The Pennsylvania State University

The Graduate School

Department of Biochemistry and Molecular Biology

RESPIRATORY PATHOGEN INTERACTIONS WITH THE

MICROBIOME: THE MOLECULAR MECHANISMS TO THE ETHICAL

IMPLICATIONS

A Dissertation in

Biochemistry, Microbiology, and Molecular Biology

by

Laura S. Weyrich

 2012 Laura S. Weyrich

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2012

i

The dissertation of Laura S. Weyrich was reviewed and approved* by the following:

Eric T. Harvill Associate Professor of Microbiology and Infectious Diseases Dissertation Advisor Chair of Committee

Robert F. Paulson Associate Professor of Veterinary and Biomedical Science

Richard Frisque Professor of Molecular Virology

Jean E. Brenchley Professor of Microbiology and Biotechnology

Jess F. Ballenger Associate Professor of Science, Technology, and Society

Craig Cameron Professor of Biochemistry and Molecular Biology Graduate Program Head for the Department of Biochemistry and Molecular Biology

*Signatures are on file in the Graduate School

iii

ABSTRACT

This work explores how interactions between the respiratory tract microbiome and common pathogens may contribute to pathogenesis and interrogates the ethical implications associated with such research. Three closely related, widespread respiratory pathogens,

Bordetella bronchiseptica, Bordetella pertussis, and Bordetella parapertussis, were implemented to understand how pathogens can utilize classical virulence factors to overcome host flora, and, conversely, how microflora can be utilized to prevent pathogen colonization. After initially characterizing a locus in B. bronchiseptica predicted to mediate bacterial competition in vitro, the Type VI Secretion System (T6SS), we created a non-functional T6SS mutant to investigate how this pathogen uses this secretion system to initially induce pathology, modulate the immune response, and persist in the respiratory tract. The mechanism underlying persistence was later investigated; the T6SS is capable of modulating lipid antigen presentation and NK T cell accumulation at the site of infection, causing deleterious downstream effects on antibody production. Interestingly, although this secretion system has significant effects on the host and the immune response, the T6SS was also shown to mediate bacterial competition in our natural infection system. B. bronchiseptica requires a function T6SS, a Type III Secretion System, and the master regulatory system, BvgAS, to displace nasal cavity microflora organisms in vivo and out-compete host organisms in vitro. We also determined that bacterial competition within the respiratory tract can be immune-mediated; cytotoxic T-cells and IL-17, which can additionally cause the release of antimicrobial compounds, are also required to displace nasal cavity flora in vivo. Interestingly, cytotoxic T cell-mediated microflora displacement was shown to be an essential mechanism for initial colonization. In stark contrast, the human restricted pathogen, B.

iv

pertussis, did not displace murine host flora, and single host organisms were capable of blocking

B. pertussis colonization. Furthermore, mice that had been antibiotic treated, and therefore were

lacking normal microflora, were shown to be more susceptible to B. pertussis infection. As a

first step to understanding how the T6SS is regulated and therefore understanding when this

factor is utilized by these pathogens, we discovered small RNA regulatory molecules that target

a putative T6SS effector. Lastly, to correlate our murine research to humans, we used

metagenomic sequencing to identify bacterial flora in the lower respiratory tract of healthy

humans, as well as patients with Chronic Obstructive Pulmonary Disease and asthma, to

understand how the microbiome can contribute to health and disease. Finally, the ethical

implications of this experimental research were examined, in terms of identifying the best

method for teaching effective ethical training during graduate research and, secondly, determining the ethical and legal implications of patenting human isolated microbes. Together, this dissertation examines both the molecular interactions between pathogens, the microflora, and the respiratory tract and investigates the ethical issues generated from this research.

v

TABLE OF CONTENTS

LIST OF FIGURES ...... viii LIST OF TABLES ...... x ABBREVIATIONS...... xiii ACKNOWLEDGEMENTS ...... xiv

Chapter 1 : Introduction ...... 1

The Respiratory Tract: The Microbiome and Disease ...... 2 Bacterial Competition ...... 4 Bordetella: A Diverse Genus of Respiratory Pathogens ...... 7 Bordetellae virulence factors and their regulation...... 11 Ethical implications of biomedical research...... 14 References ...... 18

Chapter 2 : A Type VI Secretion System Encoding Locus is Required for Bordetella bronchiseptica Immunomodulation and Persistence In Vivo ...... 29

Abstract ...... 30 Introduction ...... 31 Materials and Methods ...... 34 Results ...... 41 Discussion ...... 53 Authors and Contributions ...... 57 References ...... 58

