HELICOBACTER PYLORI COLONIZATION OF THE MOUSE GASTRIC MUCOSA: THE ENTNER-DOUDOROFF PATHWAY AND DEVELOPMENT OF A PROMOTER-TRAPPING SYSTEM

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

by

Amy E. Wanken, B.S., M.S.

* * * * *

The Ohio State University

2003

Dissertation Committee:

Dr. Kathryn Eaton, advisor Approved by Dr. Brian Ahmer

Dr. Kathleen Boris-Lawrie ______Advisor Dr. Charles Brooks Department of Veterinary Biosciences

ABSTRACT

Helicobacter pylori is a microaerophilic, Gram-negative bacterium that persistently colonizes the mucous layer overlying the gastric epithelium of humans.

Colonization with H. pylori results in chronic superficial gastritis, which increases the risk of development of peptic ulcer disease and distal gastric carcinoma. However, the majority of individuals colonized with H. pylori remain asymptomatic. Although several strain-specific factors have been identified and may be markers for the different clinical outcomes of colonization, our present understanding of the role of the bacterial factors that might affect the course of disease is largely incomplete.

The work in this dissertation approaches two different aspects of colonization.

The first is the analysis of a metabolic pathway, the Entner-Doudoroff (ED) pathway.

The enzymes of the ED pathway are present in H. pylori, but the impact of this pathway on colonization has not yet been examined. The objectives of this project were to confirm the presence and activity of 6-phosphogluconate dehydratase (6PGD), the key enzyme in the ED pathway, in H. pylori, to create a mutant strain and confirm loss of enzyme activity, and finally to use the mutant strain in mouse colonization experiments.

In the first part of this project, the activity of the ED pathway in H. pylori was evaluated by cloning the genes and assaying their activity. Results indicated that the key ED enzyme, 6-phosphogluconate dehydratase, was active in H. pylori but the

ii cloned genes were not active in E. coli, indicating the need for an unknown accessory gene or cofactor. Finally, two ED-negative strains, one in H. pylori M6 and one in H.

pylori SS1, were created by insertional mutagenesis into 6-phosphogluconate

dehydratase.

The second part of the project consisted of testing the mutant strains for

colonization ability in a mouse model. Two mouse strains, C57BL/6J and 129, were

inoculated with wild-type and/or mutant H. pylori strains M6 or SS1. All mutant and

wild-type strains colonized, but colonization was lower for both mutant strains

compared to the parental strains. In addition, minimum infectious dose was 100-1000-

fold lower for the wild-type than for the ED-negative mutants. Surprisingly, in spite of

lower colonization density and higher minimum infectious dose, co-inoculation

experiments revealed that wild-type H. pylori did not displace the mutant strain,

indicating that competition between wild-type and mutant did not occur in vivo.

Results from this study indicate that loss of the key ED enzyme in H. pylori

diminishes fitness of the organism in vivo, but that the pathway is non-essential for

colonization. However, conservation of this pathway in H. pylori and the fact that

colonization was diminished by loss of 6PGD, suggests that the ED pathway has some

function in H. pylori metabolism. Although our data show that the pathway is not vital

for colonization, it may have some other function in colonization or transmission of the

organism among hosts.

The second focus of this dissertation was the development of a technique to

screen the H. pylori genome for genes which may be important for colonization. This

technique, in vivo expression technology (IVET), has been used in other pathogens with

iii success and would be a useful tool in H. pylori. The objectives in this project were to

develop an IVET plasmid for H. pylori, and to develop methods for testing and screening an H. pylori promoter library.

The IVET plasmid was developed so that promoter fragments would be inserted

into a StuI site upstream of a promoter-less ureB gene. Expression of this construct in urease-negative H. pylori would be a marker for functional promoters. Various methods for generating a genome library were tested, and it was found that the best

method for achieving random fragments of the appropriate size was by sonication of the

genomic DNA. Agar plates were also developed for the in vitro urease testing of a

large number of colonies by replica plating. This was a necessary step for screening the

cells after they had passed through the mouse challenge, as promoters that are up-

regulated in vivo but are turned off in vitro would be of most interest for further study.

iv

Dedicated to my parents and in memory of Robert Kitchen

v

ACKNOWLEDGMENTS

I wish to thank my advisor, Dr. Kathryn Eaton, first and foremost for accepting me as a student when my primary advisor moved to another university and allowing me to continue the work that I had begun. I would also like to thank her for intellectual support, encouragement, and enthusiasm which made the completion of this dissertation possible, as well as personal support through the process. I am also grateful to Tyrrell Conway, my previous advisor, for allowing me to pursue a project that used an organism with which he was not familiar and for encouraging this project from the beginning. I would also like to thank my committee members, Dr. Brian Ahmer, Dr. Kathleen Boris-Lawrie and Dr. Charles Brooks, for input and critical comments on my work. I would like to acknowledge and thank current and past lab members, and especially Elizabeth Murray, Patrick Baker and Richard Peterson, for always answering questions and offering both intellectual and personal support. I am also grateful for the support and friendship that I found throughout my years at Ohio State from Jon Foster, Rob Walczak, Sue Ann Frank, Traci Hatcher, Jean MacDonald, Brian Mark, Tessa Moir, Julie Maybruck, Gloria Sivko, Larry Dearth, Daniel Sanford, Tiffiney Roberts and Stacey Hull. Finally, I would like to thank my family, James and Susan Wanken, Angela and Ian Kitchen, and Jamie and Kristin Wanken for their love, patience and support. I could not have done this without you!

vi

VITA

October 14, 1972...... Born – Rantoul, IL

1995...... B.S. Biology, University of Notre Dame

1995-1999 ...... Graduate Teaching and Research Associate, Ohio State University

1999...... M.S. Microbiology, Ohio State University

2000-present...... Graduate Research Associate, Ohio State University

PUBLICATIONS

Research Publications

1. A. E. Wanken, T. Conway, and K. Eaton. 2003. The Entner-Doudoroff pathway has little role in Helicobacter pylori colonization of mice. Infect. Immun. 71(5): 2920-2923.

2. K. A. Eaton, J. V. Gilbert, E. A. Joyce, A. E. Wanken, T. Thevenot, P. Baker, A. Plaut, and A. Wright. 2001. Restoration of the ability of Helicobacter pylori to colonize by in vivo complementation of ureB. Infect. Immun. 70(2): 771-778.

3. L. K. Fuhrman, A. E. Wanken, K. W. Nickerson, T. Conway. 1998. Rapid accumulation of intracellular 2-keto-3-deoxy-6-phosphogluconate in an Entner-Doudoroff aldolase mutant results in bacteriostasis. FEMS Microbiol. Lett. 159(2): 261-266.

4. K. Mo, C. O. Lora, A. E. Wanken, M. Javanmardian, X. Yang, and C. F. Kulpa. 1997. Biodegradation of methyl t-butyl ether by pure bacterial cultures. Appl. Microbiol. Biotech. 47(1): 69-72.

vii

FIELDS OF STUDY

Major Field: Veterinary Biosciences

viii

TABLE OF CONTENTS

Page Abstract...... ii

Dedication...... v

Acknowledgments...... vi

Vita...... vii

List of Tables ...... xi

List of Figures...... xii

Chapters:

1. Introduction...... 1

2. Analysis and cloning of the Entner-Doudoroff genes in Helicobacter pylori and creation of a 6-phosphogluconate dehydratase mutant ...... 51

Abstract...... 51 Introduction...... 52 Materials and Methods...... 54 Results...... 66 Discussion...... 69

3. Colonization of the mouse gastric mucosa by Helicobacter pylori strains deficient in the Entner-Doudoroff pathway ...... 72

Abstract...... 72 Introduction...... 73 Materials and Methods...... 74 Results...... 76 Discussion...... 79

ix 4. Creation of an In vivo Expression Technology (IVET) system for use in Helicobacter pylori ...... 83

Abstract...... 83 Introduction...... 84 Materials and Methods...... 88 Results...... 100 Discussion...... 116

5. Perspectives and Future Directions...... 127

Bibliography ...... 130

x

LIST OF TABLES

Table Page

1.1 Deduction of metabolic pathways by carbon-labeling of glucose if allowed to proceed to acetate ...... 18

1.2 Genes identified using IVET ...... 48

2.1 Primer sequences developed for cloning H. pylori genes or confirming plasmid constructs...... 56

2.2 Entner-Doudoroff enzyme activity measured in E. coli wild-type and mutant strains...... 67

3.1 Colonization of mice by H. pylori M6, SS1, or the EDD- mutant ...... 76

3.2 Effect of inoculum dose on colonization of H. pylori in C57BL/6 mice...... 78

3.3 Colonization of mice by a mixture of wild-type and mutant H. pylori M6, SS1, or EDD- mutants...... 79

4.1 Primer sequences developed for IVET project ...... 95

xi

LIST OF FIGURES

Figure Page

1.1 The oxidative phase of the Pentose Phosphate Pathway ...... 20

1.2 The non-oxidative phases of the Pentose Phosphate Pathway...... 21

1.3 The enzymes of the Embden-Meyerhof-Parnas pathway ...... 24

1.4 The Entner-Doudoroff Pathway...... 26

1.5 The proposed citric acid cycle of H. pylori...... 29

1.6 Diagram of the urea cycle in H. pylori ...... 32

1.7 Positive selection of in vivo-induced genes by in vivo expression technology (IVET)...... 43

1.8 Schematic of the resolution event in RIVET ...... 47

2.1 Plasmids constructed for H. pylori colonization studies in mice...... 58

2.2 Confirming the generation of an EDD mutant by PCR ...... 61

2.3 The 6-phosphogluconate dehydratase enzyme assay...... 65

2.4 6-phosphogluconate dehydratase activity of three wild-type H. pylori strains, the EDD- mutant, and E. coli 1485 (wild-type) ...... 68

4.1 Proposed IVET scheme for H. pylori ...... 88

4.2 Plasmid p342, a ureB construct ...... 94

4.3 Plasmid p342 map showing the expected FspI restriction sites...... 97

4.4 Agarose gel of various genomic DNA cutting techniques...... 101

xii 4.5 Urease positive clones on a UTA plate...... 103

4.6 PCR results of orientation of the gfp insert into p342 ...... 105

4.7 Plasmid p342GFPmut3.1 ...... 106

4.8 p342 digested with FspI...... 109

4.9 Agarose gel of pHIVET digested with StuI or FspI ...... 110

4.10 PCR confirmation of ureB and hpn in pHIVET ...... 111

4.11 The pHIVET plasmid...... 111

4.12 Confirmation by PCR of the presence of the insert fragment within pHIVET...... 114

4.13 Determination of insert orientation by PCR ...... 115

xiii

CHAPTER 1

INTRODUCTION

History

Spiral organisms in the stomachs of mammals were discovered more than 125

years ago in 1874, and spiral bacteria were first reported in the human stomach in 1906.

By the end of 1940, two additional reports of spiral gastric bacteria had appeared (64).

Although it was noted in 1924 that the human stomach contains abundant urease

activity (137) and it was shown in 1959 that this activity disappeared after antibiotic treatment (indicating that the enzyme was of bacterial origin), the connection between gastric urease and gastric spiral bacteria was not established until the culturing of H. pylori in 1983 (102, 103, 165).

Research on gastric bacteria was at first impeded by the general belief that bacteria could not live in the acidic environment of the stomach (64, 137). It was not

until 1983 that H. pylori was first cultured by a research scientist, Barry Marshall, through the encouragement of a pathologist, Robin Warren (102). Despite this, it was believed that the organism was not pathogenic because the bacterium did not invade epithelial cells and because of its widespread prevalence in the population. Warren remained interested, however, and made the important observation that most patients with gastritis or peptic ulcers were infected with curved or Campylobacter-like 1 organisms (137). He encouraged Marshall to try to culture the bacteria from endoscopic

biopsies using methods for the culture of campylobacters, but culture for the usual

period of 48 hours produced no growth. The first successful culture occurred by chance

when a biopsy was left in the incubator for 5 days over the Easter holidays in April

1982 (64, 137). By 1984, two other groups had independently reported the disease

associations of the organism (137). Marshall hypothesized that the new bacterium was a Campylobacter found in the pyloric region of the stomach and therefore called it

Campylobacter pyloridis (102), even though the new organism had different flagellar

morphology than campylobacters. C. pyloridis became C. pylori when linguists pointed

out that the former was grammatically incorrect (101), and Campylobacter became

Helicobacter when it was discovered that its 16S ribosomal RNA was clearly distinct

from other Campylobacter species tested (63).

General characteristics

Helicobacter pylori is a gram-negative, curved or slightly spiral,

microaerophilic, slow-growing organism. Its most characteristic enzyme is a potent

multi-subunit urease. H. pylori is motile and possesses four to six sheathed polar

flagella (80, 150). The bacterium’s unique feature is its ability to colonize the stomach.

According to current knowledge, there are no bacteria, with the exception of the

helicobacter genus, that can establish long-term residence in the human gastric mucosa

(34).

Because of the relevance of this organism to human health, an effort was made

to sequence the genome. H. pylori 26695, originally isolated from a gastritis patient in

2 the United Kingdom, was the strain chosen for sequencing because it colonizes piglets

and elicits immune and inflammatory responses (157). Strain J99 was later sequenced

(7). The H. pylori genomes consist of a circular chromosome approximately 1.7 Mb in size. Of the 1590 predicted coding sequences, 594 lack homology to E. coli or H. influenzae genes (76). 499 of these lacked obvious homology to any sequences in databases at the time of annotation, in 1997 (34). Some of these species-specific genes no doubt play an important role in adaptation of H. pylori to the human stomach.

Analysis of the H. pylori 26695 genome sequence reveals fewer global regulatory proteins than expected, based on an E. coli model (157). For example, there are no homologs of OxyR, SoxR, or SoxS oxidative stress regulators, and no homologue to LexA (SOS DNA damage response regulator) was found (15). Four open reading frames (ORFs) had homology to two-component sensor ORFs and six genes encoding response regulators (157). Based on structural and functional homologies, one response-regulator pair has been assigned to be the H. pylori CheA-CheY two- component regulatory system regulating chemotaxis (13, 157). The organism appears to have a limited metabolic and biosynthetic capacity. These characteristics are consistent with those of an organism that colonizes a restricted ecological niche (157).

One interesting feature of the genome is that many predicted proteins, including urease, are most closely related to corresponding proteins from Gram-positive organisms,

Archaea, or eukaryotes, rather than from other Gram-negative organisms (157), suggesting horizontal gene transfer during the evolution of H. pylori.

3 Disease and Epidemiology

For most of the twentieth century, peptic ulcers were thought to be stress-related and caused by hyperchlorhydria. The discovery that H. pylori is associated with gastric inflammation and peptic ulcer disease was initially met with skepticism. However, this discovery and subsequent studies on H. pylori have revolutionized our view of the gastric environment, the diseases associated with it, and the appropriate treatment regimens.

H. pylori is probably the most common human chronic infection and is distributed worldwide. More than half of the world’s population is persistently colonized with H. pylori (34). The prevalence of H. pylori is much higher in emergent countries than in the Western world, although it increases with age in all populations studied (54). The incidence of H. pylori infection declines with increasing standards of socioeconomic development, including sewage disposal, water chlorination, hygienic food preparation, decreased crowding, and education (134). Once established, H. pylori generally persists for decades unless eradicated by anti-microbial therapy.

Because of the worldwide prevalence of infection, transmission of the organism from person to person is a major concern. Early studies demonstrated intrafamilial clustering of infection (42) and, more recently, DNA analyses of isolates have confirmed that most transmission occurs locally within families or small population groups (86, 153). These results are consistent with an apparent inability of H. pylori to proliferate or survive for long periods in the environment.

The mode of transmission from person to person has not been proven definitively. Oral-fecal, oral-oral, and gastro-oral routes have all been considered

4 possible (109). The study of the mode of transmission is made difficult by the fact that,

unlike other infectious diseases, there is no well-defined clinical syndrome associated

with its acquisition, and expression of chronic H. pylori infection in humans is highly

variable. Hence, the usual approach of identifying cases and determining important exposures with infectious diseases is not possible.

Although most infections are asymptomatic, about 10% of cases of H. pylori

colonization lead to illness (34). H. pylori is a major cause of chronic active gastritis and peptic ulcer disease (80, 81, 141, 150) and is also an early risk factor for gastric cancer (127, 128). Infections are occasionally cleared spontaneously after a brief acute phase, but many last for years or decades, and it is these long-term infections that are most often implicated in disease (16, 159). The persistence of the H. pylori infection is surprising considering that the bacterium stimulates marked humoral and cellular immune responses in the human host, which are insufficient to clear the infection (26).

In the years since the connection was made between H. pylori and its associated diseases, numerous studies have been conducted in areas such as clinical studies, treatment, pathogenesis, molecular biology, and epidemiology. Only relatively recently has the understanding of the physiology and biochemistry of H. pylori been advancing.

Despite the slower advent of these studies, they are important and may reveal other factors relevant for colonization or pathogenesis. It is especially important to understand the processes by which basic nutrients are used, and the ability of the organism to adapt to obtain basic metabolites necessary for survival in vivo. Recent

advances in these areas have served to increase our knowledge of how H. pylori works

at a basic level, but there is still much to understand.

5 Ecological niche

The means by which H. pylori occupies its gastric niche and the basis for its

biochemical capabilities and requirements is of great fundamental interest. A thorough

understanding of H. pylori physiology and metabolism could lead to new and better

drug therapies, to the identification of potential targets for therapeutic intervention, or to

an effective vaccine. Understanding how the organism colonizes and persists in the

host is an important step in fully understanding its pathogenesis.

H. pylori has adapted to survive in the very specific and unique ecological niche of the human stomach. Because of the presence of gastric acid, which rapidly destroys the majority of bacteria, the stomach is typically colonized only by transient oral flora

(156). H. pylori appears to have evolved specific mechanisms to assist its survival in

the hostile environment. These factors include the ability to swim well in the thick,

protective mucus gel layer, the ability to transiently survive exposure to acid, and the

ability to attach to the epithelial cell layer to prevent the bacteria from being washed out

of the stomach through the mechanical action of peristalsis. As a result, other infectious

agents do not appear to successfully compete with H. pylori in this environment.

Major non-specific host defenses against microbial colonization of the stomach are gastric acid, peristalsis, and the continual shedding of the cells and mucus lining the gastric surface (34). This layer of gastric mucus, about 0.2-0.6 mm thick, is secreted by epithelial cells and plays an important protective role in the stomach, both as a lubricant and as part of the gastric mucosal barrier against acid and pepsin (62).

H. pylori thrives in a mucus environment. Mucus is a viscous, complex substance composed of glycoproteins and glycolipids. The sugars found in mucus

6 glycoproteins include N-acetylglucosamine, N-acetylgalactosamine, galactose, fucose,

sialic acids, and lesser amounts of glucuronate and galacturonate (5). Gastric mucus gel

is an important aspect of mucosal defense against acid, pepsin, and mechanical damage

(5). The mucus layer forms a stable, continuous layer over the mucosa and provides an

environment for the neutralization of lumenal acid, maintaining a near neutral pH at the

mucosal surface (5).

According to studies employing light and electron microscopy and examination of stained gastric biopsies, H. pylori can be found both in the gastric mucus and in close

proximity to the mucosal epithelial cells where it is protected from the acidic

environment of the stomach lumen (17, 68). Specifically, the ideal adhesion site

appears to be the tight intercellular junctions of the mucosal epithelial cells (126).

Biopsy specimens demonstrated that greater than 80% of the organisms were within 2

µm of a junction (68). Because H. pylori demonstrates a predilection for the gastric pits

rather than the more acidic portion of the mucus closer to the gastric lumen, it may be that the organism is not deriving its nutrients from the lumen but from nutrients

“leaking” through the epithelial junctions. It has been proposed that the source of

carbon and energy is the blood rather than the lumen of the stomach (53, 68). For

example, urea can cross the epithelium and may therefore be utilized by the organism as a growth substance (68). In addition, lesions from gastritis may increase the permeability of the epithelial layer and may therefore increase the amount of nutrients that can be obtained by the organism (39, 138).

7 Acidity and H. pylori

H. pylori is unusual among pathogenic bacteria in its ability to colonize the human stomach, an environment of high acidity. Although the stomach was once believed to be sterile due to its low pH, it is now known that a few acid-tolerant lactobacilli and yeast are present in mouse gastric mucosa (143). Still, histological studies have shown that the area of the mucosa colonized by H. pylori is relatively free of other microbes (147).

Although the pathways through which bacterial pH homeostasis is maintained are poorly understood, it is believed that H. pylori possesses several mechanisms which enable it to transiently survive in an acidic environment. It is thought that H. pylori has the ability to establish a positive inside-membrane potential so that the extrusion of protons is more easily accomplished. Studies have shown that at a pH of 7.0, its positive inside-membrane potential is more similar to that of E. coli, a bacterium that is acid sensitive. However, the positive inside-membrane potential generated by the organism at a pH of 3.0 is comparable with two organisms, Bacillus acidocaldarius and

Acidiphilium facilis, that are able to grow at that pH (104).

H. pylori also alters the secretion of gastric acid, producing hypochlorhydria in the host. The exact mechanism by which this is accomplished is still unknown, although one theory is that parietal cell function is directly impaired. It has been shown that both H. pylori whole cells and sonicate can inhibit acid secretion from rabbit parietal cells in vitro (22). Another group found that two fatty acids produced by the organism inhibit acid production as well as H+, K+-ATPase activity in parietal cells,

8 suggesting that its anti-acid effect may be caused by blocking parietal cell proton transport (14). Thus, a variety of bacterial factors may have an effect on parietal cell function in vivo.

It has also been observed that the development of inflammation coincides with the induction of hypochlorhydria, suggesting that inflammation that occurs in early infection may directly inhibit acid secretion. Increased levels of various cytokines such as interleukin (IL)-1β, IL-6, IL-8, and tumor necrosis factor (TNF)-α, are present in the gastric mucosa of infected people. At least one of these factors (IL-1β) is known to be a strong inhibitor of acid secretion when injected intravenously or intraperitoneally in rats and may show this effect in human infections as well (107). Strangely, hypochlorhydria generally resolves despite the persistence of H. pylori and gastric pH levels return to normal in most patients within several months.

Despite the fact that H. pylori colonizes the human stomach and has mechanisms with which to tolerate acidity, the bacterium is sensitive to acid. H. pylori can grow in a pH range of 5.0-7.5 (107, 150) and can survive in solutions with a pH as low as 4.5 (30). However, the gastric pH is approximately 2, and the organism cannot survive in solutions with a pH <4.0 without the addition of urea (100). Thus, the ability to produce high levels of urease has a protective function, as the highly active enzyme neutralizes acid in the microenvironment due to the generation of ammonia from urea

(111).

As H. pylori enters the gastric environment through ingestion, it is transiently subjected to the extreme pH of the lumen side of the gastric mucous layer. Diffusion of protons from the gastric lumen and secretion of bicarbonate by epithelial cells

9 underlying the mucus gel layer result in a strong pH gradient across the mucus layer,

ranging from pH<2 at the luminal surface to nearly neutral pH at the epithelial cell surface (6, 62, 135). After passage through the gastric lumen, the organism locates itself within the mucus as a means of protection from gastric acidity (150) and finally colonizes the neutral zone close to the epithelial cells (62). Because of the protective

“mucus blanket”, other acid protection mechanisms may not be utilized by the organism at all times during a chronic infection and may only be needed for protection during the initial establishment of infection.

Genetic diversity of H. pylori

Several factors could influence the pathophysiology and severity of disease associated with infection by different strains. First, strain-specific genes, such as those associated with the plasticity zone (an area in which there are significant differences between strains), could play a role. Second, differences in gene expression, perhaps mediated by slipped-strand repair, may be important and may affect the ability of the organism to colonize. Third, host factors may play a significant part in susceptibility to, and severity of, disease. In any host-parasite relationship, bacterial, host, and environmental factors influence the host’s susceptibility to the clinical outcome of infection. For example, different mouse strains exhibit markedly different susceptibilities to H. pylori colonization and clinical outcome (142).

H. pylori isolated from different patients and examined by PCR-based RAPD fingerprinting indicate that nearly every strain is unique at the DNA level (1). The level of polymorphism observed is higher than that in E. coli and similar to that in Neisseria

10 meningitidis (153). Portions of several H. pylori genes, such as vacA (encoding a vacuolating cytotoxin) (10, 162), hspA (encoding a GroES homolog) (85), cagA (161,

168), and iceA (136) show substantially higher levels of genetic diversity among strains

(60-90% nucleotide identity).

In addition to point mutations that are unique to individual strains, large gene

clusters are present in some strains but not in others. Early studies revealed that about

60% of H. pylori isolates produce an immunodominant 120-140 kDa protein of

unknown function (CagA) (35, 36). The cagA gene is located within an approximately

40 kb DNA segment called the cag pathogenicity island (PAI) (2, 23). Many of the

genes within the cag PAI help H. pylori activate proinflammatory signal transduction

pathways in gastric epithelial cells (145, 146), contributing to the host inflammatory

response. Infection with strains that contain the cag PAI is more likely to result in

clinical disease than is colonization with cag-negative strains (33, 35, 36). Therefore,

the presence or absence of the cag PAI is an important characteristic among H. pylori

strains.

The cag PAI has a lower G+C content than that of the H. pylori chromosome

(35% versus 39%, respectively) and contains terminal 31-bp direct repeats (157). These

features suggest that the PAI was acquired from an unrelated bacterial species.

However, in contrast to several of the best-studied bacterial PAIs, which tend to be

inserted within or adjacent to tRNA genes or other small RNA-related genes, the cag

PAI is located between two protein-encoding genes. Although the cag PAI gene

content and arrangement are fairly well conserved among cag+ strains (2, 23, 157), in at

least one strain, the cag PAI has been split into two parts and rearranged due to the

11 insertion of a transposable element (IS605) (2, 23). Many other cag+ strains contain a specific DNA inversion that separates cagA from the adjacent cagB gene (2).

PCR-based subtractive hybridization comparing H. pylori J166, a monkey-

colonizing strain, with H. pylori 26695, a fully sequenced reference strain, has resulted

in the identification of additional chromosomal sites of high-level diversity. Eighteen

different strain-specific DNA fragments were found, seven of which appear to encode

DNA restriction-modification enzyme systems (4). This suggests that restriction-

modification systems may vary considerably among different H. pylori strains (see

Chapter 4). Additional strain-specific genes have been identified by comparing the

complete genome sequences of H. pylori 26695 with H. pylori J99 (7). The overall

genomic organization, gene order and predicted proteomes (sets of proteins encoded by

the genome) were found to be quite similar. Both strains contained the complete cag

PAI flanked by the same chromosomal genes and a previously described 31-bp repeat

(2). The DNA-sequence differences between orthologues from the two strains are

found mainly in the third position of coding triplets, consistent with the variation seen

between H. pylori strains identified by methods dependent on the nucleotide sequence

or on the sequencing of specific loci in different strains (1). However, this nucleotide

variation does not translate into a highly divergent proteome. A total of 275 (18.4%)

J99 and 290 (18.7%) 26695 gene products have orthologues of unknown function in

other species, and 346 (23.1%) J99 and 367 (23.6%) 26695 genes are H. pylori specific,

showing no sequence similarity with genes available in public databases. Of these H.

pylori specific genes, 56 and 69 are specific to stains J99 and 26995, respectively (7).

12 The fact that strain-specific DNA-restriction/modification genes have a lower

(G+C) content than the remainder of the genome and are associated with regions that

are organized differently in the J99 and 26695 genomes indicates that these genes may

have been acquired horizontally from other bacterial species or transferred more

recently from other H. pylori strains by natural transformation (7).

