Pse380 pQE30

Kukkonen et al., 2004

T5 promoter, lac

Paper I, II and

operator, His6-tag

III

pQE30.1

pQE30 with Pla(R138E)

Paper II

This study

pQE30.2

pQE30 with Pla(R171E)

Paper II

This study

pQE30.3

pQE30 with

Paper II

This study

Qiagen

Pla(R138E/R171E) pQE30.4

pQE30 with Pla(D206A)

Paper II

This study

pREP4

lacI

Paper III

Qiagen

19

Materials and Methods

Table 3 - Methods

Method

Used in

Expression of recombinant omptin in E.

Papers I, II and III

coli Expression of recombinant His6-tagged

Papers I, II and III

omptin in E. coli Cultivation and Growth of Y. pestis

Thesis text

Cultivation and Growth of S. enterica

Paper III

Purification and refolding of His6-tagged

Papers I, II and III

omptin Reactivation of His6-tagged omptins with

Papers I, II and III

LPS Generation of point mutantions

Paper II

DNA sequencing

Paper II

Coating of fluorescent micro particles with

Paper I

His6-tagged omptin Adhesion of FMPs to immobilized

Paper I

substrates Plasminogen activation assays

Papers I, II and III

Plasmin inhibition by LPS

Paper II

Degradation of plasminogen

Paper II

Degradation of Gelatin and Collagen

Paper III

Degradation of DQ-gelatin

Paper III

Outer membrane extraction

Thesis text

20

Materials and Methods

3.1. Cultivation of Y. pestis Y. pestis strains were briefly stored on brain and heart infusion (BHI) plates and cultivated overnight in BHI broth for 3 consecutive passages at either 20oC or 37oC before each experiment. Bacterial growth was collected, washed once and resuspended in phosphate buffered saline pH 7.2 (PBS 7.2), the concentration was adjusted to 2x109 cells/ml by spectrophotometry (BioRad, Hercules, CA). 3.2. Extraction of outer membrane of Y. pestis Outer membrane proteins were extracted as described in Pouillot et al., 2005.

Briefly, IPTG-induced bacterial cultures were adjusted to 2x109 cells/ml in PBS and 1ml of each suspension was sonicated in 3 cycles of 10 seconds with 10 second interval over crushed ice. Intact cells were removed by centrifugation at 1800g for 5min. The supernatant was centrifuged again for 10min at 15000g. The pellets were resuspended in 30μl of sample buffer, separated by 12% SDS-PAGE and transferred to hybond nitrocellulose membranes (GE Healthcare). Pla protein or the LPS-binding mutant proteins were detected using anti-His6-Pla rabbit antisera (1:1000) produced in Medprobe (Norway) and alkaline phosphatase conjugated antirabbit IgG (1:5000) (Dako, Denmark) and exposed to phosphatase substrate.

21

Results and Discussion

4. Results and discussion 4.1. Reactivation of purified His6-Pla with E.coli K12 LPS and coating onto FMPs (I, III) The reactivation procedure of His6-Pla used in this thesis is based on the activation of purified OmpT described by Kramer et al., 2000b. These authors found that OmpT purified from inclusion bodies in recombinant E. coli adopts a folded state but remains enzymatic inactive in a buffer containing the detergent dodecyl-N,N’dimethyl-1-ammonio-3-propanesulfonate (DodMe2NPrSO3). Based on previous work done on refolding of the outer membrane phospholipase A of E. coli (Dekker et al., 1995), several conditions for the activation of OmpT were tested, and it was found that activation of OmpT required the presence of LPS. No major changes were found by spectroscopy in the secondary structure of OmpT after addition of LPS, which probably means that LPS induced only minor conformational changes in the highly mobile surface loops of OmpT (Brandenburg et al., 2005; Krammer et al., 2002). The crystal structure of OmpT (Vandeputte-Ruten et al., 2001) suggests that it is a monomer and that one molecule of LPS per an OmpT molecule is bound, but no in vitro binding data exists to support this conclusion. Kukkonen et al. (2004) showed that the recovery of Plg activation by Pla also is dependent on the presence of LPS, specifically R-LPS, while S-LPS was poor in the reactivation. The authors suggested that the long O-chain of S-LPS blocks the contact of the active site of the enzyme with Plg substrate. In this study, His6-Pla was isolated as inclusion bodies from E. coli BL21 (pREP4/pQE30) (Kukkonen et al., 2001) in denaturing conditions (8M urea) and 22

purified by affinity chromatography. To obtain activation, the protein was then extensively dialyzed against stepwise decreasing concentrations of urea in HEPES buffer (pH 7.2) (4M, 2M, 1M, 0,5M and buffer without urea) in the presence of 1mM of the detergent DodMe2NPrSO3 (Kramer et al., 2000; Kukkonen et al., 2001). The enzymatic activity of the purified His6-Pla was measured in an indirect assay, in which Pla cleaves plasminogen into the active form plasmin, which subsequently cleaves the chromogenic substrate SS-2251 that can be spectroscopicaly determined at 405nm (Kukkonen et al., 2001). Increasing amounts of K12 LPS (10pmol, 40pmol, 100pmol, 1nmol and 2nmol) were used to activate 200pmol of His6-Pla (Fig. 1 of II). Addition of LPS in amounts as low as 10pmol and 40pmol already lead to an increase in the activity of His6-Pla. The highest activity was achieved with 100pmol-1nmol of LPS, i.e. at nearly equimolar concentrations of His6-Pla and LPS. A higher LPS concentration of 2nmol did not further increase the activity, rather, the activity was lower (Fig. 1 of II). This was found to result from inhibition of plasmin activity and Plg cleavage by high amounts of LPS. No plasmin was formed by LPS-treated His6-Pla D206A mutant protein (Fig. 1A of II) which indicates that active Pla is needed. We also assessed the cleavage of the Plg molecule into the two-chain plasmin by His6-Pla using Western blotting (Fig. 1B of II). Plasmin formation was not seen without added LPS. This shows that LPS is needed for

Results and Discussion

the proteolysis and that this effect is not due to an incorrect cleavage of the Plg molecule. We observed the highest His6-Pla activation at nearly equimolar concentration of K-12 LPS. This is in agreement with the crystal structure of FhuA (Ferguson et al., 1998) and the modeling of the LPS-OmpT complex (Vandepute-Rutten et al., 2001), which show one molecule of LPS bound to a molecule of FhuA/OmpT. However we cannot conclude from our data that Pla binds only one LPS molecule. It is possible that both the Pla/DodMe2NPrSO3 complex as well as the LPS used here (Mukerjee et al., 1999) are in micelles and that the concentration of free LPS molecules is probably low. Kramer et al. (2002) concluded that, in their activation of OmpT, the LPS molecules formed mixed micelles with the detergent and OmpT, and it is reasonable to assume that a similar situation exists in the activation of His6-Pla. Our assay with Plg as a substrate is justified because of the high biological relevance, but adds complexity to the system because high concentrations of LPS were found to interfere with plasmin activity. This probably resulted from unspecific binding of LPS to the Plg/ plasmin molecule. Pla has been associated with bacterial adhesion to laminin (Lähteenmäki et al., 1998) and other cell or tissues targets (Lähteenmäki et al., 1998, Kienle et al., 1992). Purified OMPs are difficult to use in adhesion assays, and in this thesis work I coated activated His6-Pla onto fluorescent micro-particles (FMPs) and used them in binding studies (I). To assess the coating, I measured enzymatic activity of the FMPs (Fig. 1 of I). The His6-Pla-FMPs activated Plg, whereas no Plg activation was detected with beads coated with LPS (LPS-FMPs) or BSA (BSA-FMPs). The activity of 109