Chapter 3 : CD1d Manipulation by the Bordetella bronchiseptica Type VI Secretion System Modulates Antibody Production and Persistence ...... 65

Abstract ...... 66 Introduction ...... 67 Materials and Methods ...... 70 Results ...... 73 Discussion ...... 83 Author Contributions: ...... 88 References ...... 89

Chapter 4 : Bordetella bronchiseptica Requires Cytotoxic T-cells to Displace Upper Respiratory Tract Microflora during Infection ...... 94

Abstract ...... 95 Introduction ...... 96 Results and Discussion ...... 99 Materials and Methods ...... 112 Author Contributions: ...... 117 References ...... 118

vi

Chapter 5 : Murine Upper Respiratory Tract Microflora Inhibit Bordetella pertussis Colonization123

Abstract ...... 124 Introduction ...... 125 Material and Methods ...... 128 Results ...... 131 Discussion ...... 138 Author Contributions: ...... 142 References ...... 143

Chapter 6 : Bordetella bronchiseptica Small RNA Molecules are Predicted to Target Major Virulence Factors and Contribute to Strain Specificity ...... 147

Abstract ...... 148 Introduction ...... 149 Materials and Methods ...... 152 Results ...... 155 Discussion ...... 172 Author Contributions: ...... 176 References ...... 177

Chapter 7 : Respiratory Tract Microbiome Dysbiosis Contributes to Obstructive Pulmonary Disease and Asthma ...... 182

Abstract ...... 183 Introduction ...... 184 Materials and Methods ...... 186 Results ...... 189 Discussion ...... 199 Author Contributions: ...... 202 References ...... 203

Chapter 8 : Teaching Ethical Aptitude to Graduate Student Researchers...... 206

Author Contributions: ...... 214 References ...... 215

Chapter 9 : Is there a solution to microbiome patenting? ...... 217

Scenario #1: Corporate patenting...... 220 Scenario #2: Allowing individuals rights to their microbiome...... 222 Scenario #3: No microbiome patenting allowed...... 224 A Potential Solution ...... 225 Author Contributions: ...... 227 References ...... 228

Chapter 10 : Summary and Significance ...... 230

Overall Summary and Significance ...... 231

vii

Type VI Secretion System: A Novel Virulence Determinant in Bordetella ...... 232 Novel Competition Mechanisms between Respiratory Pathogens and Host Microflora . 236 Microbial competition as novel disease treatments and preventions...... 241 Contributions to the Human Respiratory Microbiome ...... 244 Ethical Implications of Biomedical and Microbiome Research ...... 244 References ...... 246 Appendix A: SIPHT Predicted Small RNAs ...... 250 Appendix B: TransTerm predicted small RNA terminators...... 262

viii

LIST OF FIGURES

Figure 2-1 Genetic comparison of B. bronchiseptica T6SS locus to known T6SS loci...... 41

Figure 2-2 Confirmation of RB50ΔclpV construction and T6SS expression...... 43

Figure 2-3 T6SS mediated cytotoxicity of murine macrophages...... 44

Figure 2-4 T6SS-mediated proteomic changes during in vitro macrophage infection...... 46

Figure 2-5 T6SS-mediate cytokine production in vitro...... 47

Figure 2-6 Colonization and histopathology of RB50ΔclpV compared to wild type...... 49

Figure 2-7 Effects of the T6SS on cytokine production in vivo...... 51

Figure 3-1 Lung colonization and antibody titers in RB50 vs RB50ΔclpV infected mice. .... 73

Figure 3-2 IgM and IgG2a titers from mice infected with either RB50 or RB50ΔclpV...... 74

Figure 3-3 Immune cells during RB50 or RB50ΔclpV infection over time in the cervical lymph node, mediastinal lymph node, or lungs...... 75

Figure 3-4 MHC-II and CD1d surface expression on BMDCs at 2 and 6 hours...... 77

Figure 3-5 CD1d surface expression on immune cells in the cervical lymph nodes or lungs. 78

Figure 3-6 NK T cell numbers in the lungs and antibody production in NK T cell deficient mice.80

Figure 3-7 In vitro BMDC cytotoxicity caused by either RB50 or RB50ΔclpV...... 81

Figure 4-1 B. bronchiseptica displaces culturable murine nasal cavity flora...... 99

Figure 4-2 T3SS, T6SS, and BvgAS are required to displace flora in vitro...... 104

Figure 4-3 T-cells, IFN-γ, and IL-17 are required to displace microflora in vivo...... 106

Figure 4-4 CD4+CD8+ and CD8+ T-cells are recruited to the nasal cavity and required for microflora displacement...... 108