There is increasing evidence that recombination occurs between different H.

pylori strains (10, 58, 61). Simultaneous colonization of human stomachs with more

than one strain of H. pylori is detectable in about 5-10% of patients in the United States

(57), and may occur even more commonly in other populations (82). Mixed infections, even those that are transient, provide an opportunity for genetic exchange between strains. Analysis of single cell clones from a patient who was naturally infected with two different H. pylori strains, one of which was cag+ and the other cag-, revealed evidence for at least six different genetic exchanges. One of these exchanges resulted in

the replacement of the entire cag PAI with DNA containing the “empty site allele” from

the cag-negative strain. Several others involved a region encoding putative outer

membrane proteins that could be involved in interactions with the host (86). Genetic

exchange may play an important role in the biology of H. pylori by generating new

genotypes much more rapidly than is possible by mutation alone, therefore allowing

cells to rapidly adapt to new sites in the gastric environment or to new hosts.

For several genes, including flaA, flaB, and portions of vacA, phylogenetic

analyses of orthologous sequences from different strains have yielded a “bush”-like

rather than a “tree”-like pattern, indicating considerable interstrain recombination.

Recombination is thought to occur more commonly in H. pylori than in any other

13 bacterial species analyzed thus far (153). Thus, the high level of allelic variation

observed in H. pylori can be attributed to at least two factors. First, large populations of

H. pylori have probably evolved within millions of individual human stomachs over thousands of years, resulting in considerable mutational diversity. Second, additional diversity has accumulated as a result of extensive intragenic recombination due to mixed infections (34).

As with several other bacterial species, including Neisseria spp. and

Streptococcus pneumoniae, H. pylori is naturally competent for genetic transformation in vitro (125). This property allows for the uptake of exogenous DNA, which may subsequently replicate, in the case of plasmids, or incorporate into the chromosome by homologous recombination. In addition to genetic exchange via transformation, H. pylori strains may exchange DNA via a contact-dependent mechanism resembling bacterial conjugation (89).

Substrate utilization

Little is known about the details of H. pylori catabolism and anabolism.

Because H. pylori mainly colonizes human stomachs, it has almost certainly evolved special mechanisms to exploit its unique ecological niche. Despite the importance of this pathogen, its preferred growth substrates either in vitro or in vivo are largely unknown and its key metabolic processes are only beginning to be understood.

Although studies have shown which metabolic pathways may function in the organism, little is known about the substrates that it uses to support its growth in the stomach.

Additionally, it is not known which, if any, of these metabolic pathways is important for

14 colonization. Competition is virtually non-existent in its natural habitat of the stomach,

allowing H. pylori to utilize whichever substrates it has access to as long as it contains

the necessary machinery. As a result, metabolic pathways would likely enable the use

of simple, easily obtained nutrients.

As previously stated, H. pylori may obtain its carbon and energy sources from

the blood supplying these areas rather than in the stomach lumen. Additionally, living

on the stomach epithelium and in the stomach mucus exposes the bacterium to various

products of host metabolism. These metabolites include monosaccharides, and so it is

easy to believe that the organism might utilize some of these compounds as a source of

carbon and energy.

In the past, it was believed that H. pylori metabolism was similar to that of the

Campylobacter species. Like Campylobacter, it was thought that H. pylori lacked

enzymes necessary for carbohydrate fermentation. Studies showed no evidence of

fermentative pathways using standard microbiological techniques or rapid identification

kits and thus it was assumed that the organism could not utilize carbohydrates (103,

108, 111). Since then, it has been determined through numerous studies that although

H. pylori does not use most carbohydrates, it can metabolize glucose.

The utilization of glucose can provide not only core metabolites (glucose-6-P, fructose-6-P, ribose-5-P, erythrose-4-P, triose-P, 3-P-glycerate, succinyl-CoA, α- ketoglutarate, acetyl-CoA, pyruvate, phosphoenolpyruvate, and oxaloacetate) but also generate the energy (ATP) needed for work carried out by the bacterial cell (120). Most of these metabolites can be produced via the citric acid cycle and either the Embden-

Meyerhof-Parnas (EMP) glycolytic pathway or the Entner-Doudoroff (ED) pathway.

15 These metabolites are then used as the starting materials for amino acids, nucleic acid

bases, and cofactors that the cell requires. Energy production from carbohydrate

metabolism is generally through substrate level phosphorylation and oxidative

phosphorylation.

Some of the earliest studies regarding glucose utilization in H. pylori were

radioactive tracer analysis experiments using D-[U-14C]-glucose. These studies

indicated that glucose is utilized by H. pylori and determined transport and

incorporation rates of the sugar (112, 119). Details of the transport and utilization of

glucose by H. pylori have since been elucidated by several approaches.

Utilization of carbohydrates begins with the ability to transport them across the

cell membrane and into the cell. Glucose transport into cells was analyzed using 2-

deoxy-D-glucose, a glucose analogue that utilized the same transporter but was not further metabolized by H. pylori. The transporter was found to have a Km of 4.8 +/- 0.8

-1 -1 mM and a Vmax of 146 +/- 8 pmol s (ul cell water) (112). It is thought that the use of

other nutrients available in the environment in preference to glucose, the lack of

microbial competition, and the availability of glucose may help explain the fairly high

Km for transport and the low affinity for the substrate (112). Competitive inhibition of the transporter was also studied to determine the specificity of the transporter. Most monosaccharides tested (D-arabinose, D-fructose, L-glucose, D-ribose, deoxy-D-ribose,

D-xylose, D-gluconate, D-glucuronate, D-saccharate) and all disaccharides tested

(lactose, maltose, sucrose) showed less than 15% inhibition. L-arabinose inhibited glucose transport by 20-25%. Competition for the transporters was seen with D- galactose (47%) and D-glucose (57%) (112). These results imply that binding

16 specificity to the transporter appears to be due to the presence of the 6-OH and the

configuration of the group at C-3 on the hexose ring (112). Transport of glucose into

the cell is thought to be carrier mediated because the process has a stereospecificity for

D-glucose, saturable kinetics, and temperature dependence (112).

In addition to transport, another important step in the utilization of sugars is

activation by phosphorylation during uptake. Various monosaccharides and

disaccharides were tested as substrates in incubations with bacterial lysates and ATP

using 13C and 31P nuclear magnetic resonance (NMR) spectroscopy (116). The

monosaccharides included two trioses (D-glyceraldehyde, dihydroxyacetone), three

tretoses (D-erythrose, D-threose, D-erythrulose), seven pentoses (D-ribose, D-

arabinose, L-arabinose, D-xylose, D-lyxose, D-ribulose, D-xylulose), eight hexoses (D-

allose, D-glucose, D-mannose, D-galactose, D-talose, D-fructose, D-sorbose, L-

sorbose), and D-gluconic acid and D-glucuronic acid. The disaccharides included maltose, trehalose, sucrose, lactose, cellobiose, gentobiose, and melibiose. Of the 22

sugars tested, only D-glucose was phosphorylated (116). The enzyme for

phosphorylation had a high substrate specificity, a relatively high Km, and showed no

inhibition in the presence of excess substrates. Therefore, it is suggested that it is a

highly specific glucokinase, rather than a more general hexokinase or an E-III enzyme

of the glucose phosphotransferase system (116).

To further study the characteristics of glucose utilization, loss of label of D-

[13C]-glucose was analyzed by NMR spectroscopy (119). It was found that the

disappearance of the label exhibited biphasic characteristics. Initially, glucose

metabolism was slow. This was followed by a second phase in which utilization was at

17 least one order of magnitude faster. It was during the second phase that the highest

levels of product accumulated. When glucose labeled in different locations of the molecule was used, it was found that the label at C1 was being lost. Because levels

declined nearly two times faster with D-[1-13C]-glucose versus D-[6-13C]-glucose, it

was postulated that this was probably lost as CO2 in the oxidative phase of the pentose phosphate pathway. However, products were still seen with the D-[1-13C]-glucose

incubations suggesting the presence of another catabolic pathway (119).

Following transport and phosphorylation, the catabolism of glucose could

theoretically occur via three pathways: Embden-Meyerhof-Parnas (EMP) glycolysis,

the pentose phosphate (PP) pathway, or the Entner-Doudoroff (ED) pathway. When

allowed to proceed to acetate, the three pathways differ in the metabolism of glucose

(Table 1.1). Therefore, investigation of the labeling patterns of the end-products of

Labeled carbon Detection of label Pathway 1 CH3 EMP CO2 ED or PP 2 CH3 PP COOH EMP or ED CO2 PP 3 CH3 ED COOH PP CO2 EMP or PP 4, 5, 6 no variation between pathways

Table 1.1: Deduction of metabolic pathways by carbon-labeling of glucose if allowed to proceed to acetate. EMP = Embden-Meyerhof-Parnas pathway; ED = Entner- Doudoroff pathway; PP = pentose phosphate pathway. Modified from Chalk et al. (24).

metabolism makes it possible to deduce the metabolic pathways involved. In one study,

it was observed that glucose was oxidized to acetate by H. pylori, although in small

18 amounts (24). The labeling patterns observed when differently labeled glucose

substrates were used suggested that oxidation of glucose by H. pylori takes places via the Entner-Doudoroff pathway (24).

Over the years, each step in discovering the details of H. pylori metabolism was small, but taken together they begin to complete the picture of the overall metabolic pathways that exist in this organism. It is now known that the bacterium can specifically transport and activate at least one carbohydrate, glucose. Knowledge is at a point to which the application of these pathways must be explored. Based on sequence analysis, it is known which genes for these pathways exist in the organism, and biochemical tests on cell lysates have shown enzyme activity. Whether the pathways are truly functional within the cell and serve as a means to acquire carbon and energy for the organism, and whether they are essential for colonization of the host remains to be studied. The answers to these questions are important. Even if carbohydrate metabolism does not contribute directly to H. pylori-induced disease, it is likely to affect colonization fitness to some degree. To fully answer these questions, the metabolism of H. pylori in its gastric microenvironment must be studied.

Pentose Phosphate Pathway

Growing cells have basic metabolic requirements requiring reducing power as well as components to synthesize nucleotides. These needs can be met in part through the use of the pentose phosphate pathway (114). The pentose phosphate pathway is generally divided into three stages: 1) oxidative reactions, 2) isomerization and epimerization reactions, and 3) C—C bond formation and cleavage reactions. The

19 Glucose-6-phosphate

Glucose-6-phosphate NADP+ dehydrogenase NADPH

6-P-gluconolactone

H2O Gluconolactonase H+

6-P-gluconate

NADP+ 6-P-gluconate dehydrogenase NADPH

CO2 Ribulose-5-P

Figure 1.1: The oxidative phase of the Pentose Phosphate Pathway. Each molecule of glucose-6-phosphate generates two molecules of NADPH and one molecule of CO2.

oxidative phase yields both NADPH and ribulose-5-phosphate (Ru5P) (Figure 1.1).

For each molecule of glucose-6-phosphate (G6P) that enters the pathway, two molecules of NADPH are made and are used as reducing power in other reactions.

Isomerization and epimerization produce either ribose-5-phosphate (R5P) or xylulose-5- phosphate (Xu5P) from Ru5P. When needed, R5P can be used for nucleotide biosynthesis. The final phase of the pathway results in the formation of fructose-6- phosphate (F6P) and glyceraldehyde-3-phosphate (G3P) (163) (Figure 1.2). The reactions in the second and third phases are reversible and, as such, the products made vary with the needs of the cell.

20 Ribulose-5-P

Ribose-5-P Ribose-5-P epimerase isomaerase

Xylulose-5-P Ribose-5-P

Transketolase

Sedoheptulose-7-P Glyceraldehyde-3-P

Transaldolase

Fructose-6-P Erythrose-4-P Xylulose-5-P

Transketolase

Fructose-6-P Glyceraldehyde-3-P

Figure 1.2: The non-oxidative phases of the Pentose Phosphate Pathway. In these steps, a series of sugar interconversions occur, yielding ribose-5-phosphate for nucleotide biosynthesis and fructose-6-phosphate and glyceraldehydes-3-phosphate for glycolysis/gluconeogenesis. These reactions are near equilibrium, with fluxes driven by supply and use of these three intermediates.

The enzymes of this metabolic pathway were discovered in H. pylori using 31P

NMR spectroscopy (114). This method allows for the simultaneous analysis of label distribution by divergent metabolic routes. Specific enzyme activities from both oxidative and non-oxidative phases of this pathway were detected when end products from incubations of bacterial lysates with specific substrates were analyzed (114). It was found that glucose-6-phosphate dehydrogenase and phosphogluconolactonase activities were present in H. pylori cells. These two enzymes catabolize the initial steps in both the pentose phosphate and the Entner-Doudoroff (ED) pathways. The presence 21 of 6-phosphogluconate dehydrogenase was also observed (114). The existence of these three enzymes suggests the presence of the oxidative phase of the pentose phosphate

pathway in H. pylori (163). Phosphopentose isomerase and phosphopentose epimerase

activities were also detected (114). This reveals that the non-oxidative phase of the

pathway may also function in H. pylori. Assays for enzymes of the third phase of the

pathway were conducted as well, and the presence of transketolase and transaldolase

was discovered (114).

As a result of this work, it was shown that H. pylori produces the enzymes for a

complete pentose phosphate pathway. However, the specific use of the pentose

phosphate pathway for oxidative glucose metabolism as a source of carbon and energy

has not been proven. It has thus been suggested that this pathway functions in the

organism as a means of supplying the cell with NADPH for reductive biosynthesis (i.e.,

anabolic pathways) and ribose-5-phosphate for nucleic acid synthesis (114).

During the course of these experiments, the activities of two other enzymes,

triosephosphate isomerase (TIM) and glucose-6-phosphate isomerase (also called

phosphoglucose isomerase (PGI)) were detected (114). Not part of the pentose

phosphate system, TIM enables the interconversion of G3P and dihydroxyacetone

phosphate. PGI is utilized to convert F6P to G6P, an activated form of glucose that can

enter several glycolytic pathways such as the Embden-Meyerhof-Parnas pathway, the

pentose phosphate pathway, or the Entner-Doudoroff pathway (163).

22 Embden-Meyerhof-Parnas pathway

The EMP pathway, in which there are seven common reversible reactions, can

function as either a catabolic or a gluconeogenic pathway. The two activities of the

pathway are distinguished by three opposed irreversible steps (Figure 1.3). There has

been some controversy as to whether this pathway is active in H. pylori. In earlier

NMR spectroscopy studies, EMP glycolysis did not appear to be active because glucose

labeled at the 1-C did not lead to acetate labeled at the CH3 group (Table 1.1) (24) and

the lack of detection of some enzyme activities (118). However, in a more recent study utilizing spectrophotometer-based assays, major enzymes associated with the EMP pathway, such as hexokinase, phosphoglucose isomerase, phosphofructokinase, fructose

1,6-biphosphate aldolase, glyceraldehyde-3-phosphate dehydrogenase, and pyruvate kinase were detected in H. pylori (72). Phosphoglycerate mutase activity has not been observed, although a gene coding for a protein with 44.6% similarity to the pgm gene

product has been identified. Analysis of the genome suggests that other genes coding

for the enzymes of the glycolytic pathway are present, with the important exceptions of

phosphofructokinase and pyruvate kinase (98). In spite of the apparent absence of the

gene, activity for phosphofructokinase was detected by one group, suggesting that this

enzyme either has a sequence unlike other phosphofructokinases, or the use of a

spectrophotometer as a means of assaying enzyme activity was not specific enough.

23 Glucose glucose-6- * phosphatase glucokinase Glucose-6-phosphate

phosphoglucose isomerase

Fructose-6-phosphate

fructose-1,6- phosphofructokinase * bisphosphatase Fructose-1,6-bisphophate

fructose-1,6-bisphosphate aldolase Dihydroxyacetone-phosphate

triose-phosphate isomerase

Glyceraldehyde-3-phosphate

glyceraldehydes-3-phosphate dehydrogenase

Glycerate-1,3-bisphosphate

phosphoglycerate kinase

Glycerate-3-phosphate

phosphoglycerate mutase **

Glycerate-2-phosphate

enolase Phosphoenolpyruvate phosphoenolpyruvate synthase pyruvate kinase * Pyruvate

Figure 1.3: The enzymes of the Embden-Meyerhof-Parnas pathway. Enyzmes marked with a * indicate those for which no gene has been identified in the H. pylori sequence, while the ** indicates an enzyme whose enzymatic activity has not been observed but whose corresponding gene was identified.

24 Entner-Doudoroff pathway

In addition to the pentose phosphate pathway, it was clear that at least one other

pathway must be operating in the utilization of glucose by H. pylori. When H. pylori

13 was incubated with [1- C]-glucose, the label at this carbon atom would be lost to CO2 in the oxidative phase if the pentose phosphate pathway was the only pathway utilizing this sugar. However, the appearance of the label in lactate indicated that a pathway other than the pentose phosphate pathway was active. For this reason, the presence of

Entner-Doudoroff enzymes was examined (119).

The Entner-Doudoroff pathway is an alternative to the Embden-Meyerhof-

Parnas glycolytic pathway. The ability to provide necessary metabolic precursors for biosynthesis as well as provide energy is present in both pathways. Overall, the two glycolytic pathways are similar: 6-carbon sugars are phosphorylated and subsequently cleaved by an aldolase enzyme into two 3-carbon intermediates. The difference between the two glycolytic pathways is due to the nature of the 6-carbon intermediates that are the substrates for aldol cleavage. The reactions involving further metabolism of the triose phosphate intermediates are shared by both pathways and provide energy via substrate level phosphorylation. The Entner-Doudoroff pathway (Figure 1.4) fulfills two vital roles of central metabolism by providing metabolic precursors for biosynthesis and energy via substrate level phosphorylation and/or respiration.

25 glucose-6-phosphate NADP+ glucose-6-phosphate dehydrogenase (zwf) NADPH

6-phosphogluconate 6-phosphogluconate dehydratase (edd)

2-keto-3-deoxy-6-phosphogluconate

KDPG aldolase (eda)

glyceraldehyde-3-phosphate pyruvate

Figure 1.4: The Entner-Doudoroff pathway. The two enzymes of this pathway, EDD and EDA, convert 6-phosphogluconate to glyceraldehyde-3-phosphate and pyruvate. In H. pylori, glucose-6-phosphate can be utilized by this pathway following conversion to 6-phosphogluconate by a third enzyme, ZWF.

In the first step of the Entner-Doudoroff pathway, 6-phosphogluconate dehydratase (EDD) catalyzes the dehydration of 6-phosphogluconate to form 2-keto-3- deoxy-phosphogluconate (KDPG). The second step of the pathway is an aldol cleavage of KDPG by KDPG aldolase (EDA) to form pyruvate and glyceraldehyde-3-P. The dehydratase (EDD) is unique and hence the key enzyme of the pathway, whereas the aldolase (EDA) is promiscuous and as such has other metabolic roles, including hexuronic acid metabolism and hydroxyproline biosynthesis (31). Because H. pylori utilizes this pathway for glucose rather than gluconate, it requires the presence of a third enzyme, glucose-6-phosphate dehydrogenase (ZWF), to convert glucose-6-phosphate into 6-phosphogluconate, which then enters into the pathway. 26 ZWF, EDD, and EDA have all been found to be present in H. pylori (118).

Furthermore, no evidence supported the idea that the glyceraldehyde-3-phosphate

resulting from the KDPG aldolase activity was metabolized to pyruvate by enzymes of

the Embden-Meyerhof-Parnas (EMP) pathway (118). As a result, the ED pathway is the proposed pathway for glucose catabolism in H. pylori.

Labeling patterns obtained are also consistent with the oxidation of glucose in

H. pylori by the Entner-Doudoroff pathway. The classical means for distinguishing

EMP metabolism from Entner-Doudoroff metabolism is the characteristic difference in

aerobic degradation of specifically labeled glucose as detected by radiorespirometry.

Using C1-labeled glucose, the labeled carbon is found in the carboxyl group of pyruvate

when metabolized via the Entner-Doudoroff pathway. The labeled carbon is

subsequently evolved in the form of CO2 during aerobic growth through the action of

pyruvate dehydrogenase, while the remaining two unlabeled carbons of the pyruvate

molecule enter the citric acid cycle. During EMP metabolism of C1-labeled glucose,

the methyl group of pyruvate is specifically labeled, enters the TCA cycle, and a portion

of the label is assimilated into the biosynthetic pathways. The remainder of the label is

released as CO2 during subsequent rounds of the TCA cycle. Thus, an early time course

of appearance of labeled CO2 from C1-labeled glucose serves as an indication of

Entner-Doudoroff metabolism (25).

The potential energy yield via the Entner-Doudoroff pathway is less than that

obtained by EMP glycolysis, but the Entner-Doudoroff pathway allows bacteria to

degrade aldonic acids such as gluconate. While it was thought that perhaps these acids

may represent important in vivo nutrients for H. pylori, the organism does not have any

27 known analogs of transporters or kinases for gluconate according to sequence

information now available. Alternatively, the Entner-Doudoroff pathway may represent

an efficient mechanism for the generation of pyruvate for synthetic purposes rather than a means of energy production via glucose metabolism.

Citric acid cycle

The citric acid cycle (CAC) has several functions as a metabolic pathway. It oxidizes acetyl units to product CO2, it generates reduced nucleotides useful for

reductive biosynthesis or for storage of energy in the form of ATP, and it provides

precursors for biosynthesis, such as oxaloacetate, succinyl-CoA, and α-ketoglutarate

(98, 133). As a result, the CAC is arguably the most important central metabolic

pathway in living cells. However, not all bacteria conform to the typical CAC pattern.

Pathogens, in particular, often have incomplete or otherwise unusual CACs (32, 77).

Early studies in H. pylori suggested that the CAC was either absent or

incomplete because glucose and pyruvate were oxidized to organic acids (119).

However, later experiments showed the presence of an active citrate synthase,

suggesting that acetyl CoA from the oxidation of pyruvate could be catabolized by the

CAC (72). More recently, NMR spectroscopy studies have shown that the CAC in H.

pylori is non-cyclic and branched (Figure 1.5), similar to that of anaerobic bacteria.

This cycle produces succinate in the reductive dicarboxylic acid branch and α-

ketoglutarate in the oxidative tricarboxylic branch, thus it is directed more towards the

generation of biosynthetic intermediates rather than metabolic energy (133).

28 Pyruvate

Acetyl-CoA

malate citrate synthase aconitase dehydrogenase Oxaloacetate Citrate

Malate malate Aconitate synthase aconitase fumarase glyoxylate Isocitrate Fumarate isocitrate dehydrogenase fumarate Succinate α-Ketoglutarate reductase α-ketoglutarate oxidase

Figure 1.5: The proposed citric acid cycle of H. pylori. Like that of anaerobic bacteria, this citric acid cycle is non-cyclic and branched, in which the reductive branch generates succinate while the oxidative branch generates α-ketoglutarate.

Amino acid utilization

The above discussion demonstrates that H. pylori has the capacity to take up and

metabolize glucose via multiple pathways (114, 118). However, carbohydrates are not

the only possible source of nutrition for the bacterium. H. pylori can grow in the

absence of glucose if amino acids are provided in the growth medium, suggesting that

H. pylori can utilize amino acids or peptides not only as a source of nitrogen but also as

a carbon and energy source (150).

The amino acid requirements of H. pylori were determined by two groups using defined media. The first group added different components to a defined tissue culture medium and found that omission of lipoic acid, FeSO4, Bovine Serum Albumin (BSA),

and non-essential amino acids impaired growth. Glucose was also found to be 29 necessary in this medium to support growth (140). The second group used a buffered

amino acid mixture as a solid medium with charcoal as a growth supplement. The

resulting growth showed no requirement for glucose, indicating the utilization of amino

acids as both carbon and energy sources (124). This was further confirmed in metabolic

studies by others (113, 151). Furthermore, it was found that amino acids could be used

as biosynthetic precursors for growth. Through these experiments, it was found that

amino acids produced at the end of biosynthetic pathways, including arginine, histidine,

isoleucine, leucine, methionine, phenylalanine, and valine, were required by the

organism.

Concentrations of amino acids in a liquid culture prior to and following an

extended incubation time were measured using 1H NMR spectroscopy (118). These

studies indicated that carbohydrates can be absent from a growth medium if amino acids are used as the basic nutrients. Furthermore, the resulting metabolites showed that many amino acids were utilized fermentatively, a mode of utilization typically seen in anaerobes. Anaerobic bacteria degrade amino acids by pathways involving oxidations and reductions. With the limitations imposed by the absence of oxygen or other high potential oxidants, the oxidation reactions in anaerobes are similar or identical to the corresponding ones catalyzed by aerobes. More distinctive are the reduction reactions

because the organism needs to generate electron acceptors of suitable potential from the

amino acids it can metabolize. Identification of the catabolic products in incubations with a single amino acid as the substrate indicated that fermentation was an important mode of amino acid utilization by H. pylori, and suggested a similarity between this part

30 of the organism’s physiology and that of anaerobes (118). Due to the microaerophilic

nature of H. pylori, it cannot grow in atmospheric oxygen and so some similarities to anaerobes may be expected.

Amino acids are an important source of carbon, nitrogen, and energy; however, the deamination of amino acids presents problems for the bacterium in the management of intracellular nitrogen balance. Biochemical evidence that H. pylori possesses a urea cycle of the type normally found in eukaryotes and a few other prokaryotes was found using one- and two-dimensional NMR spectroscopy and radioisotopic labeling. These experiments demonstrated the formation of ornithine and ammonium from L-arginine in bacterial lysates (117). The ornithine was then converted to citrulline by ornithine transcarbamoylase activity, and both arginosuccinate synthetase and arginosuccinase activities were also demonstrated. It was suggested that this cycle (Figure 1.6) may be involved in maintaining nitrogen balance in the cells, perhaps ridding the cells of excess nitrogen generated by the rapid catabolism of amino acids in the form of urea, which would be subsequently hydrolyzed by urease (117). Thus, urease together with a urea cycle may act to “pump” nitrogen from the cytosol into the external environment, helping to maintain an appropriate balance of cellular nitrogen.

31 Ammonium + CO2

urease Carbamoyl phosphate Urea Ornithine

Pi water ornithine arginase transcarbamoylase

Arginine Citrulline

argininosuccinate ATP + argininosuccinase synthetase Fumarate aspartate

AMP + PP Argininosuccinate i

Figure 1.6: Diagram of the urea cycle in H. pylori. Intermediate metabolites are shown in boxes and the enzymatic reactions of the cycle are shown at the corners with the arrows indicating direction of the reactions.

Colonization and Virulence Factors

Factors involved in colonization of the host have been an area of study that has received much attention in recent years. Two factors known to be involved in colonization are motility and urease activity. H. pylori is highly motile and uses its spiral shape and flagellae to migrate to areas that are only moderately acidic or neutral in pH. It has been shown that motility makes an important contribution to colonization ability. H. pylori generally produce 4-6 unipolar flagella, which are encased in a membraneous sheath and capped by terminal bulbs (59, 79). These bulbs are thought to be an extension of the sheath, which may protect the flagellar filament from depolymerization in an acidic environment (59). The flagellum has three main

32 components: filament, hook, and basal body. The flagellar filament is comprised

primarily of repeating subunits of two polypeptides, FlaA and FlaB (83, 87, 91, 152).

Flagellar-mediated movement of H. pylori through gastric mucus may be aided by the spiral shape of the organism (68). Although the movement of other motile rod-shaped organisms, such as E. coli is impaired in vitro under conditions of elevated viscosity, H. pylori cells remain motile (68).