His6-Pla-coated particles was compared against the activity of the same amount of His6-Pla that was used to coat the FMPs. The activity of His6-Pla-FMPs was clearly lower than that of the soluble His6-Pla-LPS complex. This reduction may result from protein loss during the coating process. It is also possible that the coating of His6Pla onto FMPs leads to an unnatural conformation with reduced enzymatic activity or that the immobilization may hide the active site in Pla. 4.2. Adhesion of His6-Pla-FMPs to ECM proteins (I) Lähteenmäki et al., (1998) reported that recombinant E.coli expressing Pla adhere to Matrigel and laminin but not to collagen type I or IV (Lähteenmäki et al., 1998). Pla degrades YOPs (Sodeinde et al., 1998) and thus modifies the cell wall of Y. pestis (Kienle, 1992). It remained uncertain whether Pla is an adhesin or whether the adhesiveness results cell wall modifications by Pla. No major changes in the OMPs of E. coli K-12 strain LE392 expressing Pla were observed (Sodeinde et al., 1988, Lähteenmäki et al., 1998) which supported the role of Pla as an adhesin. However, a direct demonstration of adhesive properties in the Pla protein was lacking. His6-Pla-FMPs bound to Matrigel and to laminin but not to collagen IV- or BSA-coated slides (Figs. 2 and 3 of I). The His6-Pla-FMPs bound onto Matrigel and laminin substrate were variably aggregated (Fig. 2 of I), which complicated the assays. The quantitation of the FMPs binding was based on computer-assisted counting of the total fluorescent area on the target, which was divided by the size of one FMP. Whereas the His6-Pla-FMPs were highly adhesive, the control FMPs coated 23

Results and Discussion

with purified LPS or BSA showed only a low level of binding (Figs 2 and 3 of I), which indicated specificity of the system. The intrinsic adhesive characteristic of Pla was confirmed and the paper I demonstrated that an outer membrane βbarrel protein, refolded from its denatured state, can be coated to the surface of micro particles without loosing its enzymatic activity and adhesive properties. To best of my knowledge, this is the first report on coating the FMPs with an entire β-barrel protein molecule. 4.3. Reactivation of His6-Pla with different LPS chemotypes (II) LPS molecules are composed of regions (O-antigen, outer core, inner core and lipid A) which could be involved in the LPS-Pla interaction. Moreover, the modifications of their structures in response to environmental changes take place in Y. pestis (Hitchen et al., 2002; Knirel et al., 2005b). Here, a set of structurally distinct LPS/lipid A chemotypes were compared for reactivation of His6-Pla in an attempt to identify the regions of LPS involved in the interaction. It was previously reported (Kukkonen et al., 2004) that Pla expressed in smooth enterobacteria is functionally inactive, apparently due to a steric inhibition by the O-antigen. A similar difference in activation of His6-Pla was observed in the assays with purified LPS (Fig 2 of II). Particularly interesting is the behaviour of Y. pseudotuberculosis O1b and O1bΔwb bacteria (Kukkonen et al., 2004) and LPS (Fig 2 of II). In both cases, S-LPS is associated with low activation of Pla/His6-Pla. Y. pestis has evolved from Y. pseudotuberculosis O1b (Skurnik et al., 2000), and one of the advantages of the loss of the O-antigen could be the high activity of Pla. 24

The deep rough Re-LPS from E. coli caused a poor activation of His6Pla, while the corresponding Ra LPS from E. coli K12 was highly active (Fig 2 of II). Deletion of the wb region in Y. pseudotuberculosis O1b LPS results in a complete core region and mimics the LPS structure of Y. pestis (Skurnik et al., 2000). The Y. pestis KM218, which was efficient activator of His6-Pla, also has an outer core in LPS. The Y. pestis EV11M LPS is in structure similar to Re of E. coli (Knirel et al., 2005a) and lacks the outer core region and most of the inner core, and both were poor in enhancement of His6-Pla activation (Fig 2 of II). A marginal enhancement of His6-Pla activity was seen with hexaacyl lipid A (Fig 2 of II). Taken together, these results show that an LPS with a near complete outer core region is optimal for activation of His6-Pla. On the hand, the R-LPS from Yersinia and E. coli here tested differ in oligosaccharide structure and in substitution with phosphate, the results thus suggest that the presence, rather than specific sequence, of an outer core oligosaccharide is critical for Pla functions. We next assessed a possible effect of the LPS acylation on His6-Pla activation with Re and lipid A molecules that differ in degree of acylation (Fig 3 of II). The poor activation of His6-Pla by hexaacylated lipid A was slightly increased in the pentaand tetraacylated lipid As. The effect was more obvious between the hexa- and pentaacylated Re LPS molecules. Lipid A molecules lacking the phosphates in position 4’ or 1 of the disaccharide backbone failed to activate His6-Pla. The Y. pestis 1146 LPS was poor in enhancing the His6-Pla function (Fig 2 of II). This LPS differs from the other Y. pestis LPS in having a nearly complete substitution of both lipid A phosphates with Ara4N

Results and Discussion

(Knirel et al., 2005a). Its low activity indicates a role of lipid A phosphates. One of the common modifications in LPS of Y. pestis is the substitution of phosphates in lipid A with Ara4N (Raetz, 2001; Winfield et al., 2005). Our findings are in agreement with previous reports concerning the effect of the O-chain in Pla activity. Kukkonen et al. (2004) showed that the activities of Pla and PgtE are dependent on R-LPS but are inhibited by the long O-antigen of S-LPS, possibly by steric hindrance of the catalytic site. The results in paper II support this hypothesis and also indicate that the presence of an outer core in the LPS molecule is required for Pla activity. Similarly, Kramer et al. (2002) showed that deep rough Re LPS and lipid A are poor activators of OmpT. It was also reported that phosphates bound to the heptose sugars in the outer core are important for OmpT activity (Kramer et al., 2002). Differently from E. coli K-12 LPS, Y. pestis LPS does not contain phosphates in the outer core heptoses. We found that lipid A lacking either of the phosphates is less efficient in Pla activation. Differences between the interaction of Pla and OmpT with LPS may result from the fact that the LPS-binding motif of OmpT has one more positively charged amino acid (Lys 226) in the correct spatial conformation than the LPS-binding motif in Pla. Kramer et al., 2002 found that hexacylated LPS activates OmpT more efficiently than triacylated LPS. In contrast, we found that tetra- and pentacylated lipid A are more efficient to reactivate Pla than hexaacylated lipid A. At 37oC Y. pestis produces mainly tetraand pentaacylated lipid A, which are less immunogenic than the hexaacylated form that are prevalent at20-27oC. The supra-molecular structure of the tetra- and pentacylated LPS aggregates is lamellar,

while the hexacylated LPS is an inverted cubic structure (Brandenburg et al., 2003). The LPS in the outer leaflet of the outer membrane is believed to be in a lamellar state (de Cock et al., 1999) which could explain the preference of Pla for this structure. 4.4. Effect of phosphate and NaCl on His6-Pla activity (III) The results suggested that lipid A phosphates have a role in Pla-LPS interaction. We further addressed this question by testing the effect of phosphate and salt on enhancement of His6-Pla activity. Their effects were dosedependent, and 50mM of phosphate or NaCl completely abolishes the activity (Fig. 4 of II). In control assays, the same concentration of phosphate or NaCl had no effect on the enzymatic activity of plasmin, showing that the inhibition occurred during activation of Plg by Pla. These results are similar to those detected by Kramer et al., 2002 in activation of OmpT and show the importance of ionic forces in OmpT-LPS interaction. 4.5. Effect of mutations in the LPSbinding motif of Pla on plasminogen activation (III) To further analyze the role of lipid A phosphates in the reactivation of Pla, we substituted the two arginines (138 and 171) in the lipid A-binding motif either separately or in combination. These arginines are oriented outwards in the Pla barrel (See Fig. 1) and form a part of the three-dimensional conserved motif detected in several LPS-binding prokaryotic and eukaryotic proteins (Ferguson et al., 2000). Kukkonen et al. (2001) showed that PgtE activity is abolished upon mutation 25