Figure 4-5 Host microflora, but not B. bronchiseptica, are susceptible to granzyme B in vitro. 109

Figure 5-1 104 CFU B. pertussis are required to colonize murine nasal cavity...... 131

Figure 5-2 Baytril treatment reduced B. pertussis ID50...... 132

Figure 5-3 Murine nasal cavity flora inhibits B. pertussis growth...... 133

ix

Figure 5-4 Inoculation with known microflora inhibits B. pertussis colonization...... 134

Figure 5-5 Zoonotic pathogen B. bronchiseptica can displace murine flora...... 136

Figure 6-1 Small RNAs are distributed evenly throughout the B. bronchiseptica genome. .... 161

Figure 6-2 Small RNAs are commonly transcribed near the 5’ end of an ORF...... 162

Figure 6-3 Northern blotting of 4 identified sRNAs...... 166

Figure 6-4 Bbsrna 23 and 70 bind a similar region upstream of tssJ...... 169

Figure 6-5 B. bronchiseptica sRNAs are conserved amongst the classical bordetellae...... 170

Figure 7-1 Blood Agar isolates the most sputum microorganisms...... 189

Figure 7-2 Frequency and colonization of cultured host microflora species...... 192

Figure 7-3 Read frequency according to bacterial class...... 193

Figure 7-4 Bacterial community structure and relationships amongst patients...... 195

Figure 7-5 Species-based phylogenetic analysis of all patients...... 198

x

LIST OF TABLES

Table 1 Cross-streaking competition...... 102

Table 2 Host microflora isolated from nasal cavities of healthy mice...... 133

Table 3 Small RNA molecules identified in B. bronchiseptica strain RB50...... 156

Table 4 Small RNAs conserved amongst SIPHT prediction and SOLiD sequencing...... 160

Table 5 Small RNA molecules with predicted σE binding sites upstream...... 163

Table 6 Virulence factor targets of sequenced sRNAs...... 167

Table 7 – Microorganisms cultured from COPD, asthma, NID, or healthy patients...... 190

xiii

ABBREVIATIONS

T3SS, Type III Secretion System

ACT, adenylate cyclase toxin

FHA, filamentous hemagglutinin

FIM, fimbriae

DNT, dermonecrotic toxin

TCT, tracheal cytotoxin

T6SS, Type VI Secretion System

PBS, phosphate buffered saline

PCR, polymerase chain reaction

IL, interleukin

IFN-γ, interferon-gamma

APC, antigen presenting cell

NK, natural killer

MHC, major histocompatibility complex

Bvg, Bordetella virulence genes sRNA, small RNA mRNA, messenger RNA

T2SS; Type II Secretion System

COPD, Chronic Obstructive Pulmonary Disease

NID, no identifiable disease

xiv

ACKNOWLEDGEMENTS

I would like to thank Craig, Candace, Deb and Greg at The Pennsylvania State University

Genomic Core Facility for completing DNA and RNA sequencing in Chapters 2, 4, 5, and 6, the

Microscopy and Cytometry Facility for assistance in collecting and analyzing flow cytometry

data in Chapters 3 and 4, and the Proteomics and Mass Spectrometry Core Facility for protein

identification via mass spectrometry in Chapter 2. I would like to extend my gratitude to Dr.

Mark Exley at Harvard University for supplying the mice deficient in CD1d used in Chapter 3

and to Andrew Gunderson at Penn State for providing α-CD4 and α-CD8 antibodies used in

Chapter 4. I would additionally like to thank Dr. Istvan Albert for help running Mothur software

utilized in Chapter 7 and Dr. Edward Dudley for sharing his T6SS knowledge with us in Chapter

2. Lastly, I would like to thank the entire COPD/asthma project team, including Dr. Avery

August, Dr. Sandeep Prabhu, Dr. Margherita Cantorna, Dr. Timothy Craig, Cathy Mende, and

Crystal Rhodes for obtaining funding, collecting samples, and data analysis presented in Chapter

7. For work completed in the final chapters, I would like to thank Dr. Jesse Ballenger for

insightful ethics discussions and direction.

Next, I would like to extend my full gratitude to the National Science Foundation for sponsoring me throughout my dissertation research and proving that there is a place for ethics in science. I would like to thank my committee members Dr. Richard Frisque, Dr. Jean Brenchley,

Dr. Robert Paulson, and Dr. Jesse Ballenger for their insight, suggestions, and support throughout the years. I would additionally like to thank Dr. Ballenger and Dr. Jonathan Marks for encouraging me to pursue my philosophy and ethics interests at Penn State. Lastly, I would

xv like to thank my advisor, Dr. Eric Harvill for giving me the freedom to complete the research that I love and for reining me in when I needed it.