In early studies conducted to determine the importance of motility to colonization it was found that motile variants of a pig-colonizing strain were able to colonize the gastric mucosa at a higher level and for longer periods of time than non- motile organisms (49). Later studies using strains with null insertion mutations in either flaA or flaB showed that these strains colonized piglets only in very low numbers and that the duration of infection was short compared to that of the wild type strains (51).

The low levels of colonization by flagellar mutants suggest that while flagella are not essential for viability, they may have an effect on the ability of the organism to persist in the mucosa.

Motility may function to aid colonization in several ways. It is likely that cell and mucus turnover cleared the non-motile organisms from the stomach. Also, the ability to colonize may be due to flagellar adhesion or that motile strains can more easily penetrate mucus (12, 21). Although only a small percentage of bacteria are found attached to the epithelial cells, they are able to adhere to specific receptors, e.g. glycolipids, present on the surface of epithelial cells (109). As a result, it is possible that bacteria that are unable to adhere to epithelial cells may use their flagella to maintain their location in the mucus (56).

33 Another factor involved in colonization of the host is urease activity. Urease is

the most characteristic and prevalent enzyme produced by H. pylori, constituting 5-10%

of total cell protein (75). Urease is a ~550 kDa nickel metalloenzyme composed of two

subunits, UreA (26.5 kDa) and UreB (60.3-61.6 kDa). Each urease macromolecule

contains six copies of each subunit (43, 75). H. pylori urease, like ureases from other

bacterial species, is a cytoplasmic enzyme (75). However, bacterial autolysis may

result in the release of urease, which can bind to the surface of viable bacteria or be

shed into the gastric mucosa (44, 132). Reported Km values range from 0.17-0.48 mM,

and so H. pylori urease is well suited to physiological gastric urea concentrations

(which range from 1.7-3.4 mM in the serum) and it is likely always saturated and working at maximum efficiency (43, 75, 122). As a result, it is believed that this

protein has a particularly important role in the survival of H. pylori.

Seven contiguous genes, ureABIEFGH, are transcribed in the same direction

from a 6.13 kb sequence and all of them, except for ureI, are necessary for the synthesis

of a catalytically active enzyme (29, 37, 90). The two structural subunits are encoded

by ureA and ureB. The UreA subunit of Helicobacter pylori is unique in that the

primary amino acid sequence encoded by the ureA gene is always encoded by two separate genes in all other bacterial species, as if the two smaller urease subunit genes of other species may have fused to form the H. pylori ureA. The active site of urease is found in the UreB subunit. Expression of ureA and ureB is sufficient to produce an assembled apoenzyme (73). Under these conditions, no nickel ions are inserted into the active site of the enzyme and it lacks catalytic activity.

34 For synthesis of a catalytically active urease, the accessory genes, ureE, ureF, ureG, and ureH must also be expressed. UreEFGH, by analogy to homologs from other species, serve to insert nickel ions into the apoenzyme (123). Two nickel ions are placed into the active site of each UreB subunit, yielding a total of 12 nickel ions per active macromolecule (121).

The ureI gene is unique to H. pylori, and encodes a urea-specific pore in the inner membrane (166). The pore is thought to be controlled by external pH via a shift in the periplasmic pH (139), and opens at low pH to allow passage of urea and closes at high pH to prevent access of the substrate to cytoplasmic urease (166). Although UreI is not necessary for the synthesis of active urease, it is essential for H. pylori survival in vivo and colonization of the stomach (148).

Unlike the ureases of most bacterial species, the H. pylori enzyme is not strictly cytoplasmic. In aging cultures, urease can be found adherent to the cell surface or shed into the medium (132). This appears to be due to lysis of a subset of the population and readsorption of the protein onto the cell surface of still-viable bacteria (99). The importance of external urease has been debated. In one study, bacteria with only cytoplasmic urease were more susceptible to acid (88), suggesting that surface exposed urease contributes to resistance to transient acid exposure. Another study argued that, because external urease is inactive below pH 5 and internal urease has its maximal activity below pH 5.5, internal urease is more likely to be responsible for acid tolerance

(144).

It has been suggested that urease serves to protect the organism from gastric acid

(150). Urea, the primary nitrogenous waste product of humans, diffuses freely from

35 plasma into the gastric juice and so is freely available to H. pylori. The hydrolysis of

urea produces ammonia, which may be released and directly neutralize the low gastric

pH. Ammonia is a preferred nitrogen source for bacteria, and is assimilated into protein

and other nitrogenous compounds. As previously mentioned, H. pylori is intolerant of

acid and cannot survive in vitro below a pH of 4.5 without the presence of urea (30). In

the presence of as little as 0.05 mmol/L urea, however, H. pylori can withstand a pH as

low as 1.5 (100). The rise in pH occurs within 10 minutes following the addition of

urea (30).

The mechanism by which urease aids colonization is still unknown. While

urease is not required for in vitro viability of H. pylori, it is clear that the enzyme is

necessary for colonization of the gastric mucosa and represents a critical virulence

factor. As stated above, it has been proposed that urease may function to protect the

organism from transient exposure to the acid in the lumen of the stomach until it

reaches the more alkaline gastric mucus (67). However, other studies have shown that

this is not the sole function of the enzyme. A urease-negative mutant, with 0.4% of the urease activity of the parental strain, was used to inoculate gnotobiotic piglets. The mutant strain could not colonize and no pathological lesions were observed in the piglets, while the parental strain both colonized and caused gastritis (46). However, the

urease-dependent colonization was independent of gastric acidity. In further elucidating the effect of urease on colonization, it was discovered that urease-negative H. pylori cannot colonize even in the absence of gastric acid. Additionally, the presence of urease-positive bacteria did not facilitate the growth of urease-negative bacteria in studies where these organisms were co-inoculated. Thus, urease-dependent

36 colonization is likely not affected by any diffusible factors such as ammonia or carbonic

acid from urea metabolism (48). These results suggest that urease has additional roles

necessary for colonization beyond that of raising the gastric pH in its microenvironment.

Urease may also serve as a means to obtain nitrogen (i.e., it may have a nutritive role) (150). It has been suggested that the enzyme may play a role in nitrogen metabolism by providing a nitrogen source for protein synthesis (67, 106). H. pylori

has the gene for glutamine synthetase, supporting the concept that urea-derived

ammonium ions can be added to glutamate to make glutamine. This, in turn, can be

directly incorporated into protein or converted into other amino acids (106). This has

been supported by studies utilizing labeled urea nitrogen in incubations with H. pylori

and finding that the label ultimately appears in protein (69). Additionally, the

expression of H. pylori urease genes in E. coli is regulated by nitrogen availability (37).

Therefore, a key role for urease may be a nutritional one, and in its absence the organism may starve for nitrogen. It has been suggested that the bacterium has a urea cycle (Figure 1.6), and this may allow tight control over nitrogen metabolism (117).

In vivo Assessment of Virulence Genes

Much of our knowledge about virulence determinants comes from experiments with bacteria grown in culture. In vitro assays have been useful and continue to provide much information on the mechanisms of bacterial pathogenesis, but they cannot accurately replicate all aspects of the host-pathogen interaction. In vivo conditions are

not only more complex than those in vitro, but can change at various points throughout

37 the infectious process, particularly in response to a developing host immune response.

Consequently, a gene that may seem important during in vitro studies may not be important in vivo, and genes that appear unimportant in an in vitro assay may play a critical role during a natural infection.

Upon entering the host, bacteria must be able to respond to a situation that likely differs significantly from the environment outside of the host. Patterns of gene expression must be modulated accordingly, down-regulating the expression of genes that are no longer necessary and up-regulating those that are specifically required for survival in the host. Bacterial adaptation is critical to growth within the host, and the interaction between a host and pathogen may not be easily assessed from in vitro studies alone. The production of any given virulence gene may be simultaneously modulated by a number of environmental and genetic signals. These signals provide precise anatomical and “disease state” information to the pathogen which, in turn, ensures the production of the appropriate set of virulence genes. Because of the inherent complexity of this process, bacterial responses during infection cannot be determined from in vitro studies alone. Host environments are extremely complex, and can change during the infectious process due to inflammation, breakdown of tissue, and immune responses of the host. Three main principles emerge from the fact that environmental conditions in vivo are not the same as those in vitro: 1) some candidate virulence determinants produced in vitro may not be produced during infection, 2) some virulence determinants produced during infection may not be produced in vitro, and 3) the complement of virulence determinants changes as the infection proceeds (149). As a result, the ability to study those factors that are truly critical to virulence has become

38 important and techniques have been developed for identifying potential genes that are

expressed within the host. Four methodologies have been developed to study in vivo

gene expression: 1) signature tagged mutagenesis, 2) differential display, 3) differential fluorescence induction, and 4) in vivo expression technology.

Signature tagged mutagenesis (STM) is a negative selective strategy that uses a collection of transposons, each of which is modified by the incorporation of a different

DNA sequence tag. When the transposon inserts into a bacterial gene, a tagged mutant is created. Mutagenized bacterial strains are stored individually in arrays (often in the

wells of microtiter dishes) and colony or dot blots are made from these arrays. The

animal host is then infected with a pool of these signature-tagged insertion mutants.

Mutants that are represented in the initial inoculum but that are not recovered from the

host may carry a mutation in a gene that is essential for survival or infection (27, 71).

PCR can then be used to prepare labeled probes representing the tags present in the

preselection (input) and postselection (output) pools. Hybridization of the tags from the

input and output pools to the colony or dot blots permits the identification of mutants

that were unable to survive the selective process. These strains can then be recovered

from the original arrays and the nucleotide sequence of DNA flanking the transposon insertion point can be determined (27).

The major advantage of STM is that it directly identifies genes involved in virulence rather than indirectly by in vivo expression. It does not depend on selection parameters and can be used to identify genes that are expressed transiently or at low levels. STM is relatively easy to adapt to many host-pathogen systems and is not dependent on preferential transcription in host tissues. That is, it does not measure

39 promoter activity, but measures the essential nature of the gene itself. However, mutants that are slow-growing, inviable, contain mutations in genes encoding redundant functions, or that can be complemented in a mixed population (e.g. by secreted factors) may be underrepresented (93).

A second method that has been developed, differential display (DD), is a

subtractive hybridization strategy in which bacterial cDNAs from infected tissue are hybridized against a cDNA library constructed from bacteria grown on laboratory medium. The resulting host-specific cDNAs are used as probes to isolate genes that are expressed in host tissues. DD can be used in pathogens lacking well-developed genetic tools, can be applied to two or more conditions simultaneously, and has the added advantage of being able to detect both up- and down-regulation of in vivo expressed

genes. Disadvantages include the instability of bacterial mRNA for the construction of

cDNA libraries, the low abundance of messages from transiently expressed genes, and

the difficulty in isolation of sufficient high-quality mRNA from small populations of

bacteria in vivo. PCR-based modifications may allow representation of low-abundance

mRNAs, but this may also generate false-positive results (93).

Other methods of detecting virulence genes in vivo are based on promoter

trapping. Differential fluorescence induction (DFI) is one of these promoter trap

methods. In this scheme, bacterial promoters drive the expression of green fluorescent protein (GFP), a highly sensitive reporter. GFP is a small molecule that remains fluorescent when fused to proteins of interest and, unlike beta-galactosidase and

luciferase, does not require a substrate or cofactor. DFI takes advantage of the high

throughput and semi-automation of fluorescence activated cell sorting (FACS) to carry

40 out the selection for active gene fusions by measuring intracellular fluorescence in

individual bacterial cells (160). FACS can be used to identify either fluorescent

bacteria or host cells containing fluorescent bacteria carrying transcriptional fusions to

the GFP reporter. DFI screens are not influenced by selection parameters, and thus

genes can be identified that are weakly and/or transiently expressed during infection.

However, flow cytometric analysis can be affected by aggregations of bacteria,

virulence genes and products that are preferentially regulated post-transcriptionally will

not be detected, and mutations in desired genes must be constructed to assess their role

in virulence (93).

In vivo expression technology (IVET) is another promoter trapping strategy in

which bacterial promoters drive the expression of a gene required for growth within the

host. IVET is further described in detail below.

In vivo expression technology (IVET)

The IVET strategy is a promoter trap that relies on bacterial promoters to drive the expression of a gene required for virulence. In the IVET scheme, the bacterial genome to be analyzed is first sheared into random fragments. The DNA fragments are cloned into an IVET vector 5’ to a promoterless reporter gene. This reporter gene must be one that is essential for colonization in the host. The vector is then transformed into a bacterial host that is deficient in the reporter gene activity. Because the plasmid cannot replicate in the pathogen, integration must take place by homologous recombination between the promoter-containing fragment and the corresponding region of the host chromosome. Transformants are pooled and used as the inoculum for the

41 host animal. Survival of the bacteria in vivo is dependent on activation of the reporter gene. Therefore, only strains carrying DNA fragments with promoters that are active in

vivo and drive expression of the promoterless selection gene will be recovered. After

incubation in the animal, bacterial strains are recovered from the host tissue and plated

on a selective medium to screen for in vitro activity. Genes of interest are those that are

inactive on the medium, indicating that they are active solely in vivo.

The original IVET selection system (Figure 1.7) was based on the fact that a

mutation in a biosynthetic gene can dramatically attenuate the growth and persistence of

a pathogen in host tissues. Growth of the auxotrophic mutant in the host can then be complemented by fusions to the same biosynthetic gene, thus selecting for those promoters that are expressed in host tissues. This technique was developed in

Salmonella typhimurium, using complementation of purine auxotrophy to identity in vivo induced (ivi) genes (95). S. typhimurium strains mutant in the purA gene are extremely attenuated in their ability to cause mouse typhoid or persist in animal tissues

(105). A transcriptional gene fusion library to purA was constructed in a mutant S. typhimurium strain in which the native purA was deleted. Because purine auxotrophs cannot survive within mice, only strains that carried a transcriptional gene fusion of an ivi gene to purA were able to complement the mutation and therefore survive and multiply within the mouse. Fusion strains that passed this selection were then screened for those that were transcriptionally inactive on normal laboratory medium. This subset of strains contained fusions to genes that are “on” in the mouse and “off” outside the mouse; that is, those that were specifically induced in the host.

42 purA lacZY genome fragments Bgl II pIVET plasmid

ligate fragments into Bgl II site

transformation or conjugation into host

integration into host genome

inject pool into mouse

incubation in mouse; selects for purA+ in vivo

recover bacteria

screen for Lac- in vitro

Figure 1.7: Positive selection of in vivo-induced genes by in vivo expression technology (IVET). The IVET plasmid in this scheme has two reporters, a promoterless purA gene and a promoterless lac operon, downstream of a BglII cloning site. Random genome fragments are ligated into the BglII site resulting in transcriptional fusions. The pool of fusions is introduced into a ∆purA host strain, where the plasmid must recombine into the genome. Bacteria carrying promoters that are active and permit expression of purA survive in the mouse. Recovered bacteria are then screened for a Lac- phenotype. Modified from Mahan et al (95).

43 Because the existence of an attenuating and complementable auxotrophy is not

readily available in all microbial systems, a variation on the basic principle was

established in which expression of the reporter gene provided resistance to an antibiotic,

which could then be administered to the host. Using this method, it is possible to carry

out selection for ivi genes in any tissue in which the antibiotic concentration can be

made sufficiently high to select against strains not expressing the resistance gene.

Adjustment of the antibiotic dosage may permit isolation of ivi promoters with different

levels of activity, and variation of the timing of antibiotic administration might allow

for the identification of genes expressed at a particular time or place during infection.

The first application of this modification was successfully utilized in S. typhimurium

using the chloramphenicol acetyl transferase (cat) gene as the reporter gene (96). The

technique has also been used successfully to identify ivi genes in Yersinia enterocolitica

(169).

Successfully utilizing IVET requires prior knowledge of specific reporter genes

whose expression is essential for survival within the host and that are required for growth in the same host compartment in which the putative pathogenicity gene(s) is

expressed. The main advantage of the preceding IVET variations (auxotrophic

complementation or antibiotic selection) is the use of positive selection to isolate ivi

gene fusions from a pool of fusion strains, thereby largely avoiding the labor-intensive

nature of individually screening for such loci. However, complementation in the animal

requires higher levels of gene expression compared to growth on laboratory medium

(94, 96) and genes that are highly expressed or required throughout infection may be

favored. One potential problem is that the selected genes may be required throughout

44 the entire infectious process, in which case enrichment will be for genes that are

expressed constitutively during infection, such as housekeeping genes. Thus, pathogenicity genes that are expressed only at specific stages during the infectious process or at low levels will likely be missed in this situation. Furthermore, although some genes that may be identified may not have an essential role in virulence per se,

they may still contribute to growth in restricted host tissues of compartments and therefore may play some role in pathogenicity or colonization.

As a result, a further modification of the IVET procedure was developed so that intermittently active in vivo promoters could be detected (18). This variation on the

technique is known as resolvase in vivo expression technology (RIVET) and uses

genetic recombination as the reporter activity. Instead of a gene essential to the host

survival, the RIVET system uses a promoterless site-specific recombinase gene

(resolvase) whose expression is of no consequence to the survival of the pathogen.

Resolvase expression is monitored by the ability to excise indicator genes located

between site-specific recombination sites artificially inserted elsewhere on the bacterial

chromosome. This excision is permanent and allows for the detection of promoter

activity even if the promoter is active only briefly during infection by inducing a

heritable change (i.e., conversion from antibiotic resistance to antibiotic sensitivity) that

can be detected by replica plating after the bacteria are recovered from host tissues.

Although RIVET does not have the benefits of positive selection, its main advantage is

that it can be used to identify genes that are transiently expressed during pathogenesis or

that are tissue-specific (27, 93).

45 In one of the first examples of the use of RIVET, a promoterless tnpR-lacZY

reporter was used to identify ivi genes in Vibrio cholerae (19). tnpR encodes a site- specific DNA resolvase from E. coli Tnγδ that mediates recombination between two directly repeated copies of specific target DNA sequences called res1 sites. res1 sites are inserted, flanking a reporter gene (in this case, a tetracycline resistance gene), within the chromosome of the pathogen of interest. If a particular tnpR fusion is transcriptionally induced during infection, even transiently or at low levels, the resolvase that is produced will catalyze a permanent and heritable change in the bacterium by excising the reporter from the chromosome (Figure 1.8). Resolved strains are then screened after recovering the bacteria from host tissues. This allows for the isolation of all ivi promoters regardless of the stage of infection or the duration of time during an infection over which they are active. In this way, all in vivo active genes, including those that are expressed only transiently, may be identified.

Regardless of whether IVET or RIVET has been used, a distinction between genes that are expressed constitutively and those that are not must be made following isolation of ivi genes from the animal. Constitutively-expressed genes may include housekeeping genes that, although essential for survival, are likely not critical for virulence. This can be achieved by screening the bacteria isolated from the animal on selective media to indicate which genes are up-regulated solely in vivo. The authenticity of the putative pathogenesis genes expressed in vivo must be subsequently verified by more classical techniques such as DNA sequencing, amino acid sequence comparisons, and mutation and deletion analysis.

46 Resolvase

geh res tetR res geh

tetR

geh res geh +

res

Tetracylcine -sensitive Non-replicating minicircle progeny

Figure 1.8: Schematic of the resolution event in RIVET. A gene cassette (res-tetR-res) that serves as the substrate for resolvase is placed at a neutral site in the bacterial genome. Fusions are made to a promoterless resolvase gene, such as tnpR. If an active promoter is integrated upstream of the promoterless tnpR gene, resolvase is produced and acts at the res sites, resulting in the excision of the tetR gene from the chromosome. The non-replicating minicircle is diluted out as cells divide and progeny of resolved reporter cells thus become tetracycline sensitive. geh = lipase gene; res = resolvase recognition sequences; tetR = tetracycline resistance gene.

Pathogens utilize many elaborate factors that make unique contributions to their

fitness during the infectious process. In vivo expression methods have allowed for the

identification of these functions in the context of the host. Many genes have been

identified using the various IVET strategies discussed from a variety of host/pathogen

systems (Table 1.2). The prevalent classes of ivi genes identified are involved in

nutrient acquisition, cell metabolism, and the stress response, suggesting that they may

47 Gene Protein function Contribution to Host site of expression Microbial species pathogenesis Nutrient acquisition and synthesis Metals entF siderophore Fe2+ uptake macrophage line Salmonella typhimurium fhuA; cirA; iron transport Fe2+ uptake mouse intestine; Salmonella typhimurium sitB macrophage; hepatocyte fyuA iron transport Fe2+ uptake Mouse Peyer’s patch Yersinia enterocolitica fptA iron transport Fe2+ uptake Mouse liver Pseudomonas aeruginosa pydD pyoverdin synthesis Fe2+ acquisition Rat lung; mouse liver Pseudomonas aeruginosa np20 Fur-like Fe2+ starvation response Mouse liver Pseudomonas aeruginosa mgtA; Mg2+ transport Mg2+ uptake Mouse spleen; macrophage Salmonella typhimurium mgtB iviX Cu2+ transport Cu2+ homeostasis Macrophage line Salmonella typhimurium Nucleotides carAB; pyrimidine; purine De novo requirement Mouse spleen Salmonella typhimurium purDL synthesis ndk nucleotide balance Alarmone synthesis Mouse intestine/spleen Salmonella typhimurium vacB; mRNA; tRNA Post-transcriptional Mouse spleen/intestine Salmonella typhimurium vacC processing regulation vacC tRNA processing Post-transcriptional Mouse Peyer’s patch Yersinia enterocolitica regulation Cofactors hemA Heme synthesis Peroxide resistance Mouse spleen Salmonella typhimurium hemD Heme synthesis Peroxide resistance Mouse Peyer’s patch Yersinia enterocolitica cobI Vitamin B12 synthesis Carbon source utilization Mouse liver Pseudomonas aeruginosa iviXVII Vitamin B12 synthesis Carbon source utilization Mouse spleen Salmonella typhimurium propanediol utilization Stress response and adaptation DNA repair recD Recombination and Macrophage survival Macrophage line Salmonella typhimurium repair recB Recombination and Macrophage survival Mouse Peyer’s patch Yersinia enterocolitica repair mutL DNA repair Intestine survival Mouse Peyer’s patch Yersinia enterocolitica Environmental otsA Trehalose synthesis Thermo/osmotic Mouse intestine Salmonella typhimurium protection rhi-12 IsfA-like Oxidative stress Sugar beet roots Pseudomonas fluorescens sodA Superoxide dismutase Oxidative stress Mouse hepatocyte line Salmonella typhimurium rhiI Glycine-betaine binding Osmotic protection Sugar beet roots Pseudomonas fluorescens cadC Cadaverine synthesis Acid tolerance Mouse intestine/spleen Salmonella typhimurium cadA Cadaverine synthesis Acid tolerance Rabbit intestine Vibrio cholerae fadB; hylC Secreted lipase; Fatty Clear proinflammatory Mouse spleen; mouse Salmonella typhimurium acid degradation fatty acids intestine Vibrio cholerae lip Lipid degradation Mouse kidney Staphylococcus aureus cfa; aas Membrane modifications Membrane repair Mouse intestine; spleen; Salmonella typhimurium macrophage cell ivi134-21 Membrane modifications Protein targeting Mouse liver; rat lung Pseudomonas aeruginosa phoP Virulence regulator Invasion/macrophage Mouse spleen Salmonella typhimurium survival agrA Accessory gene Virulence gene regulation Mouse kidney Staphylococcus aureus regulation vieB Response regulator Colonization Mouse intestine Vibrio cholerae pmrB Polymixin resistance Neutrophil survival Macrophage line Salmonella typhimurium hre-7 AcrR-like Efflux pump regulator Mouse Peyer’s patch Yersinia enterocolitica rhi-3 CopRS-like Copper inducible Sugar beet root Pseudomonas fluorescens regulator iviIV Signal receptor Chemoreception Mouse intestine Vibrio cholerae np9 Response regulator Chemoreception Mouse liver Pseudomonas aeruginosa iviG AraC-like Virulence gene regulation Rabbit heart Streptococcus gordonii

(Table continued on next page)

Table 1.2: Genes identified using IVET. Modified from Mahan et al. (93).

48 Table 1.2 (continued)

Microbe-host cell interactions and colonization SPI-2 Type III secretion system Invasion/systemic survival Mouse spleen; macrophage; Salmonella typhimurium hepatocyte rhi-18 Type III secretion system Colonization Sugar beet roots Pseudomonas fluorescens iviVI-A; Tia-like; PfEMP1-like Adhesion/invasion Mouse spleen Salmonella typhimurium iviVI-B ivi131-19 Hag2-like Adhesion/invasion Rat lung; mouse liver Pseudomonas aeruginosa secE Secretion Protein export Pig lung Actinobacillus pleuropneumoniae Intracellular and systemic survival spvRAD; Plasmid virulence Systemic survival Mouse spleen Salmonella typhimurium spvB SAP2 Secreted aspartic Disseminated candidiasis Mouse spleen Candida albicans proteases Bacterial cell surface structure rfb LPS synthesis Surface variation Mouse spleen Salmonella typhimurium pbp2 Penicillin binding protein Cell wall maintenance Mouse kidney Staphylococcus aureus pbp2 Peptidoglycan Cell wall assembly Mouse kidney Staphylococcus aureus crosslinking

contribute significantly to the fitness of the organism during infection. This “ecology of infection” is essential to understanding the disease process. The acquisition of metal ions and acquisition and/or synthesis of nucleotides and cofactors, together with genes involved in DNA repair and thermo, osmotic, and acid tolerance, can be critical in increasing the fitness level of a pathogen. Likewise, surface modifications may help shield the bacterium from the host immune response. The host ecology clearly serves to signal induction of bacterial genes that are suited to nutrient-limiting conditions, but it also signals the induction of other virulence genes required for immediate survival and spread to subsequent anatomical sites of infection.

Summary

Since its discovery, H. pylori has been heavily studied. Although much of this work has focused on the clinical side of infection and therapy, more basic questions have begun to be examined more recently. H. pylori is a unique organism in that it

49 colonizes only the stomach, which was previously believed to be uninhabitable. To do so, the bacterium has evolved various ways to deal with the acidity it is transiently exposed to during infection. To cause disease, however, the organism must not only be able to survive the acidity, but to effectively colonize the host. Several factors involved in colonization, such as motility and urease activity, have been found. Metabolic pathways, which have been less well studied than other bacterial characteristics, may also be important in colonization or virulence. This introduction has summarized some of the basic mechanisms whereby H. pylori survives and functions in vivo, including some known colonization factors as well as pathways for glucose metabolism.

50

CHAPTER 2

ANALYSIS AND CLONING OF THE ENTNER-DOUDOROFF GENES IN HELICOBACTER PYLORI AND CREATION OF A 6-PHOPHOGLUCONATE DEHYDRATASE MUTANT

ABSTRACT

Although H. pylori has the capability to utilize glucose, the pathways involved

and the role of specific enzymes are not well understood. For a bacterium with a

relatively small genome to continue to encode specific genes likely means that there is a

purpose for these genes at some point in the organism’s life cycle. The genes for the

Entner-Doudoroff (ED) pathway and an accessory gene necessary for catabolizing glucose via this pathway are all present within the H. pylori genome. Although much work has been done to prove which enzymes are present and functional, no studies have evaluated the specific role they play in the process of colonization.

The purpose of this study was to determine if 6-phosphogluconate dehydratase

(6PGD), the key enzyme in the ED pathway, has a role in colonization by H. pylori.

For this, the activity of the ED pathway in H. pylori was measured, the three genes of

the ED pathway were cloned from H. pylori and expressed in E. coli, and an ED- H.

pylori mutant was created. The ED pathway was active in the three H. pylori strains examined, and insertional mutation of the 6PGD gene eliminated activity. Cloned ED

51 genes were not expressed in ED-negative E. coli, suggesting that additional regulatory

mechanisms are necessary for the pathway to function.