Results and Discussion

of the two arginines in the LPS-binding motif. These positively charged amino acids interact with negatively charged phosphates in lipid A, (Ferguson, et al., 2000; Kukkonen et al., 2001). The substitution R171E considerably diminished the activity of Pla reconstituted with Y. pestis KM 218 LPS (Fig. 5A of II). The substitutions in Arg 138 and the double substitution in R138E R171E abolished the PA activity of His6-Pla (Fig. 5A of II). We also addressed the cleavage of Plg the proteins using Western blot (Fig. 5B of II). Activated His6-Pla cleaved Plg almost completely after 2h incubation, and also the His6-Pla without reactivation with LPS slowly degraded Plg. The LPSbinding mutants did not cleave Plg in the absence of LPS and the activity of the reactivated proteins remained very poor (Fig. 5B of II). The cleavage assay (in Fig. 5B of II) is less sensitive than the activity assay measurement (Fig. 5A of II), which probably explains the lack of detectable cleavage by His6-Pla R171E (lanes d and e, fig. 5B of II) We analyzed the folding of His6-Pla and its derivatives by Western blotting and found the three forms of the protein (Fig. 5C of II). α-Pla contains the unprocessed and the mature forms of Pla, β-Pla is the autoprocessed form, γ-Pla is believed to be an alternative folding form of the αPla (Kukkonen et al., 2001). Boiling of the samples in SDS-containing buffer resulted in loss of γ-Pla (Fig. 5C of II). The mutated His6-Pla proteins R138E and R171E formed much less β-Pla than did the wild type His6-Pla. The double His6-Pla mutant protein R138E R171E was unable to form either β- or γ-Pla, indicating that binding to LPS is also important for the autoprocessing and folding of the protein. Incubation of KM218 LPS with His6-Pla did not increase the amounts of β- or γ26

forms of His6-Pla (Fig. 5D of II). These results support the hypothesis that binding of the side chains of the arginines to lipid A phosphates is important in activation of His6-Pla. 4.6. Effect of growth temperature in Y. pestis on the activity of Pla (III) The plasminogen activator Pla is clearly important for the infection in the mammalian host. On the other hand, strains devoid of Pla are fully capable of colonizing and surviving in the flea (Sebanne et al., 2006) and Pla seems to be unnecessary for survival of Y. pestis in the flea. McDonough & Falkow (1989) found that the PA activity of Pla is higher at 37oC than at lower temperatures, which suggested that the activity of Pla is adjusted to the host temperature. However, microarray analyses revealed that the level of transcription of pla is similar at the low temperature as at the mammalian body temperature (Han et al., 2004; Han et al., 2005). I compared Plg activation by Y. pestis KIM D27 cells grown at 37oC or 20oC and found that cells from 37oC presented a higher PA activity (Fig. 4). Pla deficient KIM D34 did not activate Plg at all (Fig. 4A). Analysis of OM from both bacteria by Western blotting with antiHis6-Pla (Fig. 4B) and image processing revealed a 1.6 times higher amount of Pla peptides in the OM from 37oC. I adjusted both OM samples to have the same amount of His6-Pla and found that the OM from 37oC exhibited a higher PA activity (Fig. 4C). Interestingly, the Western blotting analyses revealed that β-Pla is only formed at 37oC (Fig. 4B). These results indicate that the higher PA activity of Pla at 37oC involves both a slight increase in the Pla protein amount as well as a slight increase in the activity of the protein; the latter

Results and Discussion

could involve the lower acylation and phosphate substitution in lipid A in cells growing at 37oC. The formation of β-Pla may result from the increased fluidity of LPS with the tetraacyl lipid A (Bengoechea et al., 2003), which could potentiate more contacts between the Pla molecules in the membrane. 4.7. Degradation of gelatin and collagen by Pla and PgtE (II) Surface-associated proteolysis and penetration through tissue barriers are central in cell migration (Plow et al., 1999; Lähteenmäki et al., 2005; Myöhänen & Vaheri, 2004). Collagen and its denatured form, gelatin, are major component of the ECM and hence obvious obstacles for cell migration in tissues. Matrix metalloproteinases (MMPs) form a family of collagenases (which cleave fibrillar collagens) and gelatinases (which cleave gelatin and nonfibrillar collagens) (Johansson et al., 2000), and their role in migration of eukaryotic cells is well characterized (Parks et al., 2004). MMPs are secreted as proenzymes from a number of mammalian cell types, and, interestingly for this thesis work, production of proFigure 4. Effect of growth temperature on Pla activity. A. Plasminogen activation by Y. pestis strains KIM D27 (Pla+) and KIM D34 (Pla-) grown at grown at 20oC and 37oC. B. Western blot of Y. pestis KIM D27 (Pla+) outer membrane extracts grown at a. 20oC and b. 37oC. Migration of α-Pla, β-Pla and γ-Pla are shown in the left. C. Plasminogen activation by the outer membrane extracts of Y. pestis strains KIM D27 (Pla+) and KIM D34 (Pla-) grown at 20oC and 37oC. Pla was quantified by image analyses of the Western blot of the extracts and the same amount of Pla was used.

27

Results and Discussion

MMPs by monocytes is stimulated by LPS (Parks et al., 2004). Pro-MMPS are proteolytically activated to MMPs (or gelatinases); in mammals, the active proteases include plasmin and MMPs. Production of gelatinases or collagenases has been reported for bacteria causing periodontal infections as well as for clostridial species (Harrington, 1996), and a few pathogenic species, such as Vibrio cholerae and Pseudomonas aeruginosa, express proteases that activate pro-MMPs (Okamoto et al., 1997). However, the highly invasive species Y. pestis and S. enterica studied in my thesis have not been reported to express gelatinase activity, which led us to assess possible degradation of collagens/gelatins by Pla and PgtE. The strain 14028R of S. enterica, which lacks O-antigen and hence expresses active PgtE (Lähteenmäki et al., 2005b), rapidly degraded porcine skin gelatin, whereas the pgtE deletion derivative 14028R-1 was

inactive (Fig. 2 of III). The role of active PgtE in the cleavage was further shown by the activity of the complemented strain 14028R-1 (pMRK3) and the lack of activity of 14028R-1 (pMRK31) (Fig. 2B of III). The plasmid pMRK31 encodes a nonactive PgtE D206A with a catalytic site substitution (Kukkonen et al., 2004). PgtE-positive S. enterica strains also cleaved denatured human type I collagen (gelatin) (Fig. 2D of III). Y. pestis KIM D27 with pPCP1 also cleaved the gelatin, however by a lower rate than S. enterica 14028R, and the Pla-negative strain KIM D34 was inactive (Fig. 5). Overexpression of pla in E. coli XL1 (pMRK1) resulted in active gelatin degradation (Fig. 5). PgtE-expressing S. enterica (Fig. 2 of III) as well as recombinant E. coli XL1 (pMRK3) (Fig. 5C of III) also degraded the fluorescein-conjugated DQ-gelatin substrate, which, however, was not seen with XL1 (pMRK1) expressing Pla (Fig.

Figure 5. Comparison of gelatin degradation by Pla and PgtE. Bacteria expressing either PgtE or Pla were incubated with porcine skin gelatin for 4 or 22h, and the degradation of the substrate was accessed by SDS-PAGE analyses. a. S. enterica 14028R; b. E. coli XL1 (pMRK1); c. E.coli XL1 (pSE380); d. Y. pestis KIM D27 (pPCP1+); e. Y. pestis KIM D34 (pPCP1-); f. porcine skin gelatin in PBS; g. His6-Pla reconstituted with R-LPS

28

Results and Discussion

5C of III). These results showed that PgtE and Pla both express gelatinase activity and that PgtE is more active in this novel omptin function. I then compared the gelatin degradation and Plg activation by purified and activated His6-PgtE and its derivatives. The His6PgtE protein activated Plg (Fig. 5B) but did not cause detectable degradation of

the DQ-gelatin (Fig.5A). As with Pla, reconstitution of His6-PgtE with S-LPS gave a lower level of Plg activation than with the R-LPS, and the double mutant protein His6-PgtE R138 R171 showed a low activity. The His6-PgtE protein cleaved the porcine skin gelatin (Fig. 2C of II) as well as denatured human type I collagen (Fig. 2D of II; Fig. 4C), whereas

A.