I would like to thank all of past and present Harvill lab members, graduate students and undergrads, for their never ending support, increasing motivation, incredible ideas, and for being shoulders to lean on; they really are the smartest bunch of students this side of the nut house. To past members, Dan, Liz, Anne, Girish, Tanya, and Xuqing, thank you for being outstanding examples and setting the bar high. To all of my undergraduates, Maggie, Dan, Sashah,

Mackenzie, Nate, and Sarah, you’ve made me a stronger scientist and I hope you stay in touch. I want to know how wildly successful you are some day. To Heather “Figgie Pudding” Feaga, thanks for reminding me I’m still a SD farm girl at heart and that there is room for biochemistry in this lab. To Will “Spike the Ball” Smallridge, thanks for going along with all of my intramural sports endeavors and listening to my wedding planning frustrations. To Laura “Mini- me” Goodfield, thanks for always feeding me and being hugely optimistic. To Sarah “Julia

Child” Muse, thanks for sharing your recipes and for your creative experimental insight. To

David “Rock Your Socks Off” Place, thanks for blasting your great tunes and creeping in on conversations once and awhile (your insights are beneficial). To Liron “Party Animal” Bendor, thanks for keeping us all young and spry. Your happy-go-lucky attitude at life has worn off on me for the better. To Jihye “Korean Genius” Park, your drive and determination have motivated me for years. Thank you for going along with every sequencing project I’ve ever conjured up.

To Alexia “Sparty” Karanikas, thanks for being so supportive all these years. You always had a kind thing to say and great advice to share. Go green! “To Olivier “Oh man!” Rolin, I can’t thank you enough for helping conceptualize my projects, being a spring board, and teaching me the measly amount of French and flow cytometry I know. To Sara “Polar Bear” Hester, thanks

xvi

for commiserating with me for years and always being there to help out on a project or dinner.

You’ve become a true friend. I look forward to the phenomenal scientists and people you all

become. I can’t thank you all enough.

Lastly, thanks to my friends and family. There are honestly too many to name, but you know who you are. To my parents, aunts, uncles, cousins, grandparents, childhood friends, and

new amigos, thank you for supporting me and giving me the encouragement to be somebody.

Mom and Dad, I hope you know that your encouragement, love, and the never-quit attitude you

instilled in me has gotten me far. Mom, I know now why can’t never got out of bed. To Dad, thanks for showing me that common sense is really all you need. To Keith, you’ve been the most supportive big, little brother a girl could ask for. You constantly amaze me with your strength, optimism, and outlook on life. Come home safely. To my Grandparents, thanks for being role models I can look up to and being the people I am still trying to become. To Kathleen

Donaldson, your strength, honor, and true friendship got me here. Thanks for going along with

every crazy idea I’ve ever had. Lastly, so many thanks to the love of my life, Justin VanderBerg.

You’re never ending support, encouragement, and love got me through writing all 64,763 words.

I hope you know that I seriously couldn’t have done this without you. I’ll be there when you

write yours, to remind you to lift with your back and not your legs. I love you.

1

Chapter 1 : Introduction

2

The Respiratory Tract: The Microbiome and Disease

Respiratory tract microbiome. Mammals inspire unfiltered air from the outside world into the lower respiratory tract, where oxygen replaces carbon dioxide by simple diffusion in small, delicate alveolar sacs in the lungs. The respiratory tract has numerous defense mechanisms to ward off invading bacterial species, so the lungs were once thought to be sterile in order to maintain this delicate process that supports life. The first signs of microbial respiratory tract inhabitants were observed in the 1920’s, but researchers at University of Pennsylvania and

Imperial College in London have now made it clear that the lower respiratory tract indeed is teaming with microorganisms that may affect the lung physiology or disease susceptibility [1].

These lower respiratory tract organisms include Niesseria and Haemophilus species, but interestingly, not Staphylococcus species, which frequently colonizes the upper respiratory tract

(W. Cookson, SMI ICMI, Paris, France). Higher up in the respiratory tract, the human pharynx can contain Niesseria, Cornybacterium, and Streptococcus species, while microflora from the nasal cavities of healthy humans can be much more diverse, comprised of Staphylococcus,

Streptococcus, Niesseria, Enterobacteriacea, Proteus, Cornyebacterium, and Mycobacterium species [2–6]. Interestingly, camels, donkeys, and sea lions from the United States, Mexico,

Saudia Arabia, and Ethiopia all contain similar upper respiratory tract microflora, suggesting that colonization by many of these bacterial species is conserved amongst a wide range of mammals

[7–9]. New technology utilizing the 16S rRNA encoding genome region for bacterial identification has exponentially increased our knowledge of unculturable respiratory tract inhabitants, and the Human Microbiome Project will identify the nasal cavity microbial communities of over 200 healthy human adults over the next three years [10].