INTRODUCTION

Insight into the physiology of H. pylori is critical to fully comprehend how this

organism colonizes and causes disease, and an understanding of the basic nutrients that

the organism utilizes for carbon and energy sources during an infection is central to gaining this knowledge. This information is not only of fundamental interest, but could also be of value in the development of novel therapies for H. pylori infections.

Over the past ten years, many advances have been made in the knowledge of H. pylori metabolism. Metabolic studies indicate, and the complete genome sequence of

H. pylori 26695 confirms, that the organism has a limited metabolic capacity and few regulatory networks (157). This relatively simple metabolism is compatible with its unique ecological niche in which there is little to no competition from other bacteria.

H. pylori was at first thought to lack the ability to metabolize carbohydrates (103, 111).

However, it was later discovered that the bacterium can use glucose as a growth substrate via oxidative and fermentative pathways (119). Complete pathways for carbohydrate metabolism identified in H. pylori include the Entner-Doudoroff (118) and pentose phosphate pathways (114). Although most of the Embden-Meyerhof-Parnas

(EMP) pathway enzymes are present, evidence shows that this pathway is not utilized by H. pylori (24). H. pylori also expresses the enzymes necessary for gluconeogenesis, an incomplete TCA cycle (133), and a urea cycle (117).

52 NMR spectroscopy has indicated that H. pylori most likely utilizes the Entner-

Doudoroff (ED) pathway (118, 119), for glucose metabolism. The ED pathway, an

alternative glycolytic pathway for sugar metabolism, is composed of two enzymes, 6-

phosphogluconate dehydratase (Entner-Doudoroff dehydratase, EDD) and 2-keto-3-

deoxy-6-phosphogluconate aldolase (Enter-Doudoroff aldolase, EDA), encoded by edd

and eda, respectively (31). EDD catalyzes the dehydration of the substrate, 6-

phosphogluconate (6PG) to form 2-keto-3-deoxy-6-phosphogluconate (KDPG). EDA

catalyzes an aldol cleavage of KDPG to form pyruvate and glyceraldehyde-3-phosphate

(G3P). Conversion of glucose to 6PG, the true substrate of the ED pathway, requires

the presence of a third enzyme, glucose-6-phosphate dehydrogenase, which is encoded

by zwf. Thus, glucose is activated by a glucokinase to glucose-6-phosphate (G6P),

which is then oxidized by EDD to form 6-phosphogluconate with the reduction of

NADP+ to NADPH. Metabolism of glucose to G3P thus requires all three genes, zwf, edd, and eda.

These three enzymes are present and functional in H. pylori. Bacterial lysates have been shown to contain dehydratase, aldolase, and dehydrogenase activity (118). In addition, the zwf, edd, and eda genes are homologous to the same genes in E. coli (157), supporting their potential role in H. pylori metabolism. In contrast, key enzymes of the

Embden-Meyerhof-Parnas glycolytic pathway appear to be missing from the H. pylori genome according to both NMR spectroscopy and sequence analysis methods (25, 157).

These findings give further indication that glucose is probably catabolized exclusively via the ED pathway in H. pylori.

53 The enzymes necessary for glucose metabolism are thus present in H. pylori, but

glucose metabolism is not necessary for survival. H. pylori can grow in a glucose-free

medium using only amino acids for substrates (113), demonstrating that carbohydrates

are not required if amino acids are supplied. Contrary to predictions based on genomic

analysis that glycolysis is the basis of energy production, it was found that glucose is

not utilized when amino acids are also present until the level of the other metabolites

has markedly decreased (115, 119). Despite this, the ED pathway has been conserved

in this organism. The role of the pathway in H. pylori is not yet known, but a direct role

for the ED pathway in E. coli has been demonstrated in colonization of a streptomycin-

treated mouse large intestine (154, 155), illustrating that this pathway can impact

colonization ability of other gut-associated organisms. Furthermore, there is

considerable evidence that E. coli grows in the mammalian large intestine by

metabolizing mucus-derived sugar acids via the ED pathway (129) and it is possible

that H. pylori similarly metabolizes gastric mucus. The purpose of this study was to determine whether the ED pathway is necessary for H. pylori survival and colonization

of the mouse gastric mucosa.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Three H. pylori strains, 26695, M6, and SS1, were used in this study. All three

strains were originally isolated from human stomachs and adapted for growth in piglets

(26695) or mice (SS1, M6). Bacteria were grown under microaerobic conditions (10%

o CO2, 5% O2) at 37 C on either blood agar plates (BAP, 5% sheep blood) or in Brucella

54 broth (BB) supplemented with 10% fetal calf serum (FCS). The appropriate antibiotics were surface plated when necessary at final concentrations of 20 µg/ml for kanamycin and 50 µg/ml for ampicillin. E. coli were grown aerobically at 37oC on either Luria

Burtani (LB) plates or in LB broth. The appropriate antibiotics were added when necessary by surface plating or before the plates were poured at final concentrations of

20 µg/ml for kanamycin and 50 µg/ml for ampicillin. When needed for color selection,

50 µl X-gal (2%) and 4 µl IPTG (100mM) were surface plated onto the agar plates.

Transformation of E. coli

Competent E. coli DH5α cells were prepared by the CaCl2 method (Maniatis,

o 1989). Briefly, 50 ml cultures were grown in LB at 37 C to an OD650 = 0.2-0.4

(approximately 2-3 hours). Cells were collected by centrifugation at 5000 rpm for 5 minutes at 4oC and the supernatant removed. Cells were resuspended in 10 ml cold

MgCl2 solution (50 mM) and placed on ice for 10 minutes. Cells were again collected by centrifugation as above and resuspended in 15 ml cold CaCl2 solution (50 mM) and incubated on ice for 30 minutes. Cells were then collected by centrifugation and gently

o resuspended in 5 ml cold CaCl2. Competent cells were either stored at 4 C and used within one week or were mixed 1:4 with glycerol (80%) and frozen at -80°C.

Isolation of plasmid DNA was performed using the QIAprep Spin Mini-prep Kit

(Qiagen) according to the manufacturer’s instructions. Plasmids were transformed into

E. coli by adding 1 µl plasmid DNA to 250 µl competent cells. Cells were placed on ice for 30 minutes and then heated at 42oC for 2 minutes. The mixture was held at room

55 temperature for 5 minutes and then 1 ml of LB was added. The cells were incubated at

37oC for 1 hour and the mixture was plated onto the appropriate selective medium and incubated overnight at 37oC.

Construction of plasmids

Plasmids for assaying H. pylori ED gene activity in E. coli

Each ED gene was cloned individually from H. pylori 26695 by PCR. Primers used for these reactions were eda1B, eda2, edd1B, edd2, zwf1B and zwf2 (Table 2.1).

Primers eda1B, edd1B, and zwf1B were designed with an EcoRI restriction site and primers eda2, edd2, and zwf2 were designed with a BamHI restriction site so that directional cloning could be done following isolation of the PCR products. Products of the correct sizes were obtained from PCR and the DNA was isolated from the agarose gel by using the QIAquick Gel Extraction Kit (Qiagen).

Primer Sequence eda1B GGCGAATTCaCTGAGGGAGCGTTAAATGGGG eda2 ATAGGATTCbACAAACTAATGGGCGAACGGC edd1B ACCGAATTCaATAGCTCCTTACGCCCCGATG edd2 AAAGGATCCbATAGCTCCTTACGCCCCGATG zwf1B CCGGAATTCaAATATTCGCGCAACCCAAGCT zwf2 CCCGGATCCbTTTGTAGCTCTTTGCCCCCAA kan 1 CGCCCWWGGcCAATTAACCCTCACTAA kan 2 GGACCWWGGcCTATAGGGCGAATTGGA kan 3 GCTAGGCACTTTCCCGCTCG kan 4 TCCTTTTCGCGCTAATGGGG a: EcoRI restriction site b: BamHI restriction site c: Sty I restriction site

Table 2.1: Primer sequences developed for cloning H. pylori genes or confirming plasmid constructs.

56 Each fragment was digested with EcoRI and BamHI and individually ligated into the multiple cloning site of pUC18. The ligation reactions were transformed into competent E. coli DH5α and cells were plated on LB + ampicillin + X-gal + IPTG plates to allow for color selection of colonies carrying recombinant plasmids.

Transformants were isolated and plasmid structures were confirmed by restriction

enzyme analysis and sequencing. Thus, each of the resulting plasmids, pAW107,

pAW108, and pAW109, carried one of the H. pylori ED genes, eda, edd, or zwf,

respectively. Each plasmid was transformed into an E. coli host strain suitable for each

gene. E. coli DF214 (zwf-, eda-) was used for pAW107 and pAW109, and E. coli

RW231R (edd-) was used for pAW108. The transformed strains were then used to assay H. pylori gene activity.

Plasmid for H. pylori mutant construction

To construct a plasmid with known sequence and restriction sites containing an

antibiotic resistance cassette, a 1.45 kb fragment containing the kanamycin resistance

gene (aphA-3) from Campylobacter coli was excised from pAT95 (158) using EcoRI

and HindIII. This fragment was inserted into the EcoRI/HindIII site of pBluescriptII

KS+, generating pAW102 (Figure 2.1A). Plasmid structure was confirmed by restriction enzyme analysis.

57 Figure 2.1: Plasmids constructed for H. pylori colonization studies in mice. ApR = ampicillin resistance, KanR = kanamycin resistance, eda = 2-keto-3-deoxy-6- phosphogluconate aldolase, edd = 6-phosphogluconate dehydratase, zwf = glucose-6- phosphate dehydrogenase.

To construct plasmid pAW103 (6.85 kb, ApR) (Figure 2.1B), the three genes necessary for a complete Entner-Doudoroff pathway in H. pylori were cloned from H. pylori 26695 utilizing PCR. Primers eda 2 and zwf 2 (Table 2.1) were designed so that all three genes (loci HP1099, HP 1100, and HP 1101) in the H. pylori genome were cloned as a single fragment with approximately 200 base pairs on either side of the coding region and flanked by BamHI restriction sites. Because of the length of the PCR 58 product and for higher fidelity, the eLONGase Enzyme Mix (Gibco BRL) was used in place of Taq DNA polymerase in the PCR reaction. The resulting 4165 bp gel-purified

PCR fragment was ligated into the BamHI site of pUC18. The ligation mixture was then transformed into E. coli DH5α and grown on LB + ampicillin + X-gal + IPTG plates. Cells carrying recombinant plasmids (pAW103) were selected by choosing white colonies and plasmid structure was confirmed by restriction enzyme analysis and sequencing.

A 1.45 kb fragment containing aphA-3 was generated using PCR with pAW102 as the template. Primers kan 1 and kan 2 were constructed so that StyI sites flanked the

PCR product (Table 2.1). 1 µl MgCl2, 2 mM dNTPs, 5 µl 10x PCR buffer, 10 µm each primer, 1 µl plasmid preparation as template, and 2.5 U Taq polymerase were used in the 50 µl PCR reaction. The PCR product and pAW103 were digested with StyI and ligated together, resulting in pAW105 (8.43 kb, KmR, ApR) (Figure 2.1C). The ligation mixture was transformed into E. coli DH5α and plated on LB + kanamycin + ampicillin. Colonies that grew on the plates were isolated and the plasmid was confirmed by restriction enzyme analysis and by PCR using the primers in Table 2.1.

Transformation of H. pylori/Mutant construction

Plasmid pAW105 (Figure 2.1C) was transformed into H. pylori M6 (a mouse virulent strain, kindly supplied by Dr. Steven Czinn at Case Western Reserve

University, Cleveland, OH) or SS1 (provided by Adrian Lee of the University of South

Wales, Sydney, Australia) to generate M6eddΩaphA-3 and SS1eddΩaphA-3, respectively. For this, bacteria were grown overnight in BB + FCS until the cells 59 reached mid-late log phase. Cells were diluted 1:100 in fresh BB + FCS, incubated for two hours, and 2.5 µg DNA from pAW105 was added to the broth. Bacteria were incubated overnight and then diluted 1:4 in fresh BB + FCS + kanamycin. The culture was grown overnight and 1 ml aliquots were plated onto BAP + kanamycin plates and

incubated for up to five days. The resulting colonies were pooled and streaked onto a fresh BAP + kanamycin plate and incubated for two days.

Allelic exchange and insertion of the disrupted gene into the genome was confirmed by the presence of a band of the predicted size via PCR analysis using primers kan 3 and kan 4 (Table 2.1) and template DNA from the putative mutants.

These primers are specific for portions of the edd gene and thus differentiate between transformants and non-transformants based on the size of the resulting band (Figure

2.2). The putative mutants were further confirmed by the loss of 6-phosphogluconate dehydratase activity (see below).

Complementation tests

Testing the ED genes by complementation was done by transforming E. coli

host strains deficient in the gene of interest with a plasmid carrying the H. pylori

version of the gene. Enzyme activity was measured by the host strain’s ability to use

particular sugar sources. The zwf-mutant was tested by its ability to grow on minimal

glucose (0.2%) plates and the eda-mutant was tested by its ability to grow on minimal

glucuronate (0.2%) plates. The edd-mutants were tested by color changes induced on

brom-thymol-blue + gluconate plates and tetrazolium + gluconate plates.

60

Figure 2.2: Confirming the generation of an EDD mutant by PCR. The two potential outcomes are A) no insertion of the kanamycin resistance cassette into pAW105 or B) successful insertion of the cassette. Location of the primers utilized in the PCR reaction in indicated by bars marked ‘kan3’ or ‘kan4’.

2-keto-3-deoxy-6-phosphogluconate aldolase (eda) assay

Bacteria were grown in a starter culture overnight and then subcultured to

OD550=0.05 in 30 ml. The culture was grown to mid-log phase (OD550=0.3-0.4) and

collected by centrifugation. Pellets were washed twice and resuspended in assay buffer

(50 mM MES buffer, 10 mM MgCl2⋅6H2O, pH 7.65) to a density of OD550=1.0. Two

ml of this cell suspension were pelleted in one tube and then resuspended in 500 µl

assay buffer. Cell-free extracts were prepared by sonication (5 second sonication on ice

followed by 10 second rest on ice, repeated 2-3 times or until sample cleared), followed

by centrifugation at 14,000 RPM for 30 minutes at 4oC to remove cell debris.

Supernatants were immediately assayed for enzyme activity spectrophotometrically.

61 Because eda is used in other cell reactions, a non-specific activity reading was taken by

adding 940 µl assay buffer, 10 µl NADH, and 1 µl LDH to a cuvette. The OD340 was read and then 50 µl of supernatant was added to the cuvette and the OD340 was read

again. Specific activity was measured in the same way, except that 940 µl of KDPG

substrate was used in place of the assay buffer. Change in absorbance was used in

calculating enzyme activity. Total protein in cell-free supernatants was determined by

the method of Lowry (92).

KDPG substrate is not commercially available. To make KDPG substrate for

this assay, a zwf-negative E. coli strain carrying a plasmid for expression of edd,

DF214( pTC120), was grown overnight and the density standardized to OD550=1.0 in

Tris buffer. Two ml of the culture was pelleted and then resuspended in 500 µl Tris

buffer. The suspension was sonicated (5 seconds sonication on ice and 10 seconds

resting on ice, repeated 2-3 times or until sample cleared) and cell debris removed by

centrifugation. The supernatant was transferred to 12.5 ml of 6-phosphogluconate (5

mM) and incubated for 30 minutes at 37°C. The reaction was then stopped by heat

inactivation for 5 minutes at 90°C. The supernatant was then frozen at -20°C and used

as needed.

Glucose-6-phosphate dehydrogenase (zwf) assay

Bacteria were grown in a starter culture overnight and then subcultured to

OD550=0.05 in 30 ml. The culture was grown to mid-log phase (OD550=0.3-0.4) and

collected by centrifugation. Pellets were washed twice and resuspended in assay buffer

(50 mM MES buffer, 10 mM MgCl2⋅6H2O, pH 7.65) to a density of OD550=1.0. Two

62 ml of this cell suspension was pelleted in one tube and resuspended in 500 µl assay

buffer. Cell-free extracts were prepared by sonication (5 second sonication on ice

followed by 10 second rest on ice, repeated 2-3 times or until sample cleared), followed

by centrifugation at 14,000 RPM for 30 minutes at 4oC to remove cell debris.

Supernatants were immediately assayed for enzyme activity spectrophotometrically by

adding 900 µl glucose-6-phosphate (2 mM) and 50 µl NADP to a cuvette and reading the absorbance at 340 nm. 50 µl of cell lysate was then added to the cuvette and the absorbance read again. An increase in absorbance signifies activity. Total protein in cell-free supernatants was determined by the method of Lowry (92).

6-phosphogluconate dehydratase (edd) assay

For testing activity in E. coli, cells were grown in a starter culture overnight and then subcultured to OD550=0.05 in 30 ml. The culture was grown to mid-log phase

(OD550=0.3-0.4) and collected by centrifugation. Pellets were washed twice and

resuspended in assay buffer (50 mM MES buffer, 10 mM MgCl2⋅6H2O, pH 7.65) to a

density of OD550=1.0. Two ml of this cell suspension was pelleted in one tube and

resuspended in 500 µl assay buffer. Cell-free extracts were prepared by sonication (5

second sonication on ice followed by 10 second rest on ice, repeated 2-3 times or until

sample cleared), followed by centrifugation at 14,000 RPM for 30 minutes at 4oC to remove cell debris.

For assays in H. pylori, bacteria were grown overnight in BB +FCS to mid-log

phase and collected by centrifugation. To evaluate induction of 6-phosphogluconate

dehydratase, glucose (2%) was added to the growth medium. Pellets were washed

63 twice and resuspended in 500 µl of assay buffer (50 mM MES buffer, 10 mM

MgCl2⋅6H2O, pH 7.65). Cell-free extracts were prepared by sonication (5 second

sonication on ice followed by 10 second rest on ice, repeated 2-3 times or until sample

cleared), followed by centrifugation at 14,000 RPM for 30 minutes at 4oC to remove

cell debris.

Supernatants from E. coli and H. pylori were immediately assayed spectrophotometrically for enzyme activity following sonication using a two-step

reaction (Figure 2.3) in which pyruvate formed from 6-phosphogluconate is titrated

with lactate dehydrogenase to determine the total enzyme activity (52, 55). For this

assay, 50 µl of supernatant was placed in an microcentrifuge tube with 50 µl assay

buffer. To maintain the stability of the enzyme, one µl glutathione (300 mM) was

added, the reaction was incubated for 3 minutes, and then one µl FeSO4 was added and

the reaction incubated for 10 minutes. One µl 6-phosphogluconate (700 mM) was then

added and the reaction was allowed to incubate for 15 minutes at 37°C. At this point,

900 µl assay buffer prewarmed to 90°C was added to the reaction, and then the tube was

incubated at 90°C for 3 minutes to stop any reactions. After cooling to room

temperature, the solution was placed in a cuvette. Ten µl NADH (15 mM) was added

and the absorbance read at 340 nm. One µl LDH was added and then absorbance was

read again. Change in absorbance indicated enzyme activity. Total protein in cell-free

supernatants was determined by the method of Lowry (92).

64 cell extract+ 6-phosphogluconate edd (if present)

2-keto-3-deoxy-6-phosphogluconate

eda

pyruvate glyceraldehyde-3-phosphate

heat NAD ldh A340 NAD

lactate

Figure 2.3: The 6-phosphogluconate dehydratase enzyme assay. This is a coupled, two-step assay in which the presence of edd is determined by the production of pyruvate as shown in the above reaction (eda is present in excess in the cell and does not affect the reaction). The presence of pyruvate is detected by the addition of lactate dehydrogenase (LDH) and measuring the change in absorbance as NADH is generated in the production of lactate.

Enzymes, chemicals, and media

Restriction endonucleases, T4 ligase, Taq polymerase, eLONGase, and dNTPs were obtained from Gibco BRL. Chemicals were from Sigma. Media were from Gibco

BRL, with the exception of the blood agar plates that were from BBL. Primers were from Ransom Hill Biosciences and Integrated DNA Technologies, Inc.

Statistics

Enzyme assays were conducted in triplicate except where noted. The three trials were averaged and the standard deviation was calculated. The numbers obtained were

65 then analyzed with the Mann-Whitney and 1-way ANOVA tests using the GraphPAD

Instat statistical analysis package (GraphPAD Software).

RESULTS

Enzyme assays in E. coli

Each ED gene was cloned individually so that it could be assayed in E. coli.

Two host strains were used for these tests. E. coli DF214 (zwf-, eda-) was the host strain used for pAW107 and pAW109 and E. coli RW231R (edd-) was used for pAW108. The first method was to use simple complementation assays in which the activity from the H. pylori gene carried on the plasmid would complement for the mutated gene in the host strain. Each gene could be tested by its use of a specific sugar added to an agarose plate. In each case, the controls behaved as expected: the wild-type strains grew while the mutant strains did not grow on the plates. E. coli mutant strains transformed with E. coli genes on a plasmid were used as positive controls and were able to grow. E. coli mutant strains carrying pAW107, pAW108, or pAW109 did not grow. Thus, the ED genes that were cloned separately were unable to complement the mutation.

Mixed results were seen in the spectrophotometric assays. As expected, activity was observed from the wild-type E. coli for all three ED genes. However, activity was relatively low or undetectable when the mutant E. coli strains were carrying pAW107, pAW108, or pAW109 (Table 2.2). Minimal activity was seen with pAW107 (eda) as compared to the wild-type E. coli activity, but was not significantly different from that of the mutant strain. Activity for zwf from pAW109 was approximately 4-fold lower

66 than that measured in E. coli. Activity from pAW108 (edd), however, was undetectable. When RW231R was transformed with a plasmid carrying the E. coli edd gene (pTC180) for use as a positive control, activity was detected at very high levels.

Enzyme activity (nmol/min/mg) Strain zwf eda edd 1485a 401.7 ± 45.84 91.0 ± 0.40 16.1 ± 1.61 DF214b 0.0 0.0 n/a DF214(pAW107) n/a 3.4 ± 5.89 n/a DF214(pAW109) 113.1 ± 85.19 n/a n/a RW231Rb n/a n/a 0.0 RW231R(pTC180)c n/a n/a 971.3d RW231R(pAW108) n/a n/a 0.0 a: wild-type b: negative control c: pTC180: Bluescript + 3.1 kb E. coli fragment with edd gene d: only one trial conducted

Table 2.2: Entner-Doudoroff enzyme activity measured in E. coli wild-type and mutant strains.

6-phosphogluconate dehydratase assay in H. pylori

All three wild type strains of H. pylori tested exhibited low but detectable levels of 6-phosphogluconate dehydratase activity (Figure 2.4). The average activity among

all strains tested, 4.43 +/- 0.86 nmol/min/mg total cell protein, was significantly lower

than the activity in E. coli, which averaged 16.07 +/- 1.61 nmol/min/mg total cell protein (p=0.0001). Glucose was added to the growth medium for H. pylori strains

26695 and M6 to test for induction of the pathway in the presence of its substrate, but 6- phosphogluconate dehydratase activity in the strains tested was not significantly induced by glucose and thus the activity appears to be constitutive (p=0.35 for 26695, 67 p=0.2 for M6). Insertion of aphA-3 into edd in H. pylori strain M6 to generate

M6eddΩaphA-3 eliminated all detectable 6-phosphogluconate activity (Figure 2.4, p=0.05); thus, interruption of the edd gene successfully eliminated expression of the enzyme activity.

18 *

16 no glc, no 6PG no glc, 6PG 14 glc, no 6PG 12 * glc, 6PG 10

8 * 6 * * * * 4

2 * * * * * 0 26695 M6 SS1 M6 EDD- E. coli 1485 Bacterial strain

Figure 2.4: 6-phosphogluconate dehydratase activity ± standard deviation of three wild- type H. pylori strains, the EDD- mutant (M6eddΩaphA-3), and E. coli 1485 (wild-type). H. pylori cultures were grown in either the presence or absence of 0.2% glucose (glc) (M6 and 26695 only) and/or 6-phosphogluconate (6PG) (all H. pylori strains). E. coli broths also included 0.2% gluconate, which is necessary to induce the ED pathway in E. coli. For all wild-type strains, the differences in growth in the absence or presence of glc +/- 6PG were significant (p=0.05). Growth in the absence of glc versus the presence of glc was not significant for H. pylori strains 26695 and M6 (p=0.35, p=0.2, respectively). The difference between 6-phosphogluconate dehydratase activity in the wild-type H. pylori strains and the mutant strain was significant (p=0.05). * denotes significant difference (p<0.05) between groups with and without the addition of 6PG.

68 DISCUSSION

H. pylori is unusual with its relatively small genome and its lack of many of the

metabolic pathways that are crucial for the survival of other gram-negative enteric

bacteria. Thus, understanding the role of metabolic pathways that are retained by H.

pylori is likely to reveal physiological features relevant to infection and may lead to the

development of novel therapies. Existing knowledge of H. pylori metabolism is

expanding, but there have been few studies that attempt to discover specific roles for

individual metabolic processes in colonization or in vivo survival.

The ED pathway is found in numerous organisms, including several pathogens

(31, 129). In E. coli, the ED pathway is inducible and is used to catabolize gluconate

(52). It allows E. coli to colonize the mammalian large intestine (154, 155) and it has

been suggested that oxidative glucose metabolism via the ED pathway may also permit

E. coli to survive in aerobic, aquatic environments. Thus, the ED pathway plays an

important role in the ecology of E. coli, including both colonization of host animals and survival in the environment (129).

The current study was undertaken to determine whether the ED pathway has a

role in the growth of H. pylori in vitro. Analysis of enzyme activity was first attempted

in E. coli with mixed results. Activity was not detected for strains transformed with the

cloned H. pylori edd. Minimal activity was detected with the H. pylori eda gene,

although not significant. Activity was detected from the H. pylori zwf gene, although

the levels obtained were approximately 4-fold lower than that seen in the wild-type E.

coli strain. This lower activity of the H. pylori gene may be due to the fact that glucose may not be a critical nutrient source for the organism. Failure of activity of H. pylori

69 enzymes in E. coli is not without precedence. For example, groups studying H. pylori

urease in E. coli had difficulty achieving wild-type activity levels. However, in time it was found that specific conditions, such as the overexpression of urease structural subunits, adding exogenous NiCl2, and culturing the bacteria in minimal medium that

contains no Ni2+-chelating agents was necessary for maximal recombinant urease activity in E. coli (74). Thus, lack of H. pylori enzyme activity in E. coli may indicate that additional control mechanisms or experimental conditions may be necessary for expression or detection in E. coli.

In H. pylori, low but detectable EDD activity was present in all three strains studied, indicating that the edd gene and the ED pathway are likely to be conserved within the species and may serve some essential function. Although the ED pathway was not induced by glucose in our studies, the fact that it is not essential for in vitro survival suggests that it may have a supportive role in metabolism, necessary only under specific (and as yet undefined) growth conditions. The low enzyme activity that we detected in H. pylori is consistent with reports from other groups suggesting that glucose is not a major energy source for the organism.

The role of the ED pathway in H. pylori is not known. Our studies show that, in contrast to E. coli, the pathway is not inducible in H. pylori. Because H. pylori is limited in its ability to utilize sugars and can only utilize glucose, the capability to quickly metabolize the sugar whenever it is present may be a factor in the constitutive nature of the pathway. Alternatively, it has been proposed that the ED pathway may allow the organism to utilize aldonic acids such as gluconate (118). Gluconate is present in mouse cecal mucus (129) and thus may be present in the stomach and could

70 represent an energy source for H. pylori in the absence of competition from other bacteria. However, paralogs for gluconate transporters have not been found on the H. pylori genome (data not shown), and the use of this pathway for gluconate metabolism is as yet unproven and unlikely unless novel gluconate transporters are found in H. pylori.