B.

Figure 6. Degradation of gelatin and collagen type I by His6-PgtE. A. Degradation of DQ-gelatin by reactivated His6-PgtE and its derivatives. The degradation by MMP-9 was done as in Fig 1B of paper III. B. Activation of Plg by reactivated His6-PgtE and its derivatives. C. Degradation of native () and denatured (+) collagen I by His6PgtE Lanes: a. His6-PgtE with R-LPS; b. His6-PgtE D206A with R-LPS; c. His6-PgtE R138E R171E with R-LPS; d. collagen I in buffer. Activation of His6PgtE and its derivatives was done as in paper III.

29

Results and Discussion

no activity was seen with His6-PgtE R138E R171E and His6-PgtE D206A (Fig. 6C). His6-Pla slowly degraded porcine skin gelatin (Fig. 5) but not human type I gelatin (not shown); neither did His6-Pla degraded DQ-gelatin. The results show that PgtE is a gelatinase, capable of degrading pig skin gelatin, the commonly used gelatinase substrate DQ-gelatin, as well as denatured human type I collagen (gelatin), whereas Pla shows detectable activity with the pig skin gelatin but not with the other substrates. The observed activity of PgtE suggested that it could functionally mimic mammalian gelatinases, MMPs, and later work in paper III – to be included in the PhD thesis of Päivi Ramu of this laboratory – showed that this indeed is the case. PgtE also activated pro-MMP-9, which was not seen with Pla, and Pla expressed in E. coli XL1 could be induced to express DQ-gelatin degradation and pro-MMP-9 activation through genetic substitutions of pla regions encoding residues in the

30

surface loops (Figs. 5C-5H of III). Deletion of PgtE also attenuated the strain 14028 in experimentally infected mice, which is the final evidence for a virulence role of PgtE. An interesting feature in paper III for my PhD thesis is that S. enterica cells with active PgtE degraded both gelatin as well as the DQ-gelatin substrate, whereas purified His6-PgtE was active only with the porcine and human type I gelatin but not with the DQ-gelatin (Fig. 5A and). The reason for this discrepancy remains open. The detailed structure of DQ-gelatin is not available, but it is a fluorescein-conjugated compound derived from porcine skin gelatin. It could be that degradation of gelatin requires an immobilization onto the bacterial cell surface and hence is very poor with isolated His6-PgtE; another possibility is that the degradation may require a closer PgtE-substrate contact or a high substrate density obtainable with gelatin but not with DQ-gelatin molecules.

Conclusion

5. Conclusion This thesis work underlines the importance of coordinated regulation of surface structures in creating powerful surface proteolysis. The omptins are unique OMPs in that their proteolytic activity is dependent on LPS. S-LPS inhibits their functions, and in the case of Pla of Y. pestis, this is overcome by the genetic loss of Oantigen biosynthesis genes. S. enterica is an intracellular pathogen, which modifies its surface and LPS inside macrophages in a manner that optimizes PgtE proteolysis. This study showed that the maximal enhancement of Pla activity occurs with an LPS containing the complete outer core region, i.e. an LPS type similar to the one existing in Y. pestis. Also in the case of Y. pestis, changes in the LPS structure occur when the bacteria are transformed from a lower temperature – mimicking the flea host – to the temperature of the mammalian body. It is remarkable that the changes, i.e. lowering of lipid A acylation and phosphate substitution, address the features that were, in this study, found important for Pla activity. The activity of Pla is increased in cells from higher temperature, which, as with PgtE and S. enterica, indicates that Pla proteolysis is regulated by changes in cell wall architecture. In this study, formation of βPla was detected in cells from the higher temperature; this most likely reflects the higher fluidity of tetraacyl LPS lamellae. The LPS-Pla interaction in this study was mainly analyzed with purified His6-Pla and LPS, and it remains to be established whether other structures and modifications in the cell envelope of Y. pestis also affect the proteolysis. Binding of prokaryotic and eukaryotic proteins to lipid A is dependent on a conserved motif of basic amino acids.

That the same motif is important in PlaLPS interaction and that LPS has a role in modifying the folding of Pla, is suggested by the enzymatic inactivity and lack of γ-Pla in the His6-Pla R138E R171E derivative from recombinant E. coli. Although this mutant was nearly inactive in proteolysis, the study, however, does not unequivocally demonstrate that γ-Pla is the active conformation of Pla. No changes in Pla peptide patterns were correlated with the increased enzymatic activity observed upon LPS addition to His6-Pla protein. Thus the biological significance of the γ-Pla formation remains open. It is interesting to note that the OmpT-LPS and Pla-LPS interactions share characteristics (presence of outer core, sensitivity to salts) but also exhibit differences (acylation in lipid A, presence of phosphates in the core region). Pla is a multifunctional protein and known to modify the cell envelope in Y. pestis. The biological significance of these changes has remained open, but they could contribute to communication of Y. pestis with the mammalian host, e.g. in bacterial adhesion and invasion. This study shows that the Pla molecule has intrinsic adhesive properties, which in a previous study were obseved to be blocked in bacteria with SLPS (Kukkonen et al., 2004). Thus the LPS effects also address adhesiveness of Pla-expressing Y. pestis and thereby the establishment of localized proteolysis. The novel gelatinase activity detected with Pla and PgtE can enhance damage of tissue barriers and potentiate cell migration. Yersinia and S. enterica have not been previously shown to have gelatinases. The two omptins differ in this aspect, with PgtE being more active. This may exemplify adaptation to the life style of Salmonella, 31

Conclusion

whose dissemination is thought to involve infective cycles of phagocytes, which are known to secrete pro-MMPs in response to infection. The gelatinase activity, as measured with DQ gelatin, was not detectable with isolated His6-PgtE but was

32

efficient with PgtE-expressing cells. This indicates that some functions of omptins, in this case exemplified by the gelatinase activity, may require the omptin to be expressed on the bacterial cell surface.

References

6. References Achtman, M., Zurth, K., Morelli, G., Torrea, G., Guiyoule, A. and Carniel, E. 1999. Yersinia pestis, the cause of plague, is a recently emerged clone of Yersinia pseudotuberculosis. Proc Natl Acad Sci. 96:14043-8. Anisimov, A.P., Dentovskaya, S.V., Titareva, G.M., Bakhteeva, I.V., Shaikhutdinova, R.Z., Balakhonov, S.V., Lindner, B., Kocharova, N.A., Senchenkova, S.N., Holst, O., Pier, G.B., Knirel, Y.A. 2005. Intraspecies and temperature-dependent variations in susceptibility of Yersinia pestis to the bactericidal action of serum and to polymyxin B. Infect Immun. 73:7324-31. Aumailley, M. and Gayraud, B. 1998. Structure and biological activity of the extracellular matrix. J Mol Med. 76:253-65. Bader, M.W., Sanowar, S., Daley, M.E., Schneider, A.R., Cho, U., Xu, W., Klevit, R.E., Le Moual, H. and Miller, S.I. 2005. Recognition of antimicrobial peptides by a bacterial sensor kinase. Cell. 122:461-72. Bengoechea, J.A., Lindner, B., Seydel, U., Diaz, R. and Moriyon, I. 1998. Yersinia pseudotuberculosis and Yersinia pestis are more resistant to bactericidal cationic peptides than Yersinia enterocolitica. Microbiology. 144:1509-15.