3 Obstructive Respiratory Diseases. Common respiratory tract diseases can be grouped into four categories: obstructive, restrictive, vascular, or infectious; the work of this dissertation focuses on both obstructive and infectious respiratory diseases. Obstructive respiratory disease includes asthma, chronic obstructive pulmonary disorder (COPD), emphysema, and bronchitis.

The World Health Organization (WHO) currently estimates that 35 million people suffer from

COPD, while 235 million people suffer from asthma worldwide (2004) [11]. Both COPD and asthma are caused by hardened and dysfunctional alveolar air sacs and inflamed and hardened bronchi. While the etiologies of many obstructive respiratory diseases are well understood, how the respiratory tract microbiome and pathogen colonization contribute to disease symptoms, prevention, and long-term progress is poorly understood. We currently know that COPD patients who are treated with antibiotics have lower relapse rates and decreased symptoms, suggesting that there is a largely unstudied bacterial component to these diseases [12,13]. We additionally can associate numerous pathogens with both COPD and asthma patients, namely

Pseudomonas species, but not all people colonized by Pseudomonas have COPD and asthma, indicating that there is still much to learn about how pathogens contribute to these diseases [14–

19]. Furthermore, both COPD and asthma can be exacerbated by cigarette smoking, which has been shown to decrease the number of culturable organisms within the respiratory tract, adding an additional level of complexity when attempting to understand how microbial communities can contribute to obstructive respiratory diseases [17,20].

Infectious Respiratory Diseases. Infectious respiratory diseases are caused by countless different viral, fungal, and bacterial species. Currently, pneumonia is the leading cause of death in children worldwide, killing an estimated 1.4 million children under the age of 5 per year, and is primarily caused by Mycobacterium, Pseudomonas, Bacillus, Francisella, or Bordetella

4 species [11]. Often, death from these infections is caused by uncontrolled fluid within the airway or uncontrolled inflammatory responses that cause irreversible damage to the alveolar air sacs. Although many of these diseases have murine models of infection that allow us to understand disease kinetics, the immune response, and how to control these pathogens in a monitored environment, little work has been done on how these pathogens overcome the respiratory tract microflora to cause infection, how microflora may contribute to disease severity,

and how microbes within a host may prevent pathogen colonization.

Bacterial Competition

Direct Competition. Charles Darwin noted that the toughest competition between species

is amongst similar species that share environmental niches. Within the respiratory tract, for

example, is a diverse, microbial rich environment, where invading pathogens must compete for

nutrients, space, and survival before they can cause infection. Here, direct competition will be

defined similarly to interference competition, where one species directly impedes the ability of

another organism to be successful, particularly by targeted mechanisms from one organism

toward another to achieve a competitive advantage [21]. An example of this would be the

release of colicins, or bacteria toxins, by Escherichia coli to kill other related E. coli as a

competition or survival mechanism [22,23]. Another example includes the use of phage, or

bacterial viruses, released by bacteria attempting to invade a particular environment; for

example, Salmonella, Clostridium, and E. coli O157::H7 utilize a variety of prophages to assist

in digestive tract invasion [24–26]. Interactions within biofilms have also revealed some

interesting direct competition mechanisms, such as the ability of Pseudomonas aeruginosa to

produce molecules that displace Bordetella bronchiseptica from active biofilms [27]. Recently,

5 a newly discovered bacterial secretion system in Pseudomonas, the Type VI Secretion System

(T6SS), was shown to inject a toxin into other bacterial species, while the toxin secreting strain

had an anti-toxin to protect itself [28]. This T6SS-dependent observation has also been observed

in Vibrio and Serratia, suggesting that a novel mechanism behind direct bacterial toxin secretion

has been identified [29,30]. However, how this T6SS-associated toxin/antitoxin injection system

contributes to initial pathogen colonization, especially in the respiratory tract, and the ability to

compete with host microflora in vivo has not been shown.

Indirect Competition. Another means of bacterial competition is indirect competition, or in

other words, when one species loses out because of the ability of another species to obtain limiting factors, such as nutrients. Ecologists refer to this type of competition as exploitation competition, which is a passive process, where the amount of available resources is affected by one species as a means of effectively competing for nutrients in any given environment [21].