71

CHAPTER 3

COLONIZATION OF THE MOUSE GASTRIC MUCOSA BY HELICOBACTER PYLORI STRAINS DEFICIENT IN THE ENTNER-DOUDOROFF PATHWAY

ABSTRACT

Construction of the edd mutant is the first step in determining functionality of the pathway in H. pylori. To determine whether this pathway is necessary for colonization, studies were conducted utilizing a mouse model. The following work uses the mutant created in the previous chapter for colonization studies to determine whether impairment of the Entner-Doudoroff Pathway affects colonization of the mouse gastric mucosa.

Two mouse strains, C57BL/6 and germ-free recombinase activating gene

(RAG)-knockout 129 mice, were inoculated with wild-type and/or mutant H. pylori strains M6 or SS1. All mutant and wild-type strains colonized, but colonization was lower for both mutant strains compared to the parental strains. In addition, minimum infectious dose was 100-1000-fold lower for the wild-type than for the ED-negative mutants. Surprisingly, in spite of lower colonization density and higher minimum infectious dose, co-inoculation experiments revealed that wild-type H. pylori did not displace the mutant strain, indicating that competition between wild-type and mutant did not occur in vivo.

72 Results from this study indicate that loss of the key ED enzyme in H. pylori diminishes fitness of the organism in vivo, but that the pathway is non-essential for colonization. However, conservation of this pathway in H. pylori and the fact that colonization was diminished by loss of 6PGD, suggests that the ED pathway has some function in H. pylori metabolism.

INTRODUCTION

The enzymes necessary for glucose metabolism are present in H. pylori, but glucose metabolism is not necessary for survival. H. pylori can grow in a glucose-free medium using only amino acids for substrates (113), demonstrating that carbohydrates are not required if amino acids are supplied. Predictions based on genomic analysis suggested that glycolysis is the basis of energy production. It was found, however, that glucose is not utilized when amino acids are also present until the level of the other metabolites has markedly decreased (115, 119), and it is likely that the ED pathway is the means by which glucose metabolism occurs (see Chapter 1).

The role of the ED pathway in H. pylori is not known, but experiments in other enteric bacteria suggest that it may be active in vivo. For example, a direct role for the

ED pathway in E. coli has been demonstrated in colonization of a streptomycin-treated mouse large intestine (154, 155), illustrating that this pathway can impact colonization ability of other gut-associated organisms. Furthermore, there is considerable evidence that E. coli grows in the mammalian large intestine by metabolizing mucus-derived

73 sugar acids via the ED pathway (129). The purpose of this study was to determine

whether the ED pathway in H. pylori is necessary for colonization of the mouse gastric

mucosa.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Two H. pylori strains, M6 and SS1, were used in this study. Both strains were

originally isolated from human stomachs and adapted for growth in mice. Bacteria

o were grown under microaerobic conditions (10% CO2, 5% O2) at 37 C on either blood

agar plates (BAP, 5% sheep blood) or in Brucella broth (BB) supplemented with 10%

fetal calf serum (FCS). The appropriate antibiotics were surface plated when necessary

at final concentrations of 20 µg/ml for kanamycin and 50 µg/ml for ampicillin.

Mouse colonization

H. pylori wild-type and mutant strains were grown overnight in BB + FCS with

or without kanamycin as needed and concentrated 10x. C57BL/6 mice from Jackson

laboratory, germ-free C57BL/6 mice from our own colony, or germ-free recombinase-

activating gene (RAG)-knockout 129 mice from our own colony were orally inoculated

with 109 cfu of live H. pylori and sacrificed at either two weeks or ten weeks after inoculation. Stomachs were removed and homogenized, and serial dilutions were plated on blood agar plates with or without kanamycin. The number of bacteria per gram of gastric mucosa was determined by colony count. Bacteria recovered from the mice were cultured and tested by PCR to confirm that the bacteria colonizing the gastric

74 mucosa were those used in the inoculations (wild-type and/or mutant) (data not shown).

All procedures involving mice were approved by the Ohio State University ILACUC.

Histopathology Hematoxylin and eosin (HE)-stained sections were scored for the extent of gastritis as previously described (50). Briefly, microscopic fields containing gastritis severe enough to displace glands, and fields containing neutrophilic infiltrate were enumerated and expressed as a percent of the gastric mucosa. All available fields were scored and slides were examined blind (without knowledge of their source).

Statistics

Results of the colonization trials (cfu/g gastric mucosa) were averaged for each group and the standard deviations were calculated. If the colony count was detectable but below the level of quantification, the number utilized for statistical analysis was the highest possible (e.g., 1000 assumed if the lowest level of detection was 103). Data

were analyzed with the Mann-Whitney non-parametric test from the GraphPAD

package.

Enzymes, chemicals, and media

Chemicals were from Sigma. Media were from Gibco BRL, with the exception

of the blood agar plates that were from BBL.

75 RESULTS

Thirty-one C57BL/6J mice and eight RAG-KO 129 mice were inoculated with

M6, fifteen with the parental wild-type strain and sixteen with the mutant strain,

M6eddΩaphA-3. In addition, ten C57BL/6J mice were inoculated, half with SS1 wild-

type and half with the SS1eddΩaphA-3 strain. Two weeks after inoculation, mice were

sacrificed and the number of bacteria (cfu/g gastric mucosa) was determined (Table

3.1). In C57BL/6J mice, both wild-type and mutant M6 H. pylori strains were able to colonize, although the mutant strain had a lower colonization density (p<0.05). The

SS1 strains, however, showed a greater difference in colonization ability. The wild-type strain colonized at a high density (up to 107 cfu/g gastric mucosa) while the mutant

colonized at an average of 102 cfu/g (p<0.0001). In RAG-KO 129 mice, the M6 wild-

type strain colonized to a higher density than the mutant, but in this case differences in colonization densities were not significant (p>0.05).

Average cfu/g gastric mucosa x # mice infected/ 106 (mean ± SD) # mice colonized Parental Wild-type Mutant p value Wild-type Mutant bacterial strain C57BL/6J C57BL/6J SS1 69 ± 9.23 0.0001 ± 0 p<0.0001 5/5 5/5 M6 1.5 ± 2.06 0.21 ± 0.23 p<0.05 15/15 13/16 RAG-KO 129 RAG-KO 129 M6 1.0 ± 1.28 0.0011 ± 0.00214 p>0.05 3/4 2/4

Table 3.1: Colonization of mice by H. pylori M6, SS1, or the EDD- mutant. Bacteria were inoculated separately.

76 The effect of inoculation dose was determined using C57BL/6J mice. Forty

mice were inoculated with 10-fold dilutions of SS1 or SS1eddΩaphA-3 (twenty mice

per strain, five mice per inoculation group, Table 3.2). Likewise, forty-eight mice were

inoculated with M6 or M6eddΩaphA-3. Forty of these mice were C57BL/6J mice while the remaining eight were RAG-KO 129 mice. Prior to the sacrifice date, one mouse from the wild-type SS1 group died. Mice were sacrificed two weeks post-

inoculation and cfu/g gastric mucosa determined. As shown in Table 3.2, the lowest

infectious dose for SS1eddΩaphA-3 was at least 2x105 cfu/g. This strain did not

colonize mice given less than 2x105 bacterial cells per inoculum. Wild-type SS1, however, colonized mice even at the lowest inoculation dose of 200 bacteria. The lowest infectious dose for wild-type strain M6 was higher than that of SS1 (2x104), and not all mice inoculated at this dose were infected. The infectious dose for mutant strain

M6eddΩaphA-3 strain was the highest. This strain was unable to colonize well even when mice were given 2x105 bacteria, although most mice become colonized with this

strain when given a concentrated dose (>108 cfu). Colonization results in germ-free

RAG-KO mice paralleled results in C57BL/6J mice (data not shown). Of four mice

inoculated with 2x105 bacteria of M6, 2 became colonized, while the mutant

M6eddΩaphA-3 only colonized one out of four mice.

77 Inoculum dose H. pylori strain 2 x102 2 x 103 2 x 104 2 x 105 or mutant SS1 5/5* 5/5 5/5 5/5 SS1EDD- 0/5 0/5 0/5 1/4 M6 0/5 0/5 2/5 4/9 M6EDD- 0/5 0/5 0/5 2/9 *: # mice colonized/# mice inoculated

Table 3.2: Effect of inoculum dose on colonization of H. pylori in C57BL/6 mice.

Finally, the effect of a mixed inoculation containing high levels (109 cfu) of both

wild-type and mutant strains was determined (Table 3.3). Five C57BL/6J mice were

inoculated with a mixture of SS1 and SS1eddΩaphA-3, ten C57BL/6J mice were inoculated with a mixture of M6 and M6eddΩaphA-3, and four germ-free RAG-KO 129 mice were inoculated with a mixture of M6 and M6eddΩaphA-3. In C57BL/6J mice,

both SS1 and SS1eddΩaphA-3 colonized all mice inoculated, but the wild-type

colonized at a two-fold higher density (p<0.05). In C57BL/6J mice, M6 colonized all

mice, while M6eddΩaphA-3 colonized only two of ten mice (p<0.05). In RAG-KO 129

mice, H. pylori M6 wild-type colonized, although to a very low density, and M6

mutants were unable to colonize (p>0.05).

78 Average cfu/g gastric mucosa x # mice infected/ 106 (mean ± SD) # mice colonized Parental Wild-type Mutant p value Wild-type Mutant bacterial strain C57BL/6J C57BL/6J SS1 97 ± 54.6 0.2 ± 0.34 p<0.05 5/5 5/5 M6 1.9 ± 2.31 0.05 ± 0.11 p<0.05 10/10 2/10 RAG-KO 129 RAG-KO 129 M6 0.023 ± 0.0356 0 ± 0 p>0.05 3/4 0/4

Table 3.3: Colonization of mice by a mixture of wild-type and mutant H. pylori M6, SS1, or EDD- mutants.

To evaluate the role of EDD in gastritis, ten C57BL/6J mice were inoculated with H. pylori M6, five with the wild-type and five with the mutant strain, and were sacrificed ten weeks following inoculation. At sacrifice, bacterial densities were

2.40x105 ± 1.77x104 and 9.5x104 ± 1.1x104 cfu/g gastric mucosa for the wild-type and mutant strains, respectively. The difference in densities between the two strains was not significant (p>0.05).

Histologic examination revealed no differences in the extent of gastritis in mice infected with M6 eddΩaphA-3 compared to mice infected with M6. Gastritis was absent in mice killed two weeks post-inoculation and mild to moderate in mice killed ten weeks post-inoculation.

DISCUSSION

The current study was undertaken to determine if the ED pathway has a role in

H. pylori colonization of the mouse gastric mucosa. The results of our studies indicated

79 that while the ED pathway is not essential for colonization, it does confer an added advantage by increasing fitness in vivo. In our high-colonizing strain, SS1, the EDD- negative mutant colonized significantly less densely than the wild-type SS1. In addition, the difference in fitness was clearly indicated by differences in infectious dose.

The minimal infectious dose for SS1eddΩaphA-3 was at least three logs higher than the

infectious dose for wild-type bacteria, demonstrating that although EDD-negative bacteria colonize, they do so less readily than does the wild-type strain. These differences could not be detected in the less fit mouse colonizing strain, M6, possibly because the wild-type does not colonize as well as does SS1, and therefore differences in fitness are not as evident.

Interestingly, co-colonization studies indicated that, in spite of decreased fitness

for initial colonization, there is no evidence of competition for colonization sites in vivo

between wild-type and EDD-negative H. pylori. In mice co-challenged with a large dose of both SS1 and SS1eddΩaphA-3, both strains colonized all mice. Although colonization density by SS1eddΩaphA-3 was less than colonization density by SS1, the difference was similar to colonization densities in mice challenged with each strain separately. Colonization by the poorly colonizing strain, M6, was insufficient to allow interpretation of co-colonization studies for this strain.

Overall, our results indicate that EDD is not essential for colonization but does confer an advantage for initial colonization of mice. In contrast, EDD does not appear necessary for competition with other strains, since co-colonization with the wild-type strain did not diminish colonization rate or density of the mutant. The explanation for this lack of competition is not clear, but it may reflect the multifocal distribution of H.

80 pylori in vivo. It is likely that once strains colonize a specific site in the gastric mucosa, they are not displaced by other strains. Alternatively, the mutant and wild-type bacteria may colonize different anatomic regions of the stomach, as has been shown for co- colonizing H. pylori strains (3). Finally, our results suggest that loss of EDD does not diminish the longevity of H. pylori infection. Colonization density of M6eddΩaphA-3 was not statistically different in mice killed 10 weeks after inoculation than in mice killed 2 weeks after inoculation.

In addition to providing information about the role of EDD in colonization by H. pylori, our findings are applicable to future studies using genetically altered H. pylori in mice. We found significant differences in colonization patterns depending on both the parental H. pylori strain and the mouse strain. In C57BL/6 mice, strain SS1 was clearly the better and more consistent colonizer. Thus, differences between mutant and wild-type strains were clearly defined and statistically significant. In contrast, differences between M6 and M6eddΩaphA-3 were smaller and did not reach significance in some experiments. The mouse strain also affected the colonization rate.

The same parental strain, M6, colonized 129 RAG-KO mice better than C57BL/6 mice, thus suggesting that the 129 RAG-KO mouse strain is preferable for evaluation of mutant effects. The explanation for this host difference is not addressed in this study, but previous studies indicate that immune-deficient mice support greater colonization density than do normal mice (45). These results suggest that both parental bacterial strain and mouse strain should be chosen with care when evaluating colonization mutants of H. pylori in animal models.

81 Because the absence of EDD activity does not prevent colonization of mice by

H. pylori, it is likely that glucose is not an essential carbon or energy source in vivo.

However, decreased colonization demonstrates by an EDD-negative H. pylori suggests that glucose catabolism may have a supplementary role in H. pylori colonization.

While the ED pathway does contribute to H. pylori survival in vivo, its precise role in metabolism remains unclear. It may serve as a “reserve system” for scavenging carbon or producing energy when few other sources are available. For example, when amino acids are not prevalent the organism may be able to utilize glucose, a commonly available substrate. Further work in this direction may help elucidate details on the importance of this pathway in colonization and survival of H. pylori.

82

CHAPTER 4

CREATION OF AN IN VIVO EXPRESSION TECHNOLOGY (IVET) SYSTEM FOR USE IN HELICOBACTER PYLORI

ABSTRACT

One characteristic of genes that are important for colonization and virulence is

that they may be inactive in vitro but up-regulated in the host. The purpose of this study was to develop a promoter-trapping scheme to identify H. pylori genes that are up- regulated in vivo. These genes would represent a likely target for novel means of therapy or vaccination. Promoter-trapping for genes expressed in vivo has not been used in H. pylori but has been successful in several other organisms (70, 164, 169), and our preliminary data indicate that it will be effective in H. pylori as well. Genes

identified via this method may then be compared to one another, and potential gene

families necessary for virulence can be delineated.

H. pylori urease, the most highly expressed protein in H. pylori, is essential for

colonization (46). We have shown that the ureB gene, under the control of an unrelated

promoter, can restore colonization ability to an ureB-negative H. pylori mutant (47).

This knowledge has been used to develop a promoter-trapping strategy to screen the H.

pylori genome for promoters that are up-regulated in vivo. We hypothesize that 1) there

are at least several virulence genes that are expressed only when the organism is within

83 the host, 2) that we will be able to detect several promoters with this system, and 3) that the associated genes will be directly identifiable by sequencing and comparison to the two completely sequenced H. pylori strains.

The H. pylori strain used in this study was M6∆ureB. This mutant, lacking a functional urease enzyme, is unable to colonize the mouse gastric mucosa. An IVET plasmid was constructed for this project. DNA fragments can be inserted upstream of a promoterless ureB gene carried on the plasmid. This plasmid is a suicide vector in H. pylori, and thus must integrate into the genome by homologous recombination. H. pylori homologous sequences flanking the ureB gene allow the plasmid to recombine with the bacterial chromosome and ensure that transformants will contain the plasmid sequences inserted in a specific location on the H. pylori chromosome.

Methods were developed for shearing the genomic DNA for insertion into the pHIVET plasmid and for in vitro testing of the bacteria that survive the mouse challenge. Finally, the urease promoter was cloned and ligated into the promoter- trapping plasmid to test for functionality of the plasmid.

INTRODUCTION

Understanding the physiology of pathogens during an infection and the pathogenic mechanisms they use at various stages of infection is an area of important research. Bacterial pathogenicity is a multi-faceted and ever-changing process.

Although each pathogen has its own unique means of interacting with its host, there are several events that are common during infection. These are 1) attachment to host cells or tissue, 2) invasion into host cells or tissues (intracellular pathogens) or increased

84 adherence (extracellular and intracellular pathogens), 3) avoidance of and/or resistance

to the host immune system, 4) nutrient acquisition, 5) multiplication, and 6) leaving the host for new host or reservoir. It is at one of these stages that pathogens damage the

host via toxins or host-induced immune response (20). These events occur in many

cases because of the activation or up-regulation of specific genes in vivo. The ability to

regulate gene expression throughout the course of an infection is important for the

survival of a pathogen in the host, since the expression of a virulence factor may not be

advantageous to survival during the entire course of the infectious process. In vitro

studies of these up-regulated genes are different because of the complexity of the host

environment. Parameters such as temperature, osmolarity, pH, source of nitrogen,

concentration of iron, sugars, and amino acids are known to affect the regulation of virulence genes, and effectively replicating the correct balance of these factors as they occur in vivo may be difficult in the laboratory setting.

There is great interest in knowing which genes are specifically expressed by bacterial pathogens during the pathogenic process. Early studies of bacterial virulence factors relied on randomly selecting genes that were predicted to have an important role

in colonization and/or pathogenesis and stimulating expression of these factors in vitro, which may or may not be fully accurate. This type of approach was inefficient in that novel genes could be easily overlooked and each individual gene must be tested separately. Although these types of studies have yielded results, their use in developing a complete picture of the factors involved during the course of an infection are limited by the inability to fully reproduce the often complex and changing environment present within the host. Building on the observation that virulence genes expressed in vivo may

85 be quite different from those expressed by the same pathogen during growth in the

laboratory, methods have been developed to identify genes that are expressed solely or

at an up-regulated level in vivo. Furthermore, these techniques have the advantage of

screening large numbers of genes at once.

One such method is in vivo expression technology (IVET). IVET is a strategy

designed to identify genes that are transcriptionally active only during infection. Using

an IVET vector, potential promoter-containing DNA fragments from pathogenic

microorganisms are cloned upstream of a promoterless reporter gene, and the in vivo

expression profile of this gene is monitored. For this system to be effective, the reporter

gene must be essential for pathogenic growth in the host ensuring that only clones containing a functional promoter in the cloning site will survive. The bacterial test strain carries a deletion for the corresponding gene to permit complementation. The novel feature of IVET is the use of an animal model to induce the expression of virulence genes, rather than attempting to duplicate the in vivo milieu in the laboratory.

In effect, the animal is used as a selective medium to identify these in vivo-specific

genes.

Because few genetic techniques are available for studying H. pylori, the

identification of virulence genes has been difficult. One of the most highly sought after

results of H. pylori research is to identify genes that are up-regulated in vivo and thus

could represent a potential target for novel therapy or vaccination. The rising rate of antimicrobial-resistant strains (66, 110), the emergence of multi-drug-resistant strains

(38), and the development of adverse effects to treatment (28) are important causes of treatment failures and pose challenges for the successful treatment of infection.

86 However, molecular techniques and genetic manipulation in H. pylori, such as IVET, has lagged behind that of other, well-studied organisms, and thus this goal has not been reached.

The aim of this study was to develop an IVET technique for H. pylori. For the

IVET system to work, an appropriate reporter gene that has a measurable phenotype for

the particular bacterium under study must be selected. The IVET system developed

here (Figure 4.1) is based on the fact that H. pylori requires urease for colonization.

The reporter plasmid contains a promoterless urease gene. Insertions upstream of this

reporter gene containing a promoter that is up-regulated in vivo will allow for the

production of urease. The plasmids containing putative promoters are transformed into

a urease-negative H. pylori strain and used to infect mice. If an active promoter is

present, the transformant will colonize. Transformants carrying an inactive promoter

will not colonize. The bacteria surviving the mouse challenge will be harvested and

tested for urease activity in vitro. Those bacteria carrying promoters that lack activity

in vitro (i.e., the promoters are not constitutively expressed) will be selected for further

study. In this way, we can select for chromosomal promoters that are active during the

infection.

87 pHIVET

Random genome ligate fragments Isolate plasmids Transform H. pylori M6∆ureB

Inoculate mice Pool CmR transformants

Constitutively expressed (+) Æ disregard Sacrifice mice Up-regulated in vivo Æ Isolate colonies Isolate promoter insert(s) Screen for urease activity (-) Identify related gene(s)

Figure 4.1: Proposed IVET scheme for H. pylori. Random fragments of the H. pylori genome will be generated and ligated into the pHIVET plasmid to form a promoter library. The library will be used to transform an H. pylori strain that is deficient in ureB activity, and transformants will be pooled and used to inoculate mice. After sacrifice, bacteria isolated from the gastric mucosa will be tested for urease activity. Those colonies that are urease negative in vitro (but were urease positive in vivo, thus surviving the mouse challenge) are presumed to contain a promoter that is up-regulated in vivo. These clones will be chosen for further study.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Two H. pylori strains, M6 and M6∆ureB, were used in this study. M6 was originally isolated from a human stomach and adapted for growth in mice. In the

M6∆ureB strain, the ureB open reading frame was replaced with a kanamycin resistance cassette resulting in a urease-negative, non-colonizing phenotype (47). H. pylori were

88 o grown under microaerobic conditions (10% CO2, 5% O2) at 37 C on either blood agar plates (BAP, 5% sheep blood) or in Brucella broth (BB) supplemented with 10% fetal

calf serum (FCS). The appropriate antibiotics were spread on plates or added to broth

when necessary at final concentrations of 20 µg/ml for chloramphenicol or 50 µg/ml for ampicillin. E. coli were grown aerobically at 37oC on either Luria Burtani (LB) plates

or in LB broth. The appropriate antibiotics were incorporated into or spread onto plates

at final concentrations of 50 µg/ml for ampicillin or chloramphenicol. When needed for

color selection, 50 µl X-gal (2%) and 4 µl IPTG (100mM) were spread onto the agar

plates.

Cutting genomic DNA

Various methods to cut genomic DNA were analyzed to determine which

method was most efficient in obtaining fragments of approximately 500-1000 bp. Both

endonuclease digestion (KpnI, MboI, or Sau3AI) and physical methods (vortexing and

sonication) were tested. For the restriction digests, 1 µl genomic DNA, 1 µl restriction

enzyme, 2 µl appropriate buffer, and 16 µl ddH2O was incubated at 37°C for 2 hours.

For vortexing, 1 µl genomic DNA was added to 29 µl ddH2O and vortexed at high

speed for 2 minutes. For sonication, 1 µl genomic DNA was added to 29 µl ddH2O and sonicated for 10 seconds on ice with a Virsonic 60 sonicator (The Virtis Company,

Inc.).

89 Urease Test Agar (UTA)

To make the Urease Test Agar (UTA) plates, urease broth (2 g urea, 10 ml

phenol red, 0.157 g Na2HPO4·12H2O, 0.08 g NaH2PO4·2H2O, 0.02 g NaN3, in 100 ml total volume, pH 6.3-6.5 with HCl) was filter-sterilized and added to 1.5% molten agar in water following sterilization and cooling to 57°C. The volume of urease broth tested ranged from 100 µl to 500 µl per 25 ml agar.

Ligations

Ligations were performed using T4 DNA ligase (Gibco BRL) according to the manufacturer’s specifications (4 µl 5x ligation buffer, 1 µl T4 DNA ligase (1 unit), 30 fmol p324, and 90 fmol insert DNA in a total volume of 20 µl and incubating at room temperature for 5 minutes). The ligation reactions were then diluted 5-fold prior to transformation as recommended by product literature. Alternatively, the vector and insert were mixed together and ethanol precipitated. After resuspension in ddH2O,

ligase buffer and enzyme were added and the ligation was allowed to proceed overnight

at 16°C.

Treatment of plasmid DNA with cell-free extract

H. pylori cell-free extract (CFE) was prepared by centrifuging an overnight

culture grown in BB + FCS (10%) to collect the cells. Cells were washed two times

with fresh BB, and resuspended in one ml of BB. Cells were placed on ice, and

sonicated in 5 bursts of 5 seconds with 10 seconds resting on ice between bursts. The

suspension was centrifuged at high speed in a table top centrifuge to remove cell debris

90 and the supernatant collected for use as CFE. Samples of the CFE were used in a

bicinchoninic acid (BCA) protein assay (Sigma) to determine the amount of protein in the CFE.

Plasmid DNA (12 µg), prepared from E. coli DH5α using the Midi Prep Kit

(Qiagen), was combined with H. pylori CFE (300-400 µg protein) in a 200 µl reaction

containing 20 mM Tris-acetate (pH 7.9), 50 mM potassium acetate, 5 mM Na2EDTA, 1

mM DTT, and 200 µM SAM (Sigma). Samples were incubated at 37°C for one hour

and then extracted with phenol-chloroform (97). Plasmid DNA was precipitated with

ethanol and resuspended in 10 mM Tris-HCl (pH 7.5) to a final concentration of 0.5

mg/ml.

Transformation of H. pylori

Method one (Broth method):

Bacteria were grown overnight in Brucella broth (BB) + fetal calf serum (FCS,

10%) until the cells reached mid-late log phase. Cells were diluted 1:100 in fresh BB +

FCS, incubated for two hours, and 0.5-1.0 µg plasmid DNA was then added to the broth. Bacteria were incubated overnight and then diluted 1:4 in fresh BB + FCS + chloramphenicol (Cm) (20 µg/ml). The culture was grown overnight and 1 ml aliquots were plated onto blood agar plates (BAP) + Cm (20 µg/ml) and incubated for up to five days. The resulting colonies were pooled and streaked onto a fresh BAP + Cm and incubated for two days.

91 Method two (Plate method):

Alternatively, H. pylori cells were scraped from overnight cultures grown on

BAP and suspended in 500 µl BB. Then, 25 µl of the cell suspension was spotted onto a BAP and 1 µg of CFE-treated plasmid DNA was added. Plates were incubated overnight and then the spots were transferred to BAP + Cm (20 µg/ml) plates and incubated for 3-7 days.

Transformation of E. coli

For transformations of DH5α or JM109, competent cells were thawed on ice and

100 µl of cells were aliquoted into pre-chilled tubes. 0.8 µl β-mercaptoethanol was added to each tube. The cells were incubated on ice for 10 minutes. Various amounts of DNA, ranging from 0.1-50 ng, were added to the tubes for transformation. For the transformation control, 1 µl of pUC18 was added to another tube. Tubes were incubated on ice for 30 minutes, heat-pulsed in a 42°C water bath for 45 seconds, and incubated on ice for 2 minutes. 0.9 ml of preheated SOC medium (20 g tryptone, 5 g yeast extract, and 0.5 g NaCl per liter ddH2O; autoclave, then add filter-sterilized 10 ml

MgCl2 (1 M), 10 ml MgSO4 (1 M), and 20 ml glucose (20%)) was added to the tubes, and then they were incubated for 1 hour at 37°C with shaking at 225-250 rpm. Aliquots were then plated onto LB plates with the appropriate antibiotics. The control transformation was plated onto LB+Ap plates. All plates were incubated overnight at

37°C.