Booth, N.A. Fibrinolysis and thrombosis. Baillieres Best Pract Res Clin Haematol. 1999 3:423-33. Brandenburg, K., Andra, J., Muller, M., Koch, M.H. and Garidel, P. 2003. Physicochemical properties of bacterial glycopolymers in relation to bioactivity. Carbohydr Res. 338:2477-89 Brandenburg, K. and Wiese, A. 2004. Endotoxins: relationships between structure, function, and activity. Curr Top Med Chem. 4:1127-46. Brandenburg, K., Garidel, P., Schromm, A.B., Andra, J., Kramer, A., Egmond, M. and Wiese, A. 2005. Investigation into the interaction of the bacterial protease OmpT with outer membrane lipids and biological activity of OmpT:lipopolysaccharide complexes. Eur Biophys J. 34:28-41. Cavanaugh, D.C. and Randall, R. 1959. The role of multiplication of Pasteurella pestis in mononuclear phagocytes in the pathogenesis of flea-borne plague, J. Immunol. 83:348–363.

Bengoechea, J.A., Brandenburg, K., Arraiza, M.D., Seydel, U., Skurnik, M. and Moriyon, I. 2003. Pathogenic Yersinia enterocolitica strains increase the outer membrane permeability in response to environmental stimuli by modulating lipopolysaccharide fluidity and lipid A structure. Infect Immun.71:2014-21.

Chain, P.S., Carniel, E., Larimer, F.W., Lamerdin, J., Stoutland, P.O., Regala, W.M., Georgescu, A.M., Vergez, L.M., Land, M.L., Motin, V.L., Brubaker, R.R., Fowler, J., Hinnebusch, J., Marceau, M., Medigue, C., Simonet, M., Chenal-Francisque, V., Souza, B., Dacheux, D., Elliott, J.M., Derbise, A., Hauser, L.J. and Garcia E. 2004. Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis. Proc Natl Acad Sci. 101:13826-31.

Bercovier, H., Alonso, J.M., Bentaiba, Z.N., Brault, J., Mollaret, H.H. 1979. Contribution to the definition and the taxonomy of Yersinia enterocolitica. Contrib Microbiol Immunol. 5:12-22.

Cheng, Q., Stafslien, D., Purushothaman, S.S. and Cleary, P. 2002. The group B streptococcal C5a peptidase is both a specific protease and an invasin. Infect Immun. 70: 2408-13.

Bjur, E., Eriksson-Ygberg, S. and Rhen, M. 2006. The O-antigen affects replication of Salmonella enterica serovar Typhimurium in murine macrophage-like J774-A.1 cells through modulation of host cell nitric oxide production. Microbes Infect. 8:1826-38

Cheng, X.W., Kuzuya, M., Kanda, S., Maeda, K., Sasaki, T., Wang, Q.L., TamayaMori, N., Shibata, T. and Iguchi, A. 2003. Epigallocatechin-3-gallate binding to MMP2 inhibits gelatinolytic activity without

33

References

influencing the attachment to extracellular matrix proteins but enhances MMP-2 binding to TIMP-2. Arch Biochem Biophys. 415:12632. Coleman, J.L., Roemer, E.J. and Benach, J.L. 1999. Plasmin-coated Borrelia Burgdorferi degrades soluble and insoluble components of the mammalian extracellular matrix. Infect Immun. 67:3929-36. Collen, D. and Lijnen, H.R. 1991. Basic and clinical aspects of fibrinolysis and thrombolysis. Blood. 78:3114-24. Cornelis, G.R. 2000. Molecular and cell biology aspects of plague. Proc Natl Acad Sci U S A. 97:8778-83. Darveau, R.P. 1998. Lipid A diversity and the innate host response to bacterial infection. Curr Opin Microbiol. 1:36-42. Darwin, K.H. and Miller, V.L. 1999. Molecular basis of the interaction of Salmonella with the intestinal mucosa. Clin Microbiol Rev. 12:405-28. de Cock, H. and Tommassen, J. 1996. Lipopolysaccharides and divalent cations are involved in the formation of an assemblycompetent intermediate of outer-membrane protein PhoE of E.coli. EMBO J. 15:5567-73 de Cock, H., Brandenburg, K., Wiese, A., Holst, O. and Seydel, U. 1999 Nonlamellar structure and negative charges of lipopolysaccharides required for efficient folding of outer membrane protein PhoE of Escherichia coli. J Biol Chem. 274:5114-9 Dekker, N., Cox, R.C., Kramer, R.A. and Egmond, M.R. 2001. Substrate specificity of the integral membrane protease OmpT determined by spatially addressed peptide libraries. Biochemistry. 40:1694-701. Duplantier, J.M., Duchemin, J.B., Chanteau, S. and Carniel, E. 2005. From the recent lessons of the Malagasy foci towards a global understanding of the factors involved in plague reemergence. Vet Res. 36:437-53. Egile, C., d’Hauteville, H., Parsot, C. and Sansonetti, P.J. 1997. SopA, the outer

34

membrane protease responsible for polar localization of IcsA in Shigella flexneri. Mol Microbiol. 23:1063-73. Eriksson, S., Lucchini, S., Thompson, A., Rhen, M. and Hinton, J.C. 2003. Unravelling the biology of macrophage infection by gene expression profiling of intracellular Salmonella enterica. Mol Microbiol. 47:103-18. Ferguson, A.D., Hofmann, E., Coulton, J.W., Diederichs, K. and Welte, W. 1998. Siderophore-mediated iron transport: crystal structure of FhuA with bound lipopolysaccharide. Science. 282:2215-20. Ferguson, A.D., Welte, W., Hofmann, E., Lindner, B., Holst, O., Coulton, J.W. and Diederichs, K. 2000. A conserved structural motif for lipopolysaccharide recognition by procaryotic and eucaryotic proteins. Structure. 8:585-92. Filippov, A.A., Solodovnikov, N.S., Kookleva, L.M. and Protsenko OA. 1990. Plasmid content in Yersinia pestis strains of different origin. FEMS Microbiol Lett. 55:458. Frirdich, E. and Whitfield, C. 2005. Lipopolysaccharide inner core oligosaccharide structure and outer membrane stability in human pathogens belonging to the Enterobacteriaceae. J Endotoxin Res. 11:133-44. Fuchs, H., Simon, M.M., Wallich, R., Bechtel, M. and Kramer, M.D. 1996. Borrelia burgdorferi induces secretion of pro-urokinase-type plasminogen activator by human monocytes. Infect Immun. 64:430712. Grabenstein, J.P., Marceau, M., Pujol, C., Simonet, M. and Bliska, J.B. 2004. The response regulator PhoP of Yersinia pseudotuberculosis is important for replication in macrophages and for virulence. Infect Immun. 72:4973-84. Grenier, D. 1996. Degradation of host protease inhibitors and activation of plasminogen by proteolytic enzymes from Porphyromonas gingivalis and Treponema denticola. Microbiology. 142:955-61.

References

Gruber, A., Mancek, M., Wagner, H., Kirschning, C.J. and Jerala, R. 2004. Structural model of MD-2 and functional role of its basic amino acid clusters involved in cellular lipopolysaccharide recognition. J Biol Chem. 279:28475-82. Goguen, J.D., Hoe, N.P. and Subrahmanyam, Y.V. 1995. Proteases and bacterial virulence: a view from the trenches. Infect Agents Dis. 4:47-54. Gorvel, J.P. and Meresse, S. 2001. Maturation steps of the Salmonella-containing vacuole. Microbes Infect. 3:1299-303.

Han, Y., Zhou, D., Pang, X., Zhang, L., Song, Y., Tong, Z., Bao, J., Dai, E., Wang, J., Guo, Z., Zhai, J., Du, Z., Wang, X., Wang, J., Huang, P. and Yang, R. 2005. DNA microarray analysis of the heat- and coldshock stimulons in Yersinia pestis. Microbes Infect. 7:335-48. Harrington, D.J. 1996. Bacterial collagenases and collagen-degrading enzymes and their potential role in human disease. Infect Immun. 64:1885-91. Hay, E.D. 1991 (ed.): Cell biology of extracellular matrix. Plenum Press, New York.

Groisman, E.A. and Saier, M.H. Jr. 1990. Salmonella virulence: new clues to intramacrophage survival. Trends Biochem Sci. 15:30-3.