Additionally, resource competition can be divided into two broad groups: scramble and contest, where scramble is defined as passively, yet rapidly, using a limiting resource, making it unavailable to other microbes, and contest is defined similarly to direct or interference competition, in which neighboring microbes are killed in order to be scavenged for nutrients

[31]. Indirect competition within this dissertation will only refer to the passive mechanisms of competition. Examples of this include pathogens that have remarkable abilities to scavenge iron a competitive mechanism [32]; for example, iron-scavenging siderophore production in

Pseudomonas can provide a competitive advantage over other microbes, especially when competitor species are phylogenetically distinct [33]. Nutrient availability may also provide a competitive advantage to particular microbes, as the use of pre-biotics, such as dietary carbohydrates are being investigated to alter the gut microbiotic [34]. Furthermore, complex

6 symbiotic, co-evolved relationships have been discovered using exploitation competition as a

means for interactions between species; for example the rock-paper-scissors theory, where one species feeds from another, which may provide advantages for yet another species [23].

Immune Mediated Competition. Although direct and indirect microbial competition can be

highly effective, pathogens have also evolved mechanisms to provoke the immune response as

another means to compete against other microbes, which will be defined in this dissertation as

immune-mediated competition [23]. This is similar to “apparent competition”, where a single

predator (the immune response) preys upon both species, of which one is more suited to survive,

resulting in a negative effect against the other species [21]. Numerous examples of this exist

within the gut, as pathogens must use any means possible to survive amongst over 1 trillion other

resident microbes present. For example, E. coli uses toxins to induce a beneficial immune

response, while Salmonella has been shown to trigger inflammatory responses that change gut

microbiota communities [23,35]. However, very few examples of immune-mediated

competition between pathogens and respiratory tract flora exist. We do know that, as a

competitive mechanism, Haemophilus influenze can out-compete Stretococcus pneumonia in the upper respiratory tract by recruiting and activating neutrophils, to which S. pneumonia is more susceptible [36]. Each respiratory pathogen may have a novel mechanism in which it directly or indirectly competes with microflora or manipulates the immune system to benefit itself.

Understanding how pathogens achieve colonization amidst established flora is a largely undeveloped, burgeoning field.

7 Bordetella: A Diverse Genus of Respiratory Pathogens

Bordetella species. Bordetella is a diverse genus of Gram-negative coccobacilli bacteria, consisting of Bordetella pertussis, Bordetella parapertussis, Bordetella bronchiseptica,

Bordetella holmesii, Bordetella hinzii, Bordetella avium, Bordetella trematum, Bordetella petri, and Bordetella ansorpii [37,38]. The classical, and arguably the most studied, bordetellae consist of B. pertussis, B. parapertussis, and B. bronchiseptica, which commonly cause respiratory tract infections. B. pertussis and B. parapertussis cause the acute respiratory disease, whooping cough or pertussis in humans [37,39]. Unlike the two human pathogens, B. bronchiseptica causes respiratory disease in numerous mammals, including sea-lions, tigers, bears, dogs, rabbits, cats and humans, inducing a wide range of diseases, from asymptomatic colonization to lethal pneumonia [37,40]. Most of what is currently known about the bordetellae comes from a highly robust, rigorous murine infection model, which has led to our current understanding of disease kinetics, immune response during bordetellae infection, and vaccine immunity.

Evolutionary relationships between the classical bordetellae. The evolutionary relationship between several classical bordetellae lineages has been examined using multi-locus sequence typing (MLST), where seven housekeeping genes are sequenced and the single-nucleotide polymorphisms are then used to determine evolutionary divergence and diversity. B. pertussis and B. parapertussis were each found to cluster independently, and B. bronchiseptica was found to separate into two distinct lineages [41]. Of the classical bordetellae, B. bronchiseptica has the largest genome, over 5.3 Mb in size, while the two human pathogens, B. pertussis and B. parapertussis, have much smaller genomes, at 4.1 and 4.7 Mb, respectively; however, up to 86% of their genes are conserved amongst all three genomes [42]. B. bronchiseptica maintains 602

8 unique genes, which are proposed to be involved in membrane transport, small-molecule metabolism, gene regulation, and surface structure synthesis, all of which reflect the ability of B.

bronchiseptica to survive environmentally, an ability the two human pathogens lack [40,43,44].

These observations led to the belief that B. pertussis and B. parapertussis have evolved over time

from a B. bronchiseptica-like progenitor species, perhaps driven by host adaptation. This

observation prompts questions about the energetic cost for bacterial strains to maintain genes that

allow host diversity and environmental survival.