For transformation of ultracompetent E. coli XL-10-kan cells, cells were thawed on ice and 40 µl aliquoted into pre-chilled tubes. 1.6 µl XL10-Gold β-mercaptoethanol

92 mix (Stratagene, provided in kit) was added and incubated on ice for 10 minutes. 2 µl

of the ligation reaction were added and the tubes incubated on ice for 30 minutes. The

reactions were then heat pulsed for 30 seconds at 42°C and incubated on ice for 2

minutes. Preheated (42°C) NZY+ broth (10 g NZ amine (casein hydrolysate), 5 g yeast

extract, and 5 g NaCl per liter ddH2O, pH 7.5 with NaOH; autoclave, then add filter sterilized 12.5 ml MgCl2 (1 M), 12.5 ml MgSO4 (1 M), 20 ml glucose (20%) prior to

use) (or, alternatively, SOC broth) was then added to the tubes and the tubes were incubated at 37°C for 1 hour with shaking at 225-250 rpm. 200 µl or less of the transformation reaction was plated onto LB-Cm-IPTG-X-gal plates and incubated

overnight at 37°C.

Construction of pHIVET

p342, a urease reporter plasmid (Figure 4.2) (84), was used to construct two potential IVET plasmids. For the first plasmid, the gfp gene was amplified from the

pGFPmut3.1 plasmid (Clontech) via PCR using primers GFP-F and GFP-R (Table 4.1)

that were designed with a StuI site to allow for insertion into p342. Both p342 and the

cloned gfp DNA segment were digested with StuI following standard procedures. DNA fragments were purified using the QIAquick PCR Purification Kit (Qiagen) and used in ligations. The ligation reactions were used in transformation of E. coli JM109 competent cells, the host strain recommended for the pGFPmut3.1 plasmid.

Alternatively, ligations were transformed into E. coli DH5α competent cells.

93

rrnB terminators StuI hpn ureB

H. pylori genomic DNA

CmR

p342 ~11.1 kb ColE1 H. pylori genomic DNA ApR

Figure 4.2: Plasmid p342, a ureB reporter construct. A promoterless ureB gene with its native Shine-Dalgarno sequence was introduced, along with a chloramphenicol resistance gene, into the hpn locus. In order to block any transcription originating upstream of ureB, DNA encoding tandem rrnB transcriptional terminators was inserted just downstream of the hpn coding sequence [Joyce, 2001 #315]. Genomic DNA flanking the hpn gene is included in the plasmid to allow for homologous recombination of the ureB:cat construct into the genome.

94 Primer Sequence GFP-F GCAGGCCTaGCTATGACCATGATTACGCC GFP-R GCAGGCCTaGGTCAGCTAATTAAGC GFP-2F AAGAATTCbGCTATGACCATGATTACGCC GFP-2R CGCGAATTCbGGTCAGCTAATTAAGC LacProF AAGCATCTGCAGcCACGACAGGTTTCCCGAC LacProR TTCGATGGGTACCdTTGGCGTAATCATGGTCA Hpn-1 ATGGTAATGAAGAACAGCACGGCG LacPF2 GATGGTACCdCAGGACAGGTTTCCCGAC LacPR2 GGGGGTACCdTTGGCGTAATCATGGTCA hpn2F AACCTTCTTCATGATGAGAG hpn2Ra CTTTAAGAGACAACCACACT hpn2Rb AGCCAGTGAGCTAGAATTTA hpn3 CAACCTTCTTCTTGATGATG ureB1 TCAATGTGAGCGGTAGTG ureAB-F GTTTTTATGGGGGTTTTTTT ureAB-R TAGCCAATTCTCCAGCATAG Pab-F CCCGATATCeTTTCCTTATTTTTGGCGTGG Pab-R CGGGATATCeTGCCTTTTTCTTTGCGTTGT Pab-intF CGTGGGCGTTTTATTGTTG Pab-intR TTTTGGGGTGAGTTTCATCTT Hivet/ AB-F TCACCACCACACACACCAC Hivet/ AB-R CCACGCCCACATCTATCAT a: StuI restriction site b: EcoRI restriction site c: PstI restriction site d: KpnI restriction site e: EcoRV restriction site

Table 4.1: Primer sequences developed for IVET project.

Due to unsuccessful ligations, the gfp segment was re-cloned using different primers, GFP-2F and GFP-2R (Table 4.1). These primers were designed so that they contained an EcoRI site instead of the StuI site. The cloned PCR fragment was digested with EcoRI and the overhanging ends generated by the enzyme were then filled in using the Klenow fragment (Large Fragment of DNA Polymerase I, Gibco BRL) according to manufacturer’s specifications (3 µl React 2 buffer, 0.5 mM dNTPs, 1 µg digested DNA,

0.5 U Klenow enzyme, ddH2O to 30 µl total volume). The reaction was incubated at room temperature for 15 minutes and then heated to 75°C for 10 minutes to inactivate

95 the enzyme. The reaction mixture was then purified using the Qiagen PCR Purification kit. This insert fragment was utilized in subsequent ligation reactions. Transformation of E. coli was performed as described above. The recombinant plasmids were checked via PCR for the presence of the insert using primers GFP-2F and GFP-2R (Table 4.1).

The plasmids were also checked for insert orientation using primers GFP-2R and Hpn-1

(Table 4.1). The resulting plasmid that contained gfp in the p342 StuI site and in the same orientation as ureB was named p342GFPmut3.1.

The second IVET plasmid was constructed by removing portions of p342 to create a slightly smaller plasmid. FspI, which cuts p342 into four pieces (Figure 4.3) was used to digest p324. The digest was run onto an agarose gel, and the two larger fragments (2.2 kb and 6.1 kb) were isolated from the gel using QIAquick Gel Extraction

Kit (Qiagen). These two fragments were then ligated together. The ligation reaction was transformed into E. coli and the recombinant plasmid was isolated. The plasmid was checked by restriction digestion with FspI and StuI to ensure that the expected restriction patterns were present, and the plasmid was further checked for the presence and correct orientation of the hpn gene and upstream genomic region and the ureB gene using primers hpn2F, hpn2Ra, hpn2Rb, hpn3, and ureB1 (Table 4.1). This plasmid was called pHIVET.

96 rrnB terminators

StuI hpn ureB FspI

H. pylori genomic DNA CmR

p342 2.2 kb ~11.1 kb

H. pylori ColE1 genomic DNA

FspI FspI ApR

FspI

1.2 kb 1.6 kb

Figure 4.3: Plasmid p342 map showing the expected FspI restriction sites. FspI cuts the plasmid into four segments of approximately 1.2, 1.6, 2.2, and 6.1 kb. The segments of 1.2 and 1.6 kb would be discarded and the remaining fragments ligated together to form the IVET plasmid.

Testing IVET plasmids

Both IVET plasmids were tested using known promoters. p342GFPmut3.1 was tested using the lac promoter (Plac). pGFPmut3.1 (Clontech) carries Plac and was used

as the template in the PCR reaction to clone this promoter using primers LacProF and

LacProR (Table 4.1). The primers were designed with a PstI site and a KpnI site,

respectively, to permit for directional insertion into the multiple cloning site of the

p342GFPmut3.1. The resulting PCR product and p342GFPmut3.1 were digested with 97 PstI and KpnI and ligation reactions and transformation were performed as described

above. Upon further testing, it was found that PstI digests p342GFPmut3.1 in more

than one location, and therefore another restriction enzyme was chosen as a replacement

and new primers were designed for cloning Plac. The new primers, LacPF2 and

LacPR2, were both designed with a KpnI restriction site (Table 4.1). Following PCR cloning of the lac promoter, ligations and transformations were performed. Any colonies obtained from transformations were replica plated onto LB+Ap+Cm+IPTG plates, grown overnight at 37°C, and checked for fluorescence by UV light.

pHIVET was tested using the ureAB promoter (PureAB). The promoter was

cloned from H. pylori M6 using primers ureAB-F and ureBA-R (Table 4.1) and Pfu

DNA polymerase (Stratagene) according to the manufacturer’s protocol. Due to

inefficient amplification, the protocol was modified to include MgCl2. The amount

required (2 µl of 50 mM MgCl2 per 50 µl reaction) was determined by titration of the

PCR reaction with various amounts of MgCl2. The PCR product was extracted and

purified from an agarose gel using the Qiagen Gel Extraction Kit (Qiagen). Because of low product yield, the PCR reactions were performed numerous times, and the products pooled and concentrated by ethanol precipitation.

pHIVET was digested with StuI and dephosphorylated with calf intestinal alkaline phosphatase (CIAP, Gibco BRL) according to the manufacturer’s protocol.

PureAB was treated with T4 polynucleotide kinase (Gibco BRL) according to

manufacturer’s protocol. Ligations and transformation were performed as described

previously.

98 Due to difficulties with blunt-end ligations, the PureAB fragment was re-cloned

using new primers, Pab-F and Pab-R (Table 4.1). These PCR reactions were conducted

using Taq DNA polymerase. The new primers were designed so that they contained an

EcoRV restriction site, which results in blunt ends but could also be used in the

following subcloning experiment. The PCR product was cloned into the pPCR-Script

Cam SK(-) vector from the PCR-Script Cam Cloning Kit (Stratagene) according to the

manufacturer’s protocol. Briefly, the PCR product was polished using Pfu DNA

polymerase and ligated into the provided cloning vector. Ultracompetent E. coli XL-

10-kan cells were used for the transformation.

The recombinant plasmid was isolated from the transformants using the

QIAprep Spin Miniprep Kit (Qiagen) and digested with EcoRV. The PureAB fragment

was isolated from an agarose gel using the QIAquick Gel Extraction Kit (Qiagen). pHIVET was digested with StuI, dephosphorylated with CIAP, and used in ligation

reactions with PureAB. The ligations were transformed into E. coli DH5α. Plasmids

were checked for the presence of the insert by PCR using primers Pab-intF and Pab- intR (Table 4.1). The plasmids were also checked for the correct orientation of the insert using primers Hivet/AB-F and Hivet/AB-R (Table 4.1).

“Shotgun cloning” with p342

p342 was digested with StuI and dephosphorylated using CIAP. H. pylori M6 genomic DNA was sheared by sonication and run on a gel. Bands from 500-1000 bp were purified from the gel using the QIAquick Gel Extraction Kit (Qiagen). The ends of the sheared DNA were filled in using the Klenow enzyme according to the

99 manufacturer’s protocol. The ligation reaction was set up as previously described and transformed into E. coli DH5α or H. pylori M6∆ureB cells. E. coli were plated on the

LB+Ap plates and incubated overnight at 37°C. H. pylori were plated on BAP+Ap and incubated for 3-5 days at 37°C in microaerobic conditions.

Enzymes, chemicals, and media

Restriction endonucleases, T4 DNA ligase, T4 polynucleotide kinase, CIAP,

Klenow fragment, Taq polymerase, Pfu DNA polymerase, and dNTPs were obtained from Gibco BRL. Chemicals were from Sigma. Media were from Gibco BRL, with the exception of the blood agar plates that were from BBL. Primers were from Integrated

DNA Technologies, Inc.

RESULTS

Cutting genomic DNA

Generation of the random 500-1000 bp fragments of H. pylori genomic DNA was optimized by testing several restriction enzymes and physical shearing methods

(Figure 4.4). KpnI was used because the insertion site into the plasmid is a KpnI site and this would be the easiest insertion. However, the cutting pattern observed did not provide bands of the appropriate size. MboI also did not provide appropriate fragment sizes. Vortexing the samples did shear the genome, but the fragments obtained were larger than desired. Both digestion with Sau3AI and sonication yielded a good size distribution of bands. Sonication was chosen as the method to shear the genome for the

100 1 2 345678 9 101112 13 14 15

1000

500

Figure 4.4: Agarose gel of various genomic DNA cutting techniques. Lanes 1-3 are DNA from 26695, SS1, and M6 cut with KpnI. Lanes 4-6 are DNA from 26695, SS1, and M6 digested with MboI. Lanes 7-9 are DNA from 26695, SS1, and M6 cut with Sau3AI. Lanes 10-12 are DNA from 26695, SS1, and M6 sheared by vortexing. Lanes 13-15 are DNA from 26695, SS1, and M6 sheared by sonication. KpnI and MboI are ineffective in cutting the genome into a ladder of fragments. Vortexing produced bands that were larger than desired. Sau3AI and sonication produced bands of the appropriate size.

IVET procedure because the physical method will generate fragments in a more random manner (i.e., with less bias) than using a restriction enzyme.

Urease Test Agar

Urease broth can be used to determine the presence of urease activity by color change. Using this broth, urea hydrolysis is indicated by a color change from orange/red to pink/purple, indicating that the broth is more alkaline due to the liberation of ammonia. This broth was used to make agar plates so that large numbers of colonies could be tested quickly for urease activity. Plates were tested by applying colonies

101 directly to the plates and looking for the development of a pink halo. Alternatively, colonies were transferred onto filter paper and the paper was then placed onto the UTA plate.

The best color change occurred in plates with 1.2% urease broth. Both methods of applying colonies to the UTA plates resulted in a color change, but the presence of the filter paper made for easier detection of the color change because of the white background (Figure 4.5). This is useful in that filter paper can be used to easily and quickly transfer entire plates of colonies to the UTA plates for detection of urease activity. The color change was relatively quick, with 10 minutes being sufficient for an effective change in color. Color continues to develop over time, but the diffusion of the color within the agar after one hour may become too great in plates that have a large number and/or closely-spaced colonies. Therefore, plates should be read within an hour of colony transfer.

102

Figure 4.5: Urease positive colonies on a UTA plate. A filter membrane was used to transfer colonies from a blood agar plate to the UTA plate. Color change is evident within 10 minutes.

Construction of IVET plasmid

For the first IVET plasmid, p342 was to be modified by insertion of a green fluorescent protein gene (gfp) into the StuI site. The 0.82 kb cloned gfp segment contained the gfp gene, the 5’ MCS from pGFPmut3.1 to allow for insertion of random genome fragments into the plasmid, and the synthetic ribosome binding site upstream of the gfp gene. Ligations using the insert generated by the first set of primers, which contained StuI sites, were unsuccessful. No transformants carrying the insert resulted despite numerous attempts with this procedure. In an attempt to find a more efficient enzyme for digesting the insert, the gfp gene was re-cloned using a second set of

103 primers with a different restriction site, EcoRI. Following digestion and Klenow

reaction to fill in the overhanging ends, ligations and transformations were performed.

From these ligations, six transformants were obtained that putatively carried the insert fragment. All transformants were checked for presence of the insert via PCR and

all six were confirmed to contain the cloned gfp insert. Because the reaction was a blunt-end ligation and orientation of the insert in the plasmids was unknown, the six plasmids were then checked for direction of gfp insertion via PCR. Transformants

carrying the insert in the correct orientation would result in the presence of a band of the correct size. Two of the six transformants were found to have the insertion in the correct orientation (Figure 4.6). Thus, it was confirmed that two of the plasmids carried the cloned gfp insert in the correct orientation, and that p342GFPmut3.1 had been successfully constructed (Figure 4.7).

104

pGFPmut3.1 p342

1 23456

1650

1000

Figure 4.6: PCR results of orientation of the gfp insert into p342. Lanes 1 and 2 show transformants that carry the insert in the correct direction. Lanes 3-6 show transformants carrying the insert in the incorrect orientation.

105

rrnB terminators MCS gfp hpn

H. pylori ureB genomic DNA

CmR p342GFPmut3.1 ~12.2 kb colE1

H. pylori genomic R Ap DNA

Figure 4.7: Plasmid p342GFPmut3.1. This plasmid was derived from the insertion of the 5’MCS and gfp gene from plasmid pGFPmut3.1 into the StuI site of p342. The gfp gene can be used as a reporter gene in E. coli in pre-screening promoter inserts for correct orientation of insertion and activity. The Kpn site in the MCS would be used for inserting genome fragments into the plasmid.

To test that the p342GFPmut3.1 plasmid was functional, a known promoter was

cloned for insertion into the MCS upstream of the gfp reporter gene. The lac promoter

(Plac) was chosen for this because it is a strong and inducible promoter that functions in

E. coli, in which the pre-screening step for fluorescence would be conducted. Plac was cloned from the pGFPmut3.1 plasmid as a 200 bp fragment. To enable insertion into the MCS of p342GFPmut3.1 in the correct orientation, one primer was designed with a

PstI site while the other was designed with a KpnI site.

106 The resulting PCR product and p342GFPmut3.1 were digested with PstI and

KpnI and ligation reactions were performed, but no transformants were obtained from

the ligation reactions. Upon further testing, it was found that PstI digests p342GFPmut3.1 in more than one location and so new primers were designed so that

they both contained the KpnI restriction site. Following ligation and transformation,

colonies were obtained from several plates. The colonies were transferred onto a

LB+Ap+Cm+IPTG plate and incubated overnight at 37°C. However, no colonies

fluoresced when the plates were checked under UV light. Random colonies were

checked for the presence of the insert, but no insert was detected in any of the plasmids.

While attempting to test the p342GFPmut3.1 plasmid, a quick “shotgun”

approach was attempted to test the ability to ligate fragments into the original p342

plasmid. However, no transformants were obtained. A series of ligation controls were

done to check that p342 was a viable vector. The variations used in these controls were

uncut vector, vector cut with StuI and not dephosphorylated, and cut vector that was

dephosphorylated. All controls behaved as expected, with both the uncut vector and the

cut vector (non-dephosphorylated) able to transform E. coli DH5α. No transformants

were expected from the dephosphorylated vector, and none were obtained.

Problems with ligations into both p342 and p342GFPmut3.1 may have been due

to the difference in sizes between the insert and the vector. Not only would this reduce

the efficiency of the ligation reaction, but blunt-end ligations are also naturally

inefficient. As a result, a second tactic was devised to create the IVET plasmid in

which portions of the p342 plasmid were removed to decrease its size, and increase the

chance of achieving successful ligations into the plasmid. Few choices for enzymes

107 were available due to the large size of the plasmid. However, FspI was found to cut

only four times in the plasmid and the fragments did not cut within the portions of the

plasmid that would be necessary for use as the IVET plasmid.

Using sequence analysis programs, one of the FspI restriction sites was found to

be 1091 bases upstream of the hpn gene, in the genomic DNA portion of the plasmid,

one site was 1752 bases downstream of the CmR gene, one site was between the lacZ

gene and the f1 origin in the pBluescript backbone, and the last site was located in last

fourth of ApR gene in pBluescript backbone. Because the plasmid carries a CmR gene, it was determined that the ApR gene from the Bluescript backbone would be

unnecessary. Furthermore, the large amount of H. pylori genomic DNA in the plasmid was also not needed. The expected sites would leave sufficient genomic DNA for homologous recombination, the hpn, ureB, and cat genes, as well as the ColE1 origin on the Bluescript backbone.

p342 was digested with FspI and when the products were analyzed on an agarose gel bands appeared to be of the expected size. However, there was also the presence of an extra band of approximately 800 bp (Figure 4.8). The extra fragment and the two smaller expected fragments were discarded. These two smaller fragments contained a portion of the ApR gene, half of the MCS from the pBluescript backbone,

and 1.4 kb of the genomic DNA region downstream of the hpn gene. The two larger

fragments contained the genomic DNA region upstream of the hpn gene, the ColE1

R origin and half of the MCS from the pBluescript backbone, the hpn, ureB, and Cm

genes, and 1.75 kb of the genomic DNA region from downstream of the hpn gene.

Because of the unexpected presence of the extra fragment, the two larger fragments

108 3

1.6

extra band

Figure 4.8: p342 digested with FspI. Four fragments were expected based on analysis of the published sequence of the segments contained in plasmid p342. However, a fifth fragment (approximate size of 800 bp) was detected in the digest. The four remaining fragments appeared to be the appropriate sizes: 1200 bp, 1600 bp, 2200 bp, and 6100 bp.

were checked via PCR to ensure that the hpn gene and its immediate upstream region were still intact and that the extra fragment did not contain a critical piece of the plasmid. This hpn locus and upstream genomic DNA region is important so that the necessary segment of the plasmid can be integrated into the genome of the mutant strain via homologous recombination. Following confirmation of the necessary elements in the two larger fragments, they were then ligated together and transformed into E. coli

DH5α. A single transformant was obtained from the transformation and the plasmid was isolated. The plasmid was digested separately with StuI and FspI to check for

109 plasmid size as well as expected restriction patterns. StuI was expected to cut once and

FspI to cut twice. The plasmid cut as expected (Figure 4.9) and fragments were of the expected size.

123

6 2

Figure 4.9: Agarose gel of pHIVET digested with StuI or FspI. Lane 1 is pHIVET digested with FspI, and shows the expected bands of the appropriate sizes. Lane 2 is pHIVET digested with StuI, and shows a band of the expected size. Lane 3 shows undigested pHIVET.

The resulting plasmid was checked for presence of hpn and ureB in proper orientation by PCR. Bands of the expected sizes were obtained (Figure 4.10), so the genes are present in the plasmid in the correct orientation. The confirmed plasmid was called pHIVET (Figure 4.11).

110 12

2000

200

Figure 4.10: PCR confirmation of ureB and hpn in pHIVET. pHIVET was used as the template in PCR reactions using primers to detect the presence of the hpn and ureB genes. Lane 1 shows the appropriate band for hpn, and lane 2 shows the expected band for ureB.

StuI hpn

H. pylori genomic DNA ureB

pHIVET ~8.4 kb

ColE1

CmR

H. pylori genomic DNA

Figure 4.11: The pHIVET plasmid. This plasmid was constructed by removing two portions of the p342 plasmid by restriction digest with FspI. 111 The next step was to test pHIVET with a known promoter. The promoter

chosen for testing was the ureAB promoter (PureAB). This promoter was chosen because

it is a strong promoter in H. pylori and its activity is easy to measure. A region of

approximately 850 bp was cloned from H. pylori M6 using PCR. This fragment size was chosen for two reasons: 1) to ensure that any upstream elements necessary for promoter activity were present, and 2) to approximate the size of the random fragments

that would be cloned in from the sheared genomic DNA (ranging in size from about

500-1000 bp). Pfu DNA polymerase was used because it does not add additional bases

to the ends of the PCR products and a blunt-ended product was needed for insertion into

the StuI site of the pHIVET plasmid. Following the protocol that came with the Pfu

enzyme, little to no product was obtained. The protocol does not include the addition of

MgCl2 because it is already added to the buffer. However, titration of the PCR

reactions by the addition of various amounts of extra MgCl2 showed that the addition of

MgCl2 was needed to obtain product in these reactions.

pHIVET was digested with StuI and dephosphorylated with CIAP to prevent re-

circularization. The cloned PureAB was treated with T4 polynucleotide kinase to add a

phosphate group onto its ends to allow for effective insertion into the plasmid.

Following the kinase reaction, the ligation was set up to maximize the amount of insert relative to the amount of plasmid, and then transformed into E. coli DH5α. pUC19 was

provided with the competent cells and used as the transformation control. The

transformation efficiency calculated from the control was 8x107 transformants/µg

plasmid DNA. According to the protocol, efficiencies from ligations are expected to be

112 approximately 10-fold lower. However, no transformants were obtained from the

ligation reactions. Subsequent attempts at ligations and transformations were also

unsuccessful.

In an attempt to find an easier way to generate the promoter insert, the primers

used to clone the PureAB fragment were redesigned so that they contained an EcoRV

restriction site. This would allow for the PCR to be conducted using Taq DNA

polymerase and would permit restriction digestion to obtain blunt ends that would not

need to have a phosphate added on to the end in a subsequent reaction. To help assure

efficient digestion of the product, it was subcloned using the PCR-Script Cam Cloning

Kit (Stratagene). Insertion of the product into the pPCR-Script Cam SK(-) vector allowed for a step to monitor digestion of the insert fragment. Unlike digesting the PCR

product directly, in which there is no way to be certain that the fragment ends were

restricted, this digestion step could be checked by running the digestion on an agarose

gel and confirming that the fragment was efficiently cut from the vector. Following

ligation and transformation, aliquots of 1-150 µl of cells were plated onto LB-Cm-

IPTG-X-gal plates and incubated overnight at 37°C. After incubation, the plates looked

at expected, with a mixture of white and blue colonies. The plate with 1 µl of cells had

no colonies, and the plates with 100 µl and 150 µl of cells had lawns. The ratio of

white:blue colonies on the plates with 10 µl and 50 µl of cells was approximately 1:2.

Six white colonies were chosen randomly from these plates and grown overnight. The

plasmids were isolated and digested with EcoRV and run on an agarose gel. All six

plasmids showed the presence of two bands of the expected sizes.

113 The insert fragment was isolated from the gel and used in a ligation reaction

with pHIVET (digested with StuI and dephosphorylated). The ligation reaction was

transformed into E. coli DH5α cells and ten transformants resulted. These ten colonies

were then checked for the presence of the insert. These primers are located internal to

the PureAB fragment, and the presence of a band of the correct size should indicate presence of the insert. All ten plasmids showed a band of the correct size, while the pHIVET control plasmid had no band (Figure 4.12).

1 23 4 5678910 11

500

400

Figure 4.12: Confirmation by PCR of the presence of the insert fragment within pHIVET. Plasmids carrying the insert should have a band of approximately 430 bp. The DNA ladder shown is a 1 kb plus ladder. Plasmids used in lanes 1-10 all contain the insert. pHIVET was used as a control in lane 11.

To check for the presence of the fragment in the correct orientation, another

PCR was conducted using primers located so that a band would result only if the

114 fragment was present in the correct orientation. Using these primers, bands of the appropriate size were obtained with plasmids 1, 3, and 5-10 (Figure 4.13).

1 2345 6 78910

1000

Figure 4.13: Determination of insert orientation by PCR. The agarose gel shows plasmids that have the insert in the correct orientation by the presence of a band at 1000 bp. Lanes 1, 3, and 5-10 all have bands of the correct size.

Because several of the plasmids did appear to carry the insert according the above PCR confirmation reactions, the plasmids were used for transformation of H. pylori M6∆ureB. This is a mutant derived from the M6 strain, which is a mouse- adapted, easily transformed H. pylori strain. M6∆ureB has been used successfully in transformations in previous studies (47). In the M6∆ureB strain, the ureB open reading frame was replaced with a kanamycin resistance cassette (47), rendering the strain incapable of colonization unless a functional ureB is present to complement for the mutation.

115 Transformations of this strain by the putative pHIVET(PureAB) plasmids were conducted using transformation method one, in which broth cultures were used.

Multiple attempts using this method were unsuccessful. In order to circumvent possible restriction barriers, the plasmids were treated with cell-free extract (CFE) and transformations were attempted again using both the broth and plate methods of transformation, but no transformants were obtained.

In summary, methods for shearing the genomic DNA and testing bacteria following the mouse screen were successfully developed. Furthermore, two different

IVET plasmids were constructed. However, the introduction of a test promoter into p342GFPmut3.1 was not achieved. A test promoter was inserted into pHIVET, but the recombinant plasmid was not successfully transformed into H. pylori for testing.

DISCUSSION

Virulence genes and their functions will be completely understood only when we have full knowledge of the signals provided in the host. In the meantime, techniques that allow us to probe bacterial-host interactions in the absence of full knowledge about the relevant chemical and physical cues in vivo can be advantageous in pointing us in directions of further study. These methods aid us in identifying genes that are important under a given set of conditions without requiring that the exact circumstances be replicated accurately within a test tube. While these approaches do not necessarily elucidate the roles of the gene products, in some cases knowing when or where a particular factor required can lead to reasonable and testable hypotheses of gene function.