Hinnebusch, B.J., Perry, R.D. and Schwan, T.G. 1996. Role of the Yersinia pestis hemin storage (hms) locus in the transmission of plague by fleas. Science 273: 367–370

Groisman, E.A. and Ochman, H. 1996. Pathogenicity islands: bacterial evolution in quantum leaps. Cell. 87:791-4.

Hinnebusch, B.J., Rudolph, A.E., Cherepanov, P., Dixon, J.E., Schwan, T.G. and Forsberg, A. 2002. Role of Yersinia murine toxin in survival of Yersinia pestis in the midgut of the flea vector. Science. 296:7335.

Groisman, E.A., Kayser, J., Soncini, F.C. 1997. Regulation of polymyxin resistance and adaptation to low-Mg2+ environments. J Bacteriol. 179:7040-5. Groisman, E.A. 2001. The pleiotropic twocomponent regulatory system PhoP-PhoQ. J Bacteriol. 183:1835-42. Guex, N. and Peitsch, M.C. 1997. SWISSMODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis. 18:2714-23 Guina, T., Yi, E.C., Wang, H., Hackett, M. and Miller, S.I. 2000. A PhoP-regulated outer membrane protease of Salmonella enterica serovar typhimurium promotes resistance to alpha-helical antimicrobial peptides. J Bacteriol. 182:4077-86. Han, Y., Zhou, D., Pang, X., Song, Y., Zhang, L., Bao, J., Tong, Z., Wang, J., Guo, Z., Zhai, J., Du, Z., Wang, X., Zhang, X., Wang, J., Huang, P. and Yang, R. 2004. Microarray analysis of temperature-induced transcriptome of Yersinia pestis. Microbiol Immunol. 48:791805.

Hohmann, E.L. 2001. Nontyphoidal salmonellosis. Clin Infect Dis. 32:263-9. Holst, O., Muller-Loennies, S., Lindner, B. and Brade, H. 1993. Chemical structure of the lipid A of Escherichia coli J-5. Eur J Biochem. 214:695-701. Humphries, A.D., Townsend, S.M., Kingsley, R.A., Nicholson, T.L., Tsolis, R.M. and Baumler, A.J. 2001. Role of fimbriae as antigens and intestinal colonization factors of Salmonella serovars. FEMS Microbiol Lett. 201:121-5 Johansson, N., Ahonen, M. and Kahari, V.M. 2000. Matrix metalloproteinases in tumor invasion. Cell Mol Life Sci. 57:5-15 Jones, H.A., Lillard, J.W. and Perry, R.D. 1999. HmsT, a protein essential for expression of the haemin storage (Hms+) phenotype of Yersinia pestis. Microbiology 145: 2117–2128

35

References

Kawahara, K., Tsukano, H., Watanabe, H., Lindner, B. and Matsuura, M. 2002. Modification of the structure and activity of lipid A in Yersinia pestis lipopolysaccharide by growth temperature. Infect Immun. 70:40928. Kienle, Z., Emödy, L., Svanborg, C. and O’Toole, P.W. 1992. Adhesive properties conferred by the plasminogen activator of Yersinia pestis. J. Gen. Microbiol. 138:16791687. Kleinman, H.K., McGarvey, M.L., Liotta, L.A., Robey, P.G., Tryggvason, K. and Martin, G.R. 1982. Isolation and characterization of type IV procollagen, laminin, and heparan sulfate proteoglycan from the EHS sarcoma. Biochemistry. 21:6188-93. Kleinschmidt, J.H. 2003. Membrane protein folding on the example of outer membrane protein A of Escherichia coli. Cell Mol Life Sci. 60:1547-58. Knirel, Y.A., Lindner, B., Vinogradov, E.V., Kocharova, N.A., Senchenkova, S.N., Shaikhutdinova, R.Z., Dentovskaya, S.V., Fursova, N.K., Bakhteeva, I.V., Titareva, G.M., Balakhonov, S.V., Holst, O., Gremyakova, T.A., Pier, G.B. and Anisimov, A.P. 2005a. Temperature-dependent variations and intraspecies diversity of the structure of the lipopolysaccharide of Yersinia pestis. Biochemistry. 44:1731-43. Knirel, Y.A., Lindner, B., Vinogradov, E., Shaikhutdinova, R.Z., Senchenkova, S.N., Kocharova, N.A., Holst, O., Pier, G.B. and Anisimov, A.P. 2005b. Cold temperatureinduced modifications to the composition and structure of the lipopolysaccharide of Yersinia pestis. Carbohydr Res. 340:1625-30 Koebnik, R., Locher, K.P. and Van Gelder, P. 2000. Structure and function of bacterial outer membrane proteins: barrels in a nutshell. Mol Microbiol. 37:239-53. Kramer, R.A., Dekker, N. and Egmond, M.R. 2000. Identification of active site serine and histidine residues in Escherichia coli outer membrane protease OmpT. FEBS Lett. 468:220-4.

36

Kramer, R.A., Zandwijken, D., Egmond, M.R. and Dekker, N. 2000b. In vitro folding, purification and characterization of Escherichia coli outer membrane protease ompT. Eur J Biochem. 267:885-93. Kramer, R.A., Vandeputte-Rutten, L., de Roon, G.J., Gros, P., Dekker, N. and Egmond, M.R. 2001. Identification of essential acidic residues of outer membrane protease OmpT supports a novel active site. FEBS Lett. 505:426-30. Kramer, R.A., Brandenburg, K., Vandeputte-Rutten, L., Werkhoven, M., Gros, P., Dekker, N. and Egmond, M.R. 2002. Lipopolysaccharide regions involved in the activation of Escherichia coli outer membrane protease OmpT. Eur. J. Biochem. 269:1746-52. Kukkonen, M. and Korhonen, T.K. 2004. The omptin family of enterobacterial surface proteases/adhesins: from housekeeping in Escherichia coli to systemic spread of Yersinia pestis. Int J Med Microbiol. 294:7-14. Kukkonen, M., Lähteenmäki, K., Suomalainen, M., Kalkkinen, N., Emödy, L., Lang H. and Korhonen, T.K. 2001 Protein regions important for plasminogen activation and inactivation of alpha2-antiplasmin in the surface protease Pla of Yersinia pestis. Mol. Microbiol. 40:1097-111. Kukkonen, M., Suomalainen, M., Kyllönen, P., Lähteenmäki, K., Lång, H., Virkola, R., Helander, I.M., Holst, O. and Korhonen, T.K. 2004 Lack of O-antigen is essential for plasminogen activation by Yersinia pestis and Salmonella enterica. Mol. Microbiol. 51:215225 Lamping, N., Hoess, A., Yu, B., Park, T.C., Kirschning, C.J., Pfeil, D., Reuter, D., Wright, S.D., Herrmann, F. and Schumann RR. 1996. Effects of site-directed mutagenesis of basic residues (Arg 94, Lys 95, Lys 99) of lipopolysaccharide (LPS)-binding protein on binding and transfer of LPS and subsequent immune cell activation. J Immunol. 157:464856

References

Lillard, J.W. Jr, Bearden, S.W., Fetherston, J.D. and Perry, R.D. 1999. The haemin storage (Hms+) phenotype of Yersinia pestis is not essential for the pathogenesis of bubonic plague in mammals. Microbiology. 145:197209. Lähteenmäki, K., Virkola, R., Pouttu, R., Kuusela, P., Kukkonen, M. and Korhonen, T.K. 1995. Bacterial plasminogen receptors: in vitro evidence for a role in degradation of the mammalian extracellular matrix. Infect Immun. 63:3659-64. Lähteenmäki, K., Virkola, R., Saren, A., Emödy, L. and Korhonen, T.K. 1998. Expression of plasminogen activator pla of Yersinia pestis enhances bacterial attachment to the mammalian extracellular matrix. Infect. Immun. 66:5755-5762. Lähteenmäki, K., Kukkonen, M., Korhonen, T.K. 2001. The Pla surface protease/adhesin of Yersinia pestis mediates bacterial invasion into human endothelial cells. FEBS Lett. 504:6972. Lähteenmäki, K., Kyllönen, P., Partanen, L., Korhonen, T.K. 2005a. Antiprotease inactivation by Salmonella enterica released from infected macrophages. Cell Microbiol. 4:529-38. Lähteenmäki, K., Edelman, S., and Korhonen, T.K. 2005b. Bacterial metastasis: the host plasminogen system in bacterial invasion. Trends Microbiol. 13:79-85. Makoveichuk, E., Cherepanov, P., Lundberg, S., Forsberg, A. and Olivecrona, G. 2003. pH6 antigen of Yersinia pestis interacts with plasma lipoproteins and cell membranes. J Lipid Res. 44:320-30. Mangel, W.F., Lin, B.H. and Ramakrishnan, V. 1990. Characterization of an extremely large, ligand-induced conformational change in plasminogen. Science. 248:69-73 McCarter, J.D., Stephens, D., Shoemaker, K., Rosenberg, S., Kirsch, J.F., Georgiou, G. 2004. Substrate specificity of the Escherichia coli outer membrane protease OmpT. J Bacteriol. 186:5919-25.