Human Bordetella disease. B. pertussis and B. parapertussis cause an acute respiratory

disease marked by the “whoop” sounds made as patients rapidly inspire air in between coughing

fits. Whooping cough starts in the catarrhal stage, which lasts up to one week and looks similar

to the common cold (sneezing, running nose, low-grade fever, and occasional light cough).

Next, the paroxysmal stage begins when massive coughing bursts occur, because of the inability

to expel thick mucus, which can induce vomiting and cyanosis. Surprisingly, most patients do

not feel ill in between attacks, and infants may not even cough at all, making this difficult to

diagnose, especially in small children [45].

Pertussis vaccination and epidemiology. A highly successful, effective whole-cell vaccine against B. pertussis was developed in the 1930’s, which reduced pertussis cases by 80% in the

United States, decreasing the number of cases from over 200,000 in 1940 to fewer than 3,000 per year in the 1980’s [45]. An effective acellular B. pertussis vaccine was also developed in the

1980s and is the current recommended vaccine in the United States and numerous other developed countries [45–47]. Although this vaccine is highly effective in developed countries, whooping cough still kills over 240,000 people annually worldwide [11]. However, even despite high vaccination coverage in developed countries, whooping cough cases has made a comeback

9 in the past decade, causing over 25,000 cases and 27 deaths in the United States in 2010, most of

which were in children under one year of age. Whooping cough cases in Europe have also largely increased in recent years [48–51]. The resurgence of B. pertussis may be caused by waning immunity in adolescents or immune evasion through strain evolution [52–55].

Increasing global pertussis cases has provoked the construction of novel vaccines and vaccination strategies, including a live, attenuated strain of B. pertussis as an intranasal vaccine

[56].

Whooping cough diagnosis and control are further compounded by B. parapertussis, which is currently not reportable to the Center for Disease Control in the United States. B. parapertussis infections induce nearly identical symptoms compared to B. pertussis infections and are usually only distinguished by PCR testing [57,58]. Interestingly, a large immunogenic

protein on the outer surface of B. parapertussis, O-antigen, allows B. parapertussis to evade B. pertussis-induced immunity by blocking antibody binding [59]. Furthermore, acellular B.

pertussis vaccines are also not highly effective against B. parapertussis, because B.

parapertussis lacks pertussis toxin, a toxin unique to B. pertussis species and currently the main

component of several acellular vaccines, such as Tdap [37]. Recently, our laboratory also showed that acellular B. pertussis vaccination provides an advantage for B. parapertussis in the host, suggesting that current B. pertussis whooping cough vaccines may predispose patients to B. parapertussis infection [60]. Whooping cough infections are even further compounded by the recent identification of B. holmesii from the respiratory tracts of patients suffering with whooping cough like symptoms, suggesting an additional bordetellae species also has the ability to evade B. pertussis immunity and may have carved a human niche within a highly B. pertussis- vaccinated population [61–63]. Novel approaches to blocking all whooping cough infection

10 sources should additionally be investigated if we are to effectively control this disease in the

future.

B. bronchiseptica disease and vaccination. B. bronchiseptica is a highly promiscuous

pathogen, infecting immunocompromised humans, but more commonly a wide range of

mammals, causing kennel cough in dogs and cats, snuffles in rabbits, and rhinitis in pigs [37,64–

67]. Although B. bronchiseptica is commonly associated with domestic animals, it has been

isolated from mice, rats, guinea pigs, skunks, opossums, raccoons, ferrets, foxes, hedgehogs,

sheep, koalas, leopards, horses, sea-lions, sea otters, marmosets, lesser bushbabies, turkeys and

primates, suggesting that this pathogen is capable of infecting a broad range of hosts [40,68–70].

Interestingly, B. bronchiseptica strains appear to transmit amongst different species, i.e. from humans to rabbits or from dogs to cats within the same home, raising interestingly questions concerning disease transmission and effective vaccine development [71]. Currently, both

injectable and intra-nasal vaccines are available to prevent B. bronchiseptica infections in dogs

in the United States [72]. Both vaccines induce high anti-B. bronchiseptica IgA titers; however,

they have not been shown to prevent intra-nasal colonization, a predicted likely source of transmission [73].

Agricultural swine production industries have been significantly affected by B. bronchiseptica, primarily because it can cause both pneumonia in piglets and rhinitis in adult swine [40,74]. B. bronchiseptica has additionally been shown to exacerbate several porcine respiratory diseases, including bacterial agents Pasteurella multocida, Streptococcus suis, and

Haemophilus parasuis and viral agents, such as porcine reproductive and respiratory syndrome virus (PRRSV) [75–79]. In swine, very little efficacy data on licensed vaccines exists [72], although recently, two of the licensed vaccines have been shown to generate high anti-B.