116 Previous studies in our lab have shown that random evaluation of eight promoters resulted in one that was up-regulated in vivo (47). Statistically, our data imply that at least 100 other promoters (out of approximately 1600 open reading frames in H. pylori) can be expected to be up-regulated in vivo. Additionally, our experience has shown that a minimum level of urease expression is required to permit colonization, and thus weak promoters will likely not be detected. However, this is acceptable as the more strongly expressed genes are likely to be those that are most important in virulence. Isolating these promoters will allow us to identify virulence factors that are upregulated in vivo, and thus represent suitable targets for novel treatments or vaccination.

H. pylori research is made more difficult by the fact that there remains a lack of reproducible systems for genetic manipulation with which to perform studies quickly and easily in the organism. This has impaired the evaluation of many aspects of H. pylori research, including that of factors that are up-regulated in vivo and thus could represent targets for therapy or vaccination. Our goal was to attempt to construct an

IVET system for screening the genome for genes that are up-regulated in vivo, and thus likely to be an important factor in colonization and/or pathogenesis. The effective use of this tool in H. pylori depends on the ability to transform strains efficiently with a plasmid that is heterologous to the organism.

Cutting the genomic DNA was found to be somewhat problematic when using restriction enzymes. The target size of fragment for insertion was 500-1000 bp. This size was chosen because it represents the average gene size in H. pylori (157). Some enzymes did not appear to cut well, and resulted in fragments that were larger than the

117 desired size. In addition, cutting by restriction digestion would bias the distribution of

fragments because cuts would result in specific locations each time the genome was cut.

Physically shearing the DNA by sonication consistently resulted in fragments of the

appropriate size and would result in the generation of truly random fragments.

However, the ends of the DNA generated by this method would then need to be polished to ensure that the ends were blunt and could be ligated into the recipient

plasmid. This could be achieved by either filling in the ends or removing any

overhanging ends, although detecting the efficiency of either of these polishing

reactions would be difficult.

The first plasmid, p342GFPmut3.1, was constructed to be used as an indirect

method of screening the genome. The gfp reporter gene would allow the use of this

plasmid in E. coli to screen for promoters that were inserted in the correct orientation

prior to use in H. pylori. The H. pylori hpn gene allows for homologous recombination

of the plasmid into the chromosome. Hpn is a protein that has a high affinity for

binding nickel, and although Hpn deletion mutants grow normally in vitro and are able

to colonize, the exact role of this gene in the cell is unknown (60). Thus, the resulting

plasmid, p342GFPmut3.1 could be transformed into E. coli and colonies that fluoresce

under UV light (i.e. those that contain a promoter in the proper orientation) would be isolated. The plasmids from fluorescing colonies would be isolated and transformed

into a ureB-negative H. pylori mutant strain. Transformed bacteria would be pooled

and inoculated into mice. Bacteria carrying a promoter that is up-regulated in vivo will

produce active urease and be able to survive the mouse challenge. Following sacrifice,

surviving bacteria will be collected from the gastric mucosa and will be screened again

118 for UV fluorescence. At this step, the colonies of interest will be the non-fluorescent

colonies as this will signify that the promoter is off in vitro. This in vitro screening step

will decrease the chance that isolated genes are those that are constitutively expressed,

such as housekeeping genes.

The potential weakness of this method is that promoters that are not active in E.

coli will not fluoresce and thus some H. pylori promoters of importance may be missed.

Furthermore, because the gfp gene was inserted into the StuI site of p342, the promoter

inserts would have to be inserted into the KpnI site of the MCS. This site was chosen

because there is a KpnI site in the MCS and not elsewhere in the plasmid. This would

necessitate using KpnI to shear the genome, which would introduce bias into sampling

the genome.

Plasmid p342GFPmut3.1 was successfully constructed. However, testing the

plasmid with a known promoter was problematic. The lac promoter was initially

chosen because it is a strong and inducible promoter and should be active in E. coli.

However, transformants carrying the cloned lac promoter were never obtained.

Ligation reactions were attempted with varying amounts of both insert and plasmid

DNA, but changing the ratios of insert:plasmid DNA did not affect the outcome. The

restriction enzymes used were evaluated separately and KpnI was determined to be

efficient and cut the plasmid at only one location.

However, despite the apparent reproducibility of each aspect of the

transformation protocol, transformants carrying the Plac insert in the correct orientation were not obtained. In the one instance where colonies were obtained, none fluoresced under UV light. Thus, the inserts were either all in the incorrect orientation, the

119 plasmids did not contain any inserts (self-ligation of the plasmid), or the GFP system

was not functional. When the plasmids were checked for the presence of the insert,

none was found, indicating that the problem was either self-ligation or a problem with

the gfp gene.

Because of the difficulty in obtaining transformants and the potential

weaknesses of the gfp reporter (see above), this plasmid system was abandoned in favor

of creating an IVET plasmid that would directly screen the H. pylori genome. For this

IVET plasmid, the ureB gene would be used as both the in vivo and in vitro reporter.

Urease is a good choice for a reporter gene in H. pylori because, although it is not essential for growth in vitro, it is essential for colonization. The original p342 plasmid contained a promoterless ureB gene followed by a chloramphenicol resistance gene and preceded by a StuI restriction site. This entire construct is flanked by gene sequences homologous to the H. pylori hpn gene. The plasmid contains an E. coli origin of

replication and is unable to replicate in H. pylori. The StuI cloning site in p342 is

located approximately 30 bp upstream of the ureB gene and this short region contains a

ribosome binding site. Thus, when H. pylori M6∆ureB is transformed with this

plasmid, chloramphenicol resistant colonies would be generated following a double

crossover event leading to insertion of the ureB:cat construct just downstream of the

hpn gene. UreB would be expressed only when an active promoter is inserted into the

StuI site. It has been shown in previous studies in our laboratory that non-ureB

promoters can effectively initiate transcription of ureB so that colonization takes place

(47), and so this construct is a likely candidate as an IVET system. In order to increase

the efficiency of ligation and transformation, p342 was modified to decrease its size.

120 Approximately 2.5 kb of p342, including part of the pBluescript backbone and part of

the H. pylori genomic DNA flanking the hpn region, was deleted. However, the partial

multiple cloning site and ColE1 region from the pBluescript backbone, the hpn, ureB

and CmR genes, and sufficient genomic DNA for homologous recombination remained

intact in the new plasmid, pHIVET.

Construction of pHIVET and cloning PureBA into the plasmid for testing was

successful. However, obtaining H. pylori transformants with the plasmid was not achieved. Because the H. pylori M6∆ureB strain is known to be transformable, it was

determined that the plasmid must be causing the difficulties. Although the plasmid size was somewhat large, previous studies on cag genes fused to the ureB gene used the larger p342 plasmid from which pHIVET was derived and so the size of pHIVET should not be an issue in lack of transformation. The pHIVET(PureAB) plasmids are

stably maintained in E. coli DH5α, and so it is presumed that the backbone structure

with the ColE1 region is sound and that the plasmid can self-replicate. Finally, the E.

coli cells carrying the plasmids grow on LB+Cm and so the CmR gene also appears to

be functional in E. coli. The ureB gene is not necessary for in vitro growth of H. pylori

and thus should not cause problems with the transformation. Because the plasmid is a

suicide vector in H. pylori, it is critical that the plasmid integrate by homologous

recombination into the genome. The hpn region of pHIVET was checked by PCR,

however, and appeared to be intact. Thus, loss of the recombination site does not

explain the lack of transformation.

The most likely explanation for failure of pHIVET to transform H. pylori is the

presence of restriction barriers. H. pylori strains are known to be naturally competent

121 for chromosomal DNA and, to a lesser extent, plasmid DNA. The strain used here, M6,

is highly naturally competent, and has been used in a number of different studies

requiring multiple transformation steps (47). However, the repeated passage of

pHIVET in E. coli could have resulted in methylation with resulting loss in

transformation efficiency in H. pylori.

Bacteria have evolved defenses to protect their own DNA. This includes

restriction endonucleases to recognize and cleave foreign DNA paired with methylases

that are used to differentiate self-DNA from foreign DNA (9). These restriction-

modification (R-M) systems have been classified as types I, II, or III, according to

several factors (167). Studies have shown that H. pylori DNA is highly methylated at

both adenine (A) and cytosine (C) residues (131), and analysis of the genome sequence

predict 14 or 15 potential type II R-M systems for H. pylori strains 26695 and J99,

respectively (7, 157). This large number of R-M systems may impact the ability to

transform this organism with plasmids derived from other bacterial species.

Many H. pylori strains are known to be naturally competent for transformation

in vitro by chromosomal DNA (125). Further studies into aspects of transformation in

H. pylori revealed several characteristics (78). H. pylori strains could be transformed at

a high frequency by chromosomal DNA from both related and unrelated H. pylori

strains. That is, H. pylori has no strong barriers to transformation by chromosomal

DNA. A high degree of transformation occurred with concentrations of DNA as low as

8 8 0.6 ng DNA/10 cells, and saturation occurred at approximately 6 ng DNA/10 cells.

The maximum frequency of transformation was about 1.3x104 transformants/µg

DNA/cfu suggesting that, even when DNA is in excess, only a fraction of H. pylori cells

122 in the culture are transformed. Competition experiments using other H. pylori DNA, H.

bilis DNA, and E. coli DNA showed that only E. coli DNA did not cause a change in

the transformation frequency. This indicates that H. pylori likely has a mechanism for

differentiating Helicobacter DNA from the DNA of other species.

Recent studies have found strong barriers to transformation of H. pylori cells by

plasmids from heterologous H. pylori strains (8). Chromosomal DNA transformed all

strains well, indicating that the barriers were not due to a lack of competence in the host

strains. Electroporation had no detectable effect on heterologous transformations even

though it increased the efficiency of homologous transformations. This indicates that

the strong barriers to transformation by heterologous plasmid DNA is not easily

overcome. In various H. pylori strains analyzed, it was found that each had its own

varied complement of functioning R-M systems, suggesting that these differences

among strains affect transformation by plasmids (8). More detailed studies found that

an MboI R-M system is present in some H. pylori strains (8). Most E. coli strains contain three site-specific DNA methylases, Dam, Dcm, and EcoKI. MboI methylases and Dam both methylate adenines in the recognition sequence GATC. Plasmid DNA isolated from dam+ E. coli strains (such as DH5α) should be completely resistant to

GATC cleavage by MboI. Indeed, Ando et al. found that MboI.R+ H. pylori were transformed by a plasmid isolated from DH5α but were unable to be transformed by a

plasmid isolated from a dam- strain, SCS110. MboI.R- strains, on the other hand, were transformable by both plasmids (8). These data help illustrate the fact that barriers to

transformation of H. pylori need to be taken into consideration, especially in light of the fact that H. pylori 26695 and J99 each contain at least nine complete putative type I, II,

123 or III R-M systems (7, 157). If these are functional, they may be a strong barrier to

efficient transformation of H. pylori by plasmid DNA from other bacteria.

The large number of potential R-M systems in known strains, and the degree of

genetic diversity found in the genus make it impractical to isolate restriction-

deficient/modification proficient H. pylori strains. Instead, one group attempted to find

a method to protect plasmid DNA from restriction by specific methylation in vitro before natural transformation of H. pylori. Specifically, they describe a method for strain-specific methylation of plasmid DNA isolated from E. coli that significantly increased the efficiency of transformation of several H. pylori strains (40). This technique uses cell-free extract generated from the host strain to treat plasmid DNA from the donor strain. Experiments showed that incubation of plasmid DNA isolated from E. coli DH5α with CFE from an H. pylori strain resulted in acquisition of the restriction pattern of genomic DNA from the source strain. This technique allowed transformation of H. pylori strains that were not previously transformable by several plasmids that had been isolated from E. coli. Thus, it was shown that H. pylori-strain- specific restriction barriers can be overcome by methylation of plasmid DNA in vitro with a CFE from the homologous strain before introduction into H. pylori (40).

Because the pHIVET plasmid in our study was being propagated in E. coli prior to transformation, it is possible that the plasmid had been modified by E. coli methylating enzymes. This may result in being recognized as foreign by H. pylori and

being degraded by restriction endonucleases. The method described by Donahue et al.

(40) for overcoming the restriction barrier in transforming H. pylori with heterologous

plasmids was attempted in the hopes that this would allow for transformation. The cell-

124 free extract (CFE) was prepared from M6 and used to treat two of the pHIVET(PureAB) plasmids. Treating the plasmid with CFE should protect it from restriction by modifying the plasmid so that it acquires the restriction pattern characteristic of genomic DNA from the source strain (40). Transformations were attempted by both broth and plate methods, but neither resulted in transformants. A potential problem in the technique was that the amount of protein in the CFE was somewhat low, and this necessitated using a large volume to obtain the amount of protein specified in the published method for treating the plasmids with the CFE. Further attempts with the method would modify the protocol so that the amount of protein obtained was more concentrated.

Furthermore, there are two additional issues that could have affected the results using CFE to treat plasmid DNA prior to transformation. First, there are potential strain-specific differences in expression levels of methyltransferases versus their related restriction enzymes (40). To obtain maximum protection of the plasmid DNA, methyltransferase activity in the CFE must be sufficient to methylate most target sites.

If a specific methyltransferase is expressed at low levels at the time of extract preparation, but the expression of the related restriction activity is high in the recipient cells at the time of transformation, the extent of site-specific methylation achieved in vitro may not be sufficient for protection and thus could affect transformation efficiency. Secondly, the conditions established for extract preparation and plasmid treatment by Donahue et al. may not be optimal for the enzymatic activity of all

125 methyltransferases required for protection in a specific recipient strain. Either of these situations could have potentially affected the efficacy of using M6 CFE-treated pHIVET plasmid DNA in our transformations.

126

CHAPTER 5

PERSPECTIVES AND FUTURE DIRECTIONS

The studies conducted for this dissertation approached aspects of colonization

from two separate viewpoints. The first, covered in chapters two and three, was based on the possibility that a metabolic pathway that has been maintained by H. pylori may affect its ability to colonize its host. Although we found that the Entner-Doudoroff pathway is active in H. pylori and does affect fitness of the organism to some small

degree, it was not enough to fully disrupt colonization ability. The presence of this

pathway in the organism is likely important in some way to the bacterium’s survival. It

is possible that, instead of being critical in vivo, it is utilized in some way during

transmission of the organism from host-to-host, as the transmission route for this

pathogen remains unidentified. Because of the unique niche that H. pylori occupies, the

bacterium may have developed novel or unusual genes for metabolic pathways.

Furthermore, pathways that are still being debated, such as the EMP pathway, must be studied further. When it is fully understood how these pathways work together, they may enable the development and use of novel therapies, perhaps by somehow affecting

the organism’s ability to obtain enough nutrients to effectively colonize the host.

127 In the fourth chapter of this dissertation, we describe the creation of an IVET

plasmid for promoter-trapping and the restriction barriers that hinder transformation of

H. pylori with this plasmid. As described above, it is likely that passage of pHIVET in

E. coli markedly diminished efficiency of transformation. Thus, treatment with the

plasmid with concentrated CFE, screens of larger pools of putative transformants, and

possibly transformation of alternative H. pylori strains will be necessary.

Once pHIVET has been satisfactorily tested, sheared genomic DNA can then be

ligated into the plasmid to generate a promoter library that would be used to transform

H. pylori M6∆ureB. Methods for shearing DNA have been established (above) and

methods for mouse inoculation and bacterial recovery are well established (47, 50). In the current study, we developed a method for rapidly screening mouse-recovered isolates using the UTA plates that will allow efficient identification of urease-positive and –negative isolates.

The clones identified as urease-positive in vivo (by surviving the mouse challenge) and urease-negative in vitro (by UTA plate analysis) would contain promoters that are up-regulated in vivo, and would thereby be the clones of interest.

The promoter region carried by the plasmid would be isolated by PCR with primers flanking the cloning site and then sequenced. The resulting sequence information would be checked against gene databases to discover the identity of the genes associated with those promoters. When genes have been identified, they will be compared for similarities or links that may identify a specific gene family essential for virulence. Any genes that are identified via this method would then need to be further

128 tested individually by more conventional methods to determine whether the identified genes have any effect on colonization or pathogenesis, or whether they may be an appropriate target for therapy or vaccination.

Although the interest in research on basic metabolism appears to wax and wane over the years, perhaps the major impact of methods developed for studying bacterial pathogenesis in vivo will emphasize the overall importance of these types of studies. Of numerous genes expressed in vivo by various organisms and detected by IVET and other in vivo systems, the most common are those involved with the acquisition of metals, the synthesis and acquisition of nucleotides and cofactors, DNA repair, membrane modification, protein targeting, thermotolerance, osmotic tolerance and acid tolerance. From these and other studies, it appears that bacteria use many different metabolic pathways to grow in vivo, and that their metabolic characteristics influence tissue tropism and host specificity. Thus, the importance of metabolism in colonization and pathogenesis should not be disregarded. Beyond the interest of basic science lies a greater goal: the development of a suitable vaccine against H. pylori. Because of the worldwide scope of H. pylori infection, the significance of any studies that can lead to resolution of this goal should not be understated or overlooked.

129

BIBLIOGRAPHY

1. Akopyants, N., N. O. Bukanov, T. U. Westblom, S. Kresovich, and D. E. Berg. 1992. DNA diversity among clinical isolates of Helicobacter pylori deteted by PCR-based RAPD fingerprinting. Nucleic Acids Res. 20:5137-5142.

2. Akopyants, N. S., S. W. Clifton, D. Kersulyte, J. E. Crabtree, B. E. Youree, C. A. Reece, N. O. Bukanov, E. S. Drazek, B. A. Roe, and D. E. Berg. 1998. Analyses of the cag pathogenicity island of Helicobacter pylori. Mol. Microbiol. 28:37-53.

3. Akopyants, N. S., K. A. Eaton, and D. E. Berg. 1995. Adaptive mutation and cocolonization during Helicobacter pylori infection of gnotobiotic piglets. Infection and Immunity 63:116-121.

4. Akopyants, N. S., A. Fradkov, L. Diatchenko, J. E. Hill, P. D. Siebert, S. A. Lukyanov, E. D. Sverdlov, and D. E. Berg. 1998. PCR-based subtractive hybridization and differences in gene content among strain of Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 95:13108-13113.

5. Allen, A. 1984. The structure and function of gastrointestinal mucous, p. 3-11. In E. C. Boedeker (ed.), Attachment of organisms to the gut mucosa, vol. II. CRC Press, Boca Raton.

6. Allen, A., J. Newton, L. Oliver, N. Jordan, V. Strugala, J. P. Pearson, and P. W. Dettmar. 1997. Mucus and Helicobacter pylori. Journal of Physiology and Pharmacology 48:297-305.

7. Alm, R., L.-S. L. Ling, D. T. Moir, B. L. King, E. D. Brown, P. C. Doig, D. R. Smith, B. Noonan, B. C. Guild, B. L. deJonge, G. Carmel, P. J. Tummino, A. Caruso, M. Uria-Nickelsen, D. M. Mills, C. Ives, R. Gibson, D. Merberg, S. D. Mills, Q. Jiang, D. E. Taylor, G. F. Vovis, and T. J. Trust. 1999. Genomic-sequence comparison of two unrelated isolates of the human gastric pathogen Helicobacter pylori. Nature 397:176-180.

130 8. Ando, T., Q. Xu, M. Torres, K. Kusugami, D. A. Israel, and M. J. Blaser. 2000. Restriction-modification system differences in Helicobacter pylori are a barrier to interstrain plasmid transfer. Mol. Microbiol. 37:1052- 1065.

9. Arber, W. 1974. DNA modification and restriction. Prog. Nucleic Acid Res. Mol. Biol. 14:1-37.

10. Atherton, J. C., P. Cao, J. R. M. Peek, M. K. R. Tummuru, M. J. Blaser, and T. L. Cover. 1995. Mosaicism in vacuolating cytotoxin alleles of Helicobacter pylori. Association of specific vacA types with cytotoxin production and peptic ulceration. J. Biol. Chem. 270:17771-17777.

11. Atherton, J. C., R. M. Peek, K. T. Tham, T. L. Cover, and M. J. Blaser. 1997. The clinical and pathological importance of heterogeneity in vacA, encoding the vacuolating cytotoxin of Helicobacter pylori. Gastroenterology 112:92-99.

12. Attridge, S. R., and D. Rowley. 1983. The role of the flagellum in the adherence of Vibrio cholerae. Journal of Infectious Disease 147:864-72.

13. Beier, D., G. Spohn, R. Rappuoli, and V. Scarlato. 1997. Identification and characterization of an operon of Helicobacter pylori that is involoved in motility and stress adaptation. Journal of Bacteriology 179:4676- 4683.

14. Beil, W., C. Birkholz, S. Wagner, and K.-F. Sewing. 1994. Interation of Helicobacter pylori and its fatty acids with parietal cells and gastric H+/K+- ATPase. Gut 35:1176-1180.

15. Berg, D. E., P. S. Hoffman, B. J. Appelmelk, and J. G. Kusters. 1997. The Helicobacter pylori genome sequence: Genetic factors for long life in teh gastric mucosa. Trends in Microbiology.

16. Blaser, M. J. 1992. Perspectives on the pathogenesis of Helicobacter pylori infections, p. 276-280. In B. J. Rathbone and R. V. Heatley (ed.), Helicobacter pylori and Gastroduodenal Disease, 2nd ed. Blackwell Scientific Publishing, Oxford.

17. Boren, T., S. Normark, and P. Faulk. 1994. Helicobacter pylori: Molecular basis for host recognition and bacterial adherence. Trends in Microbiology 2:221-8.

18. Camilli, A., D. T. Beattie, and J. J. Mekalanos. 1994. Use of genetic recombinations as a reporter of gene expression. Proc. Natl. Acad. Sci. U.S.A. 91:2634-2638. 131 19. Camilli, A., and J. J. Mekalanos. 1995. Use of recombinase gene fusions to identify Vibrio cholerae genes undiced during infection. Mol. Microbiol. 18:671-683.

20. Camilli, A., D. S. Merrell, and J. J. Mekalanos. 2001. Strategies to identify bacterial pathogenicity factors, p. 133-177. In E. A. Groisman (ed.), Principles of Bacterial Pathogenesis. Academic Press, San Diego.

21. Carsiotis, M., D. L. Weinstein, H. Karch, I. A. Holder, and A. D. O'Brien. 1984. Flagella of Salmonella typhimurium are a virulence factor in infected c56bl/6j mice. Infection and Immunity 46:814-8.

22. Cave, D. R., and M. Vargas. 1989. Effect of a Campylobacter pylori protein on acid secretion by parietal cells. Lancet 2:187-9.

23. Censini, S., C. Lange, Z. Xiang, J. E. Crabtree, P. Ghiara, M. Borodovsky, R. Rappuoli, and A. Covacci. 1996. cag, a pathogenicity island of Helicobacter pylori, encodes type I-specific and disease-associated virulence factors. Proc. Natl. Acad. Sci. U.S.A. 93:14648-14653.

24. Chalk, P. A., A. D. Roberts, and W. M. Blows. 1994. Metabolism of pyruvate and glucose by intact cells of Helicobacter pylori studied by 13C NMR spectroscopy. Microbiology 140:2085-2092.

25. Chalk, P. A., A. D. Roberts, A. A. Davison, D. J. Kelly, and P. J. White. 1997. Use of NMR to study H. pylori metabolism. In C. L. Clayton and H. L. T. Mobley (ed.), Helicobacter pylori protocols. Humana Press, Totowa, N.J.

26. Chen, M., and A. Lee. 1993. Vaccination possibilities and probabilities, p. 158-169. In T. C. Northfield, M. Mendall, and P. M. Goggin (ed.), Helicobacter pylori infection. Kluwer Academic Publishers, Boston.

27. Chiang, S. L., J. J. Mekalanos, and D. W. Holden. 1999. In vivo genetic analysis of bacterial virulence. Annu. Rev. Microbiol. 53:129-154.

28. Chiba, N., B. V. Rao, J. W. Rademaker, and R. H. Hunt. 1992. Meta- analysis of the efficacy of antibiotic therapy in eradicating Helicobacter pylori. Am. J. Gastroenterol. 87:1716-1727.

29. Clayton, C. L., M. J. Pallen, H. Kleanthous, B. W. Wren, and S. Tabaqchali. 1990. Nucleotide sequence of two genes from Helicobacter pylori encoding for urease subunits. Nucl. Acids Res. 18:362.

132 30. Clyne, M., A. Labigne, and B. Drumm. 1995. Helicobacter pylori requires an acidic environment to survive in the presence of urea. Infect. Immun. 63:1359-1372.

31. Conway, T. 1992. The Entner-Doudoroff pathway: history, physiology and molecular biology. FEMS Microbiology Reviews 103:1-28.

32. Cordwell, S. J. 1999. Microbial genomes and "missing" enzymes: redefining biochemical pathways. Arch. Microbiol. 172:269-279.

33. Covacci, A., S. Censini, M. Bugnoli, R. Petracca, D. Burroni, G. Macchia, A. Massone, E. Papini, Z. Xiang, N. Figura, and e. al. 1993. Molecular characterization of the 128-kDa immunodominant antigen of Helicobacter pylori associated with cytotoxicity and duodenal ulcer. Proc. Natl. Acad. Sci. U.S.A. 90:5791-5795.

34. Cover, T. L., D. E. Berg, M. J. Blaser, and H. L. T. Mobley. 2001. H. pylori pathogenesis, p. 509-558. In E. A. Groisman (ed.), Principles of Bacterial Pathogenesis. Academic Press, San Diego.

35. Cover, T. L., C. P. Dooley, and M. J. Blaser. 1990. Characterization of and human serologic response to proteins in Helicobacter pylori broth culture supernatants with vacuolizing cytotoxin activity. Infect. Immun. 58:603-610.

36. Crabtree, J. E., J. D. Taylor, J. I. Wyatt, R. V. Heatley, T. M. Shallcross, L. S. Tompkins, and B. J. Rathbone. 1991. Mucosal IgA recognition of Helicobacter pylori 120 kDa protein, peptic ulceration, and gastric pathology. Lancet 338:332-335.

37. Cussac, V., R. L. Ferrero, and A. Labigne. 1992. Expression of Helicobacter pylori urease genes in Escherichia coli grown under nitrogen limiting conditions. J. Bacteriol. 174:2466-2473.

38. Debets-Ossenkopp, Y. J., A. J. Herscheid, R. G. J. Pot, E. J. Kuipers, J. G. Kusters, and C. M. J. E. Vandenbroucke-Grauls. 1999. Prevalence of Helicobacter pylori resistance to metronidazole, chlarithromycin, amoxycillin, tetracycline and trovafloxin in The Netherlands. J. Antimicrob. Chemother. 43:511-515.

39. Dixon, M. F., R. M. Genta, J. H. Yardley, and P. Correa. 1996. Classification and grading of gastritis - the updated Sydney system. Am. J. Surg. Pathol. 20:1161-1181.

40. Donahue, J. P., D. A. Israel, J. R. M. Peek, M. J. Blaser, and G. G. Miller. 2000. Overcoming the restriction barrier to plasmid transformation of Helicobacter pylori. Mol. Microbiol. 37:1066-1074. 133 41. Dooley, C. P., P. L. Fitzgibbons, H. Cohen, M. D. Appleman, G. I. Perez-Perez, and M. J. Blaser. 1989. Prevalence of Helicobacter pylori infection and histologic gastritis in asymptomatic persons. N. Engl. J. Med. 321:1562-1566.

42. Drumm, B., G. I. Perez-Perez, M. J. Blaser, and P. M. Sherman. 1990. Intrafamilial clustering of Helicobacter pylori infection. New Engl. J. Med. 322:359-363.