McDonough, K.A. and Falkow, S. 1989. A Yersinia pestis-specific DNA fragment encodes temperature-dependent coagulase and fibrinolysin-associated phenotypes. Mol Microbiol. 3:767-75. McGuire, P.G. and Seeds, N.W. 1989. The interaction of plasminogen activator with a reconstituted basement membrane matrix and extracellular macromolecules produced by cultured epithelial cells. J Cell Biochem. 40:215-27. Mignatti, P. and Rifkin, D.B. 1993. Biology and biochemistry of proteinases in tumor invasion. Physiol Rev. 73:161-95. Miller, S.I., Ernst, R.K. and Bader, M.W. 2005. LPS, TLR4 and infectious disease diversity. Nat Rev Microbiol. 3:36-46. Monack, D.M., Bouley, D.M. and Falkow, S. 2004. Salmonella typhimurium persists within macrophages in the mesenteric lymph nodes of chronically infected Nramp1+/+ mice and can be reactivated by IFNgamma neutralization. J Exp Med. 199:231-41. Mukerjee, P., Kastowsky, M., Obst, S. & Takayama, K. 1999. Lipopolysaccharide preparations in aqueous media: implications for solution versus suspension. In Endotoxin in Health and Disease (Brade, H., Opal, S.M., Vogel, S.N. & Morrison, D.C., eds), pp. 221-228. Marcel Dekker, Inc., New York. Myohanen H. and Vaheri A. 2004. Regulation and interactions in the activation of cellassociated plasminogen. Cell Mol Life Sci. 61:2840-58. Naylor, H.B., Fukui, G.M. and McDuff, C.R. 1961. Effect of temperature on growth and virulence of Pasteurella pestis. Physical and nutritional requirements for restoration of virulence. J Bacteriol. 81:649-55. Ngampasutadol, J., Ram, S., Blom, A.M., Jarva, H., Jerse, A.E., Lien, E., Goguen, J., Gulati, S. and Rice, P.A. 2005. Human C4b-binding protein selectively interacts with Neisseria gonorrhoeae and results in speciesspecific infection. Proc Natl Acad Sci USA. 102:17142-7.

37

References

Nummila, K., Kilpelainen, I., Zahringer, U., Vaara, M. and Helander IM. 1995. Lipopolysaccharides of polymyxin B-resistant mutants of Escherichia coli are extensively substituted by 2-aminoethyl pyrophosphate and contain aminoarabinose in lipid A. Mol Microbiol. 16:271-8. Ohl, M.E., Miller, S.I. 2001. Salmonella: a model for bacterial pathogenesis. Annu Rev Med.;52:259-74. Okamoto, T., Akaike, T., Nagano, T., Miyajima, S., Suga, M., Ando, M., Ichimori, K. and Maeda, H. 1997. Activation of human neutrophil procollagenase by nitrogen dioxide and peroxynitrite: a novel mechanism for procollagenase activation involving nitric oxide. Arch Biochem Biophys. 342:261-74 Okuno, K., Yabuta, M., Ohsuye, K., Ooi, T. and Kinoshita, S. 2002. An analysis of target preferences of Escherichia coli outermembrane endoprotease OmpT for use in therapeutic peptide production: efficient cleavage of substrates with basic amino acids at the P4 and P6 positions. Biotechnol Appl Biochem. 36:77-84. Oyston, P.C., Dorrell, N., Williams, K., Li, S.R., Green, M., Titball, R.W. and Wren BW. 2000. The response regulator PhoP is important for survival under conditions of macrophage-induced stress and virulence in Yersinia pestis. Infect Immun. 68:3419-25. Parkhill, J., Wren, B.W., Thomson, N.R., Titball, R.W., Holden, M.T., Prentice, M.B., Sebaihia, M., James, K.D., Churcher, C., Mungall, K.L., Baker, S., Basham, D., Bentley, S.D., Brooks, K., Cerdeno-Tarraga, A.M., Chillingworth, T., Cronin, A., Davies, R.M., Davis, P., Dougan, G., Feltwell, T., Hamlin, N., Holroyd, S., Jagels, K., Karlyshev, A.V., Leather, S., Moule, S., Oyston, P.C., Quail, M., Rutherford, K., Simmonds, M., Skelton, J., Stevens, K., Whitehead, S. and Barrell, B.G. 2001. Genome sequence of Yersinia pestis, the causative agent of plague. Nature. 413:523-7. Parkkinen, J., Hacker, J. and Korhonen, T.K. 1991. Enhancement of tissue plasminogen

38

activator-catalyzed plasminogen activation by Escherichia coli S fimbriae associated with neonatal septicaemia and meningitis. Thromb. Haemost. 65:483-6. Parks, W.C., Wilson, C.L. and LopezBoado, Y.S. 2004. Matrix metalloproteinases as modulators of inflammation and innate immunity. Nat Rev Immunol. 4:617-29 Perry, R.D. and Fetherston, J.D. 1997. Yersinia pestis--etiologic agent of plague. Clin Microbiol Rev. 10:35-66. Potempa, J., Banbula, A., Travis, J. 2000. Role of bacterial proteinases in matrix destruction and modulation of host responses. Periodontol 2000. 24:153-92. Pujol, C. and Bliska, J.B. 2003. The ability to replicate in macrophages is conserved between Yersinia pestis and Yersinia pseudotuberculosis. Infect Immun. 71:5892-9. Raetz, C.R. and Whitfield, C. 2002. Lipopolysaccharide endotoxins. Annu Rev Biochem. 71:635-700. Rautemaa, R. and Meri, S. 1999. Complementresistance mechanisms of bacteria. Microbes Infect. 1:785-94. Rijken, D.C., Sakharov, D.V. 2001. Basic principles in thrombolysis: regulatory role of plasminogen. Thromb Res. 1:S41-9. Ringdahl, U., Svensson, M., Wistedt, A.C., Renne, T., Kellner, R., Muller-Esterl, W. and Sjobring, U. 1998. Molecular co-operation between protein PAM and streptokinase for plasmin acquisition by Streptococcus pyogenes. J Biol Chem. 273:6424-30 Robbins, K.C., Summaria, L., Hsieh, B., Shah, R.J. 1967. The peptide chains of human plasmin. Mechanism of activation of human plasminogen to plasmin. J Biol Chem. 242:2333-42 Saksela, O. 1985 Plasminogen activation and regulation of pericellular proteolysis. Biochim Biophys Acta. 823:35-65. Sansonetti, P.J. and Egile, C. 1998. Molecular bases of epithelial cell invasion by Shigella

References

flexneri. Antonie Van Leeuwenhoek. 74:1917.

to and has evolved from Y. pseudotuberculosis serotype O:1b. Mol Microbiol. 37:316-30.