11 bronchiseptica titers in piglets, even though their efficacy in stopping lethal pneumonia and the ability of B. bronchiseptica to exacerbate other respiratory disease remains unknown [80].

Bordetellae virulence factors and their regulation.

The bordetellae all use common strategies to cause disease, most of which have been teased apart using a robust murine infection model. First, the bacteria must colonize the upper respiratory tract, competing with resident microflora. As the bacteria move into the trachea and lower respiratory tract, they adhere to epithelial cells and produce toxins that paralyze the cilia.

Numerous mechanisms are then used to cause inflammation, which interfere with pulmonary secretions; pro-inflammatory cytokines, such as interleukin-1β (IL-1β), tumor necrosis factor-α

(TNF-α), interferon-γ (IFN-γ), are all induced during bordetellosis, recruit immune cells to the site of infection, and can promote pathology in the lower respiratory tract. Lymphocytes are recruited to the site of infection as early as day 3 and then translocate to the surrounding lymphoid tissues, where an essential adaptive immune response is mounted. Antibody production, which is essential for clearance of the infection, begins around day 7, assisting both complement mediated killing and phagocytosis by numerous innate immune cells. B. pertussis and B. parapertussis are cleared and infection resolved in the murine model within 28 days, while B. bronchiseptica is cleared from the lower murine respiratory tract within 49 days but persists in the nasal cavity for the life of the animal [45,72].

Because so much of the classical bordetellae genomes are conserved, many of the virulence factors essential for diseases are also conserved. All three classical bordetellae have filamentous hemagglutinin, pertactin, and fimbriae, which have all been shown to promote adherence to respiratory tract epithelia [72,81–84]. Of the known bordetellae toxins, pertussis

12 toxin (Ptx) is arguably the most studied, but B. pertussis is the only bordetellae to express and utilize Ptx to block lymphocyte chemotaxis [85–88]. B. bronchiseptica and B. parapertussis contain the genes for Ptx, but genetic alterations have disrupted their expression [42,85].

Adenylate cyclase toxin (Act), which blocks phagocytosis and causes eukaryotic cell death by inducing supraphysiological levels of cyclic AMP from ATP, and dermonectrotic toxin (Dnt), which activates the GTP-binding protein Rho and causes cytoskeleton rearrangements and can block antibody production, are conserved amongst the three bordetellae species [72,89–93].

Virulence factors not conserved amongst the bordetellae include tracheal colonization factor

(Tcf), which causes ciliated-cell cytopathology and disrupts ciliostasis, O-antigen, an LPS associated molecule that blocks complement binding and contributes to immune evasion, and the

Type III Secretion System, shown to induce IL-1β, block interleukin-10 production, and cause eukaryotic cell death [94–100]. The ability of bordetellae to use different virulence factor repertoires may explain the differential diseases and some strain variation amongst the bordetellae.

The majority of known bordetellae virulence factors are regulated by a two component phosphorelay system, Bordetella virulence genes, Bvg. This master regulatory system is composed of a sensory protein, BvgS, and a transcriptional activator and regulator, BvgA, and is required for virulence in all three classical bordetellae [43,101–104]. Temperature, magnesium sulfate, and nicotinic acid have all been shown to modulate the BvgAS system, which alters virulence by initiating phase variations [105]. The Bvg- phase is thought to be a non-virulent state, where metabolism, transport, and motility genes are up-regulated and known virulence factors are down-regulated; the Bvg+ phase is just the opposite, where motility and starvation genes are down-regulated while toxin and adhesion genes are up-regulated [43,106–109]. In

13 between these two phases, an intermediate phase, Bvgi, exists, where adherence factors and metabolism genes are both transcribed while transcription of toxins and motility genes are not

[103,110–112]. Because the Bvg+ phase is sufficient for virulence and the human pathogens are

believed to transmit from human to human, why bordetellae maintain this phase variation

remains unclear. The Bvgi and Bvg- phases were thought to be required for environmental

survival outside of the host and has been implicated in transmission, but T. Nicholson recently

showed that Bvgi phase locked mutants can still transmit between piglets (T. Nicholson,

unpublished data), further making it unclear why Bordetetella maintain the Bvg- phase [43,113].

Additionally, how these respiratory pathogens fine tune transcriptional responses, respond to

environmental stresses, and use novel non-BvgAS regulated factors remains undetermined.

Numerous Gram-positive and Gram-negative pathogens are capable of fine-tuning their

transcription response with small RNA molecules (sRNAs) [114]. By definition, sRNAs include

ribosomal and transfer RNA molecules