43. Dunn, B. E., G. P. Campbell, G. I. Perez-Perez, and M. J. Blaser. 1990. Purification and characterization of urease from Helicobacter pylori. J. Biol. Chem. 265:9464-9469.

44. Dunn, B. E., N. B. Vakil, B. G. Schneider, J. B. Zitzer, T. Puetz, and S. H. Phadnis. 1997. Localization of Helicobacter pylori urease and heat shock protein homolog in human gastric biopsies. Infect. Immun, 65:1181-1188.

45. Eaton, K. A. 1999. Animal models of Helicobacter gastritis. Current Topics in Microbiology and Immunology 241:123-153.

46. Eaton, K. A., C. L. Brooks, D. R. Morgan, and S. Krakowka. 1991. Essential role of urease in pathogenesis of gastritis induced by Helicobacter pylori in gnotobiotic piglets. Infection and Immunity 59:2470-5.

47. Eaton, K. A., J. V. Gilbert, E. A. Joyce, A. E. Wanken, T. Thevenot, P. Baker, A. Plaut, and A. Wright. 2002. In vivo complementation of ureB restores the ability of Helicobacter pylori to colonize. Infect. Immun. 70:771- 778.

48. Eaton, K. A., and S. Krakowka. 1994. Effect of gastric pH on urease- dependent colonization of gntobiotic piglets of Helicobacter pylori. Infection and Immunity 62:3604-3607.

49. Eaton, K. A., D. R. Morgan, and S. Krakowka. 1992. Motility as a factor in the colonisation of gnotobiotic piglets by Helicobacter pylori. Journal of Medical Microbiology 37:123-7.

50. Eaton, K. A., S. R. Ringler, and S. J. Danon. 1999. Murine splenocytes induce severe gastritis and delayed-type hypersensitivity and suppress bacterial colonization in Helicobacter pylori-infected SCID mice. Infect. Immun. 67:4594-4602.

51. Eaton, K. A., S. Suerbaum, C. Josenhans, and S. Krakowka. 1996. Colonization of gnotobiotic piglets by Helicobacter pylori deficient in two flagellin genes. Infection and Immunity 64:2445-2448.

134 52. Eisenberg, R. C., and W. J. Dobrogosz. 1967. Gluconate metabolism in Escherichia coli. Journal of Bacteriology 93:941-9.

53. Ferrero, R. L., S. L. Hazell, and A. Lee. 1988. The urease enzymes of Campylobacter pylori and a related bacterium. Journal of Medical Microbiology 27:33-40.

54. Forman, D., and P. Webb. 1993. Geographic distribution and association with gastric cancer, p. 11-20. In T. C. Northfield, M. Mendall, and P. M. Goggin (ed.), Helicobacter pylori infection. Kluwer Academic Publishers, Boston.

55. Fraenkel, D. G., and B. L. Horecker. 1964. Pathways of D-glucose metabolism in Salmonella typhimurium. The Journal of Biological Chemistry 239:2765-71.

56. Freter, R. 1988. Mechanisms of bacterial colonization of the mucosal surfaces of the gut, p. 45-60. In J. A. Roth (ed.), Virulence mechanisms of bacterial pathogens. American Society for Microbiology, Washington, D. C.

57. Fujimoto, S., B. Marshall, and M. J. Blaser. 1994. PCR-based restriction fragment length polymorphism typing of Helicobacter pylori. J. Clin. Microbiol. 34:331-334.

58. Garner, J. A., and T. L. Cover. 1995. Analysis of genetic diversity in cytotoxin-producing and non-cytotoxin-producing Helicobacter pylori strains. J. Infect. Dis. 172:290-293.

59. Geis, G., S. Suerbaum, B. Forsthoff, H. Leying, and W. Opferkuch. 1993. Ultrastructure and biochemical studies of Helicobacter pylori flagellar sheath. J. Med. Microbiol. 38:371-377.

60. Gilbert, J. V., J. Ramakrishna, J. F. W. Sunderman, A. Wright, and A. G. Plaut. 1995. Protein Hpn: Cloning and Characterization of a Histidine- Rich Metal-Binding Polypeptide in Helicobacter pylori and Helicobacter mustelae. Infec. Immun. 63:2682-2688.

61. Go, M. F., V. Kapur, D. Y. Graham, and J. M. Musser. 1996. Population genetic analysis of Helicobacter pylori by multilocus enzyme electrophoresis: Extensive allelic diversity and recombinational population structure. J. Bacteriol. 178:3934-3938.

62. Goggin, P., and T. Northfield. 1993. Mucosal defence, p. 40-. In T. C. Northfield, M. Mendall, and P. M. Goggin (ed.), Helicobacter pylori infection. Kluwer Academic Publishers, Boston.

135 63. Goodwin, C. S. 1994. How Helicobacter pylori acquired its name, and how it overcomes gastric defence mechanisms. Journal of Gastroenterology and Hepatology 9:S1-3.

64. Goodwin, S. 1993. Historical and microbiological perspectives, p. 1-10. In T. C. Northfield, M. Mendall, and P. M. Goggin (ed.), Helicobacter pylori infection: Pathophysiology, epidemiology, and management. Kluwer Academic Publishers, Boston.

65. Goossens, H., Y. Glupczynski, A. Burette, J.-P. Lambert, L. Vlaes, and J.-P. Butzler. 1992. Role of the vacuolating toxin from Helicobacter pylori in the pathogenesis of duodenal and gastric cancer. Med. Microbiol. Lett. 1:153- 159.

66. Graham, D. Y. 1998. Antibiotic resistance in Helicobacter pylori: implications for therapy. Gastroenterology 115:1272-1277.

67. Hazell, S. L. 1990. Urease and catalase as virulence factors of Helicobacter pylori, p. 3-14. In H. M. e. al. (ed.), Helicobacter pylori. Springer- Verlag, Berlin.

68. Hazell, S. L., A. Lee, L. Brady, and W. Hennessy. 1986. Campylobacter pyloridis and gastritis: Association with intracellular spaces and adaptation to an environment of mucus as important factors in colonization of the gastric epithelium. The Journal of Infectious Diseases 153:658-663.

69. Hazell, S. L., and G. L. Mendz. 1993. The metabolism and enzymes of Helicobacter pylori: function and potential virulence effects, p. 115-141. In C. S. Goodwin and B. W. Worsley (ed.), Helicobacter pylori: Biology and Clinical Practice. CRC Press, London.

70. Heithoff, D. M., C. P. Conner, P. C. Hanna, S. M. Julio, U. Hentschel, and M. J. Mahan. 1997. Bacterial infection as assessed by in vivo gene expression. Proc. Natl. Acad. Sci. U.S.A. 94:934-939.

71. Hensel, M., J. E. Shea, C. Gleeson, M. D. Jones, E. Dalton, and D. W. Holden. 1995. Simultaneous identification of bacterial virulence genes by negative selection. Science 269:400-403.

72. Hoffman, P. S., A. Goodwin, J. Johnsen, K. Magee, and S. J. O. Veldhuyzen Van Zanten. 1996. Metabolic activities of metronidazole-sensitive and -resistant strains of Helicobacter pylori: repression of pyruvate oxidoreductase and expression of isocitrate lyase activity correlate with resistance. Journal of Bacteriology 178:4822-4829.

136 73. Hu, L.-T., P. A. Foxall, R. Russell, and H. L. T. Mobley. 1992. Purification of recombinant Helicobacter pylori urease apoenzyme encoded by ureA and ureB. Infect. Immun. 60:2657-2666.

74. Hu, L. T., and H. L. Mobley. 1993. Expression of catalytically active recombinant Helicobacter pylori urease at wild-type levels in E. coli. Infect. Immun. 61:2563-2569.

75. Hu, L.-T., and H. L. T. Mobley. 1990. Purification and N-terminal analysis of urease from Helicobacter pylori. Infect. Immun. 58:992-8.

76. Huynen, M., T. Dandelkar, and P. Bork. 1998. Differential genome analysis applied to the species-specific features of Helicobacter pylori. FEBS Letters 426:1-5.

77. Huynen, M. A., T. Dandekar, and P. Bork. 1999. Variation and evolution of the citric-acid cycle: a genomic perspective. Trends Microbiol. 7:281-291.

78. Israel, D. A., A. S. Lou, and M. J. Blaser. 2000. Characteristics of Helicobacter pylori natural transformation. FEMS Microbiol. Letts. 188:275- 280.

79. Jones, D. M., and A. Curry. 1989. The ultrastructure of Campylobacter pylori, p. 48-62. In J. Rathbone and R. V. Heatley (ed.), Campylobacter pylori and Gastroduodenal Disease. Blackwell Scientific, Boston.

80. Jones, D. M., A. Curry, and A. J. Fox. 1985. An ultrastructural study of the gastric Campylobacter-like organism 'Campylobacter pyloridis'. Journal of General Microbiology 131:2335-41.

81. Jones, D. M., A. M. Lessells, and J. Eldridge. 1984. Campylobacter like organisms on the gastric mucosa: culture, histological, and serological studies. Journal of Clinical Pathology 37:1002-6.

82. Jorgensen, M., G. Daskalopoulos, V. Warburton, H. M. Mitchell, and S. L. Hazell. 1996. Multiple strain colonization and metronidazole resistance in Helicobacter pylori-infected patients: Identification from sequential and multiple biopsy specimens. J. Infect. Dis. 174:631-635.

83. Josenhaus, C., A. Labigne, and S. Suerbaum. 1995. Comparative ultrastructural and functional studies of Helicobacter pylori and Helicobacter mustelae flagellin mutants: Both flagellin subunits, FlaA and FlaB, are necessry for full motility in Helicobacter species. J. Bacteriol. 177:3010-3020.

137 84. Joyce, E. A., J. V. Gilbert, K. A. Eaton, A. Plaut, and A. Wright. 2001. Differential gene expression from two transcriptional units in the cag pathogenicity island of Helicobacter pylori. Infect. Immun. 69:4202-4209.

85. Kansau, I., F. Guillain, J. M. Thiberge, and A. Labigne. 1996. Nickel binding and immunological properties of the C-terminal domain of the Helicobacter pylori GroES homologue (HspA). Mol. Microbiol. 22:1013-1023.

86. Kersulyte, D., H. Chalkauskas, and D. E. Berg. 1999. Emergence of recombinant strains of Helicobacter pylori during human infection. Mol. Microbiol. 31:31-43.

87. Kostrzynska, M., J. D. Betts, J. W. Austin, and T. J. Trust. 1991. Identifications, characterization, and spatial localization of two flagellin species in Helicobacter pylori flagella. J. Bacteriol. 173:937-946.

88. Krishnamurthy, P., M. Parlow, J. B. Zitzer, N. B. Vakil, H. L. T. Mobley, M. Levy, S. H. Phadnis, and B. E. Dunn. 1998. Helicobacter pylori containing only cytoplasmic urease is susceptible to acid. Infect. Immun. 66:5060-5066.

89. Kuipers, E. J., D. A. Israel, J. G. Kusters, and M. J. Blaser. 1998. Evidence for a conjugation-like mechanism of DNA transfer in Helicobacter pylori. J. Bacteriol. 180:2901-2905.

90. Labigne, A., V. Cussac, and P. Courcoux. 1991. Shuttle cloning and nucleotide sequences of Helicobacter pylori genes responsible for urease activity. J. Bacteriol. 173:1920-1931.

91. Leying, H., S. Suerbaum, G. Geis, and R. Haas. 1992. Cloning and genetic characterization of a Helicobacter pylori flagellin gene. Mol. Microbiol. 6:2863-2874.

92. Lowry, O. H., N. J. Rosbrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265- 275.

93. Mahan, M. J., D. M. Heitoff, R. L. Sinsheimer, and D. A. Low. 2000. Assessment of bacterial pathogenesis by analysis of gene expression in the host. Annu. Rev. Genet. 34:139-164.

94. Mahan, M. J., J. M. Slauch, P. C. Hanna, A. Camilli, J. W. Tobias, M. K. Waldor, and J. J. Mekalanos. 1993. Selection for bacterial genes that are specifically induced in host tissues: The hunt for virulence factors (review). Infect. Agents Dis. 2:263-268.

138 95. Mahan, M. J., J. M. Slauch, and J. J. Mekalanos. 1993. Selection of bacterial virulence genes that are specifically induced in host tissues. Science 259:686-688.

96. Mahan, M. J., J. W. Tobias, J. M. Slauch, P. C. Hanna, R. J. Collier, and J. J. Mekalanos. 1995. Antibiotic-based selection for bacterial genes that are specifically induced during infection of a host. Proc. Natl. Acad. Sci. U.S.A. 92:669-673.

97. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1989. Molecular Cloning: A Laboratory Manual, 2 ed, vol. 1. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.

98. Marais, A., G. L. Mendz, S. L. Hazell, and F. Megraud. 1999. Metabolism and genetics of Helicobacter pylori: the genome era. Microbiol. Mol. Biol. Rev. 63:642-674.

99. Marcus, E. A., and D. R. Scott. 2001. Cell lysis is responsible for the appearance of extracellular urease in Helicobacter pylori. Helicobacter 6:93-99.

100. Marshall, B. J., L. J. Barrett, C. Prakash, R. W. McCallum, and R. L. Guerrant. 1990. Urea protects Helicobacter (Campylobacter) pylori from the bactericidal effect of acid. Gastroenterology 99.

101. Marshall, B. J., and C. S. Goodwin. 1987. Revised nomenclature of Campylobacter pyloridis. Int. J. Systematic Bacteriol. 37:68.

102. Marshall, B. J., H. Royce, D. I. Annear, C. S. Goodwin, J. W. Pearman, J. R. Warren, and J. A. Armstrong. 1984. Original isolation of Campylobacter pyloridis new species from human gastric mucosa. Microbios Lett. 25:83-88.

103. Marshall, B. J., and J. R. Warren. 1984. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet 1:1311-5.

104. Matin, A., E. Zychlinsky, M. Keyhan, and G. Sachs. 1996. Capacity of Helicobacter pylori to generate ionic gradients at low pH is similar to that of bacteria which grow under strongly acidic conditions. Infect. Immun. 64:1434- 1436.

105. McFarland, W. C., and B. A. Stocker. 1987. Effect of different purine auxotrophic mutations on mouse virulence of a vi-positive strain of Salmonella dublin and of two strains of Salmonella typhimurium. Microb. Pathogenesis 3:129-142.

139 106. McGee, D. J., and H. L. Mobley. 1999. Mechanisms of Helicobacter pylori infection: bacterial factors. Current Topics in Microbiology and Immunology 241:155-180.

107. McGowan, C. C., T. L. Cover, and M. J. Blaser. 1996. Helicobacter pylori and gastric acid: biological and therapeutic implications. Gastroenterology 110:926-38.

108. McNulty, C. A. M., and J. C. Dent. 1987. Rapid identification of Campylobacter pylori (C. pyloridis) by preformed enzymes. Journal of Clinical Microbiology 25:1683-6.

109. Megraud, F. 1992. Epidemiology of Helicobacter pylori infection, p. 107-123. In B. J. Rathbone and R. V. Heatley (ed.), Helicobacter pylori and Gastroduodenal Disease, 2 ed. Blackwell Scientific Publications, Oxford.

110. Megraud, F. 1997. Resistance of Helicobacter pylori to antibiotics. Aliment. Pharmacol. Ther. 11:43-53.

111. Megraud, F., F. Bonnet, M. Garnier, and H. Lamouliatte. 1985. Characterization of "Campylobacter pyloridis" by culture, enzymatic profile, and protein content. Journal of Clinical Microbiology 22:1007-10.

112. Mendz, G. L., B. P. Burns, and S. L. Hazell. 1995. Characterisation of glucose transport in Helicobacter pylori. Biochimica et Biophysica Acta 1244:269-276.

113. Mendz, G. L., and S. L. Hazell. 1995. Amino acid utilization by Helicobacter pylori. International Journal of Biochemistry and Cell Biology 27:1085-1093.

114. Mendz, G. L., and S. L. Hazell. 1991. Evidence for a pentose phosphate pathway in Helicobacter pylori. FEMS Microbiology Letters 84:331-336.

115. Mendz, G. L., and S. L. Hazell. 1994. Glucose metabolism by Helicobacter pylori. Microbiology 140:2179-80.

116. Mendz, G. L., and S. L. Hazell. 1993. Glucose phosphorylation in Helicobacter pylori. Archives of Biochemistry and Biophysics 300:522-525.

117. Mendz, G. L., and S. L. Hazell. 1996. The urea cycle of Helicobacter pylori. Microbiology 142:2959-67.

118. Mendz, G. L., S. L. Hazell, and B. P. Burns. 1994. The Entner- Doudoroff pathway in Helicobacter pylori. Archives of Biochemistry and Biophysics 312:349-356. 140 119. Mendz, G. L., S. L. Hazell, and B. P. Burns. 1993. Glucose utilization and lactate production by Helicobacter pylori. Journal of General Microbiology 139:3023-3028.

120. Moat, A. G., and J. W. Foster. 1995. Microbial Physiology. Wiley- Liss, Inc., New York.

121. Mobley, H. L. T. 2001. Helicobacter pylori urease, p. 155-170. In M. Achtman and S. Suerbaum (ed.), Helicobacter pylori: molecular and cellular biology. Horizon Scientific Press, Wymondham.

122. Mobley, H. L. T., M. J. Cortesia, L. E. Rosenthal, and B. D. Jones. 1988. Characterization of urease from Campylobacter pylori. J. Clin. Microbiol. 26:831-836.

123. Mobley, H. L. T., M. D. Island, and R. P. Hausinger. 1995. Molecular biology of microbial ureases. Microbiol. Rev. 59:451-480.

124. Nedenskov, P. 1994. Nutritional requirements for growth of Helicobacter pylori. Appl. Environ. Microbiol. 60:3450-3453.

125. Nedenskov-Sorensen, P., G. Bukholm, and K. Bovre. 1990. Natural competence for genetic transformation in Campylobacter pylori. J. Infect. Dis. 161:365-366.

126. Noach, L. A., T. M. Rolf, and G. N. J. Tytgat. 1994. Electron microscopic study of association between Helicobacter pylori and gastric and duodenal mucosa. Journal of Clinical Pathology 47:699-704.

127. Nomura, A., G. N. Stemmermann, P.-H. Chyou, I. Kato, G. I. Perez- Perez, and M. J. Blaser. 1991. Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. The New England Journal of Medicine 325:1132-6.

128. Parsonnet, J., G. D. Friedman, D. P. Vandersteen, Y. Chang, J. H. Vogelman, N. Orentreich, and R. K. Sibley. 1991. Helicobacter pylori infection and the risk of gastric carcinoma. The New England Journal of Medicine 325:1127-31.

129. Peekhaus, N., and T. Conway. 1998. What's for Dinner? Entner- Dourdoroff Metabolism in Escherichia coli. Journal of Bacteriology 180:3495- 3502.

130. Peterson, W. L. 1991. Helicobacter pylori and peptic ulcer disease. N. Engl. J. Med. 324:1043-1048.

141 131. Phadnis, S. H. 1993. Molecular cloning of Helicobacter pylori DNA: important differences between mcrBC deletion host strains. mol. Microbiol. 10:1151.

132. Phadnis, S. H., M. H. Parlow, M. Levy, D. Ilver, C. M. Caulkins, J. B. Connors, and B. E. Dunn. 1996. Surface localization of Helicobacter pylori urease and a heat shock protein homolog requires bacterial autolysis. Infect. Immun. 64.

133. Pitson, S. M., G. L. Mendz, S. Srinivasan, and S. L. Hazell. 1999. The tricarboxylic acid cycle of Helicobacter pylori. European Journal of Biochemistry 260:258-67.

134. Pounder, R. E., and D. Ng. 1995. The prevalence of Helicobacter pylori in different countries. Aliment. Pharmacol. Ther. 9:S33-S39.

135. Quigley, E. M. M., and L. A. Turnberg. 1987. pH of the microclimate lining human gastric and duodenal mucosa in vivo. Gastroenterology 92:1876- 1883.

136. R. M. Peek, J., S. A. Thompson, J. P. Donahue, K. T. Tham, J. C. Atherton, M. J. Blaser, and G. G. Miller. 1998. Adherence to gastric epithelial cells induces expression of a Helicobacter pylori gene, iceA, that is associated with clinical outcome. Proc. Assoc. Am. Phys. 110.

137. Rathbone, B. J., and R. V. Heatley. 1992. The historical associations between bacteria and peptic ulcer disease, p. 1-4. In B. J. Rathbone and R. V. Heatley (ed.), Helicobacter pylori and gastroduodenal disease, 2nd ed. Blackwell Scientific Publications, Oxford.

138. Rauws, E. A. J., W. Langenberg, H. J. Houthoff, H. C. Zanen, and G. N. Tytgat. 1988. Campylobacter pyloridis-associated chronic active antral gastritis. Gastroenterology 94:33-40.

139. Rektorschek, M., A. Buhmann, D. Weeks, D. Schwan, K. W. Bensch, S. Eskandari, D. Scott, G. Sachs, and K. Melchers. 2000. Acid resistance of Helicobacter pylori depends on the UreI membrane protein and an inner membrane protein barrier. Mol. Microbiol. 36:141-152.

140. Reynolds, D. J., and C. W. Penn. 1994. Characteristics of Helicobacter pylori growth in a defined medium and determination of its amino acid requirements. Microbiology 140:2649-56.

141. Rollason, T. P., J. Stone, and J. M. Rhodes. 1984. Spiral organisms in endoscopic biopsies of the human stomach. Journal of Clinical Pathology 37:23- 6. 142 142. Sakagami, T., M. Dixon, J. O'Rourke, R. Howlett, F. Alderuccio, J. Vella, T. Shimoyama, and A. Lee. 1996. Atrophic gastric changes in both Helicobacter felis and Helicobacter pylori infected mice are host dependent and separate from antral gastritis. Gut 39:639-648.

143. Savage, D. C. 1977. Microbial ecology of the gastrointestinal tract. Annual Review of Microbiology 31:107-33.

144. Scott, D. R., D. Weeks, C. Hong, S. Postius, K. Melchers, and G. Sachs. 1998. The role of internal urease in acid resistance and Helicobacter pylori. Gastroenterology 114:58-70.

145. Segal, E. D., S. Falkow, and L. S. Tompkins. 1995. Helicobacter pylori attachment to gastric cells induces cytoskeletal rearrangements and tyrosine phosphorylation of host cell proteins. Proc. Natl. Acad. Sci. U.S.A. 93:1259-1264.

146. Segal, E. D., C. Lange, A. Covacci, L. S. Tompkins, and S. Falkow. 1997. Induction of host signal transduction pathways by Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 94:7595-7599.

147. Sipponen, P., and H. Hyvarinen. 1993. Role of Helicobacter pylori in the pathogenesis of gastric, peptic ulcer and gastric cancer. Scand. J. Gastroentero. 28 (S. 196):1-6.

148. Skouloubris, S., J.-M. Thiberge, A. Labigne, and H. DeReuse. 1998. The Helicobacter pylori UreI protein in not involved in urease activity but is essential for bacterial survival in vivo. Infect. Immun. 66:4517-4521.

149. Smith, H. 1998. That happens to bacterial pathogens in vivo? Trends Microbiol. 6:239-243.

150. Stark, R. M., J. Greenman, and M. R. Millar. 1995. Physiology and biochemistry of Helicobacter pylori. British Journal of Biomedical Science 52:282-290.

151. Stark, R. M., M.-S. Suleiman, I. J. Hassan, J. Greenman, and M. R. Millar. 1997. Amino acid utilisation and deamination of glutamine and asparagine by Helicobacter pylori. Journal of Medical Microbiology 46:793- 800.

152. Suerbaum, S., C. Josenhaus, and A. Labigne. 1993. Cloning and genetic characterization of the Helicobacter pylori and Helicobacter mustelae flaB flagellin genes and construction of H. pylori flaA- and flaB-negative mutants by electroporation-mediated allelic exchange. J. Bacteriol. 175:3278- 3288. 143 153. Suerbaum, S., J. M. Smith, K. Bapumia, G. Morelli, N. H. Smith, E. Kunstmann, I. Dyrek, and M. Achtman. 1998. Free recombination within Helicobacter pylori. Proc. Natl. Acad. Sci. U.S.A. 95:12619-12624.

154. Sweeney, N. J., P. Klemm, B. A. McCormick, E. Moller-Nielsen, M. Utley, M. A. Schembri, D. C. Laux, and P. S. Cohen. 1996. The Escherichia coli K-12 gntP gene allows E. coli F-18 to occupy a distinct niche in the streptomycin-treated mouse large intestine. Infection and Immunity 64:3497- 3503.

155. Sweeney, N. J., D. C. Laux, and P. S. Cohen. 1996. Escherichia coli F- 18 and E. coli K-12 eda mutants do not colonize the streptomycin-treated mouse large intestine. Infection and Immunity 64:3504-3511.

156. Thompkins, D. S. 1992. Isolation and characterization of Helicobacter pylori, p. 19-28. In B. J. Rathbone and R. V. Heatley (ed.), Helicobacter pylori and Gastroduodenal Disease, 2 ed. Blackwell Scientific Publications, Oxford.

157. Tomb, J.-F., others, and J. C. Venter. 1997. The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature 388:539-547.

158. Trieu-Cuot, P., G. Gerbaud, T. Lambert, and P. Courvalin. 1985. In vivo transfer of genetic information between gram-positive and gram-negative bacteria. The EMBO Journal 4:3583-3587.

159. Tytgat, G., and M. Dixon. 1993. Overview: Role in peptic ulcer disease, p. 75-87. In T. C. Northfield, M. Mendall, and P. M. Goggin (ed.), Helicobacter pylori infection. Kluwer Academic Publishers, Boston.

160. Valdivia, R. H., and S. Falkow. 1997. Fluorescence-based isolated of bacterial gene expressed within host cells. Science 277:2007-2011.

161. van der Ende, A., Z.-J. Pan, A. Bart, R. W. M. van der Hulst, M. Feller, S.-D. Xiao, G. N. J. Tytgat, and J. Dankert. 1998. cagA-positive Helicobacter pylori populations in China and the Netherlands are distinct. Infect. Immun. 66:1822-1826.

162. van Doorn, L.-J., C. Figueiredo, R. Sanna, S. Pena, P. Midolo, E. K. Ng, J. C. Atherton, M. J. Blaser, and W. G. Quinn. 1998. Expanding allelic diversity of Helicobacter pylori vacA. J. Clin. Microbiol. 36:2697-2603.

163. Voet, D., and J. G. Voet. 1990. Biochemistry. John Wiley & Sons, Inc., New York.

144 164. Wang, J., S. Lory, R. Ramphal, and S. Jin. 1996. Isolation and characterization of Pseudomonas aeruginosa genes inducible by respiratory mucus derived from cystic fibrosis patients. Mol. Microbiol. 22:1005-1012.

165. Warren, J. R., and B. Marshall. 1983. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. The Lancet:1273-5.

166. Weeks, D. L., S. Eskandara, D. R. Scott, and G. Sachs. 2000. A H+- gated urea channel: the link between Helicobacter pylori urease and gastric colonization. Science 287:482-485.

167. Wilson, G. G., and N. E. Murray. 1991. Restriction and modification systems. Annu. Rev. Genet. 25:585-627.

168. Yamoaka, Y., T. Kodama, K. Kashima, D. Y. Graham, and A. R. Apeulveda. 1998. Variants of the 3' region of the cagA gene in Helicobacter pylori isolates from patients with different H. pylori-associated diseases. J. Clin. Microbiol. 36:2258-2263.

169. Young, G. M., and V. L. Miller. 1997. Identification of novel chromosomal loci affecting Yersinia enterocolitica pathogenesis. Mol. Microbiol. 25:319-328.

145