Schechter, I. and Berger, A. 1967. On the size of the active site in proteases. I. Papain. Biochem Biophys Res Commun. 27:157-62

Sodeinde, O.A., Subrahmanyam, Y.V., Stark, K., Quan, T., Bao, Y. and Goguen, J.D. 1992. A surface protease and the invasive character of plague. Science 258:1004-1007

Schulz, G.E. 2000. Beta-Barrel membrane proteins. Curr Opin Struct Biol. 10:443-7. Sebbane, F., Gardner, D., Long, D., Gowen, B.B. and Hinnebusch, B.J. 2005. Kinetics of disease progression and host response in a rat model of bubonic plague. Am J Pathol. 166:1427-39. Sebbane, F., Jarrett, C.O., Gardner, D., Long, D. and Hinnebusch, B.J. 2006. Role of the Yersinia pestis plasminogen activator in the incidence of distinct septicemic and bubonic forms of flea-borne plague. Proc Natl Acad Sci USA. 103:5526-30 Sebbane, F., Lemaitre, N., Sturdevant, D.E., Rebeil, R., Virtaneva, K., Porcella, S.F. and Hinnebusch, B.J. 2006. Adaptive response of Yersinia pestis to extracellular effectors of innate immunity during bubonic plague. Proc Natl Acad Sci USA. 103:11766-71. Sharma V.K. and Dellinger, R.P. 2006. Treatment options for severe sepsis and septic shock. Expert Rev Anti Infect Ther. 4:395403. Sheppard, M., Webb, C., Heath, F., Mallows, V., Emilianus, R., Maskell, D. and Mastroeni, P. 2003. Dynamics of bacterial growth and distribution within the liver during Salmonella infection. Cell Microbiol. 5:593-600 Skrzypek E., Cowan C. and Straley S.C. 1998. Targeting of the Yersinia pestis YopM protein into HeLa cells and intracellular trafficking to the nucleus. Mol Microbiol. 30:1051-65. Skurnik, M. and Toivanen, P. 1993. Yersinia enterocolitica lipopolysaccharide: genetics and virulence. Trends Microbiol. 1:148-52 Skurnik, M., Peippo, A. and Ervela, E. 2000. Characterization of the O-antigen gene clusters of Yersinia pseudotuberculosis and the cryptic O-antigen gene cluster of Yersinia pestis shows that the plague bacillus is most closely related

Sodeinde, O.A., Sample, A.K., Brubaker, R.R. and Goguen, J.D. 1988. Plasminogen activator/coagulase gene of Yersinia pestis is responsible for degradation of plasmidencoded outer membrane proteins. Infect Immun. 56:2749-52 Steltmann, G. and Holst, O. 2002 (ed). The bacterial cell wall. Springer Press. Germany Stephens, R.W. and Vaheri, A. 1993. Plasminogen. In: Guidebook to the Extracellular Matrix and Adhesion Proteins (Kreis, T. and Vale, R., Eds.), pp. 81–82. Oxford University Press, New York. Straley, S.C. and Harmon, P.A. 1984. Yersinia pestis grows within phagolysosomes in mouse peritoneal macrophages. Infect Immun. 45:655-9. Straley, S., Skrzypek, E., Piano, G.V. and Bliska, J.B. 1993. Yops of Yersinia spp pathogenic to humans. Infect and Immun. 61:3105–3110 Studier, F.W. and Moffatt, B.A. 1986. Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol. 189:113-30 Sugimura, K. and Nishihara, T. 1988. Purification, characterization, and primary structure of Escherichia coli protease VII with specificity for paired basic residues: identity of protease VII and OmpT. J Bacteriol. 170:562532. Svensson, M.D., Sjöbring, U., Luo, F. and Bessen, D.E. 2002. Roles of the plasminogen activator streptokinase and the plasminogenassociated M protein in an experimental model for streptococcal impetigo. Microbiology 148:3933-45. Tan, L. and Darby, C. 2005. Yersinia pestis is viable with endotoxin composed of only lipid A. J Bacteriol. 187:6599-600

39

References

Thomsen, L.E., Chadfield, M.S., Bispham, J., Wallis, T.S., Olsen, J.E. and Ingmer H. 2003. Reduced amounts of LPS affect both stress tolerance and virulence of Salmonella enterica serovar Dublin. FEMS Microbiol Lett. 228:225-31 Titball, R.W., Hill, J., Lawton, D.G. and Brown, K.A. 2003. Yersinia pestis and plague. Biochem Soc Trans. 31:104-7.

Wimley, W.C. 2002. Toward genomic identification of beta-barrel membrane proteins: composition and architecture of known structures. Protein Sci. 11:301-12 Wimley, W.C. 2003. The versatile beta-barrel membrane protein. Curr Opin Struct Biol. 13:404-11.

Travis, J., Potempa, J. and Maeda, H. 1995. Are bacterial proteinases pathogenic factors? Trends Microbiol. 3:405-7.

Winfield, M.D., Latifi, T., Groisman, E.A. 2005. Transcriptional regulation of the 4amino-4-deoxy-L-arabinose biosynthetic genes in Yersinia pestis. J Biol Chem. 280:1476572.

Une, T. and Brubaker, R.R. 1984. In vivo comparison of avirulent Vwa- and Pgm- or Pstr phenotypes of yersiniae. Infect Immun. 43:895-900.

Word Health Organization. 1990. WHO consultation on plague New Delhi, India, 11 to15 September 1989, WHO report No. WHO/ MIM/PLA/90.1, 21 p.

Vandeputte-Rutten, L., Kramer, R.A., Kroon, J., Dekker, N., Egmond, M.R. and Gros, P. 2001. Crystal structure of the outer membrane protease OmpT from Escherichia coli suggests a novel catalytic site. EMBO J. 20:5033-9

Wren, B.W. 2003. The Yersiniae - a model genus to study the rapid evolution of bacterial pathogens. Nat Rev Microbiol. 1:55-64.

Vassalli, J.D., Wohlwend, A. and Belin, D. 1992. Urokinase-catalyzed plasminogen activation at the monocyte/macrophage cell surface: a localized and regulated proteolytic system. Curr Top Microbiol Immunol. 181:6586. Viboud, G.I. and Bliska, J.B. 2005. Yersinia Outer Proteins: Role in Modulation of Host Cell Signaling Responses and Pathogenesis. Annu. Rev. Microbiol. 59:69–89 Welkos, S.L., Friedlander, A.M. and Davis, K.J. 1997. Studies on the role of plasminogen activator in systemic infection by virulent Yersinia pestis strain C092. Microb Pathog. 23:211-23 Wick, M.J., Harding, C.V., Normark, S.J. and Pfeifer, J.D. 1994. Parameters that influence the efficiency of processing antigenic epitopes expressed in Salmonella typhimurium. Infect Immun. 62:4542-8. Wilkinson, S.G. 1996. Bacterial lipopolysaccharides--themes and variations. Prog Lipid Res.35:283-343.

40

Yao, J., Mackman, N., Edgington, T.S. and Fan, S.T. 1997. Lipopolysaccharide induction of the tumor necrosis factor-alpha promoter in human monocytic cells. Regulation by Egr-1, c-Jun, and NF-kappaB transcription factors. J Biol Chem. 272:17795-801. Yen, M.R., Peabody, C.R., Partovi, S.M., Zhai, Y., Tseng, Y.H. and Saier, M.H. 2002. Protein-translocating outer membrane porins of Gram-negative bacteria. Biochim Biophys Acta. 1562:6-31. Zhou, D., Han, Y., Qin, L., Chen, Z., Qiu, J., Song, Y., Li, B., Wang, J., Guo, Z., Du, Z., Wang, X. and Yang, R. 2005. Transcriptome analysis of the Mg2+-responsive PhoP regulator in Yersinia pestis. FEMS Microbiol Lett. 250:85-95. Zhou, D., Han, Y. and Yang, R. 2006. Molecular and physiological insights into plague transmission, virulence and etiology. Microbes Infect. 8:273-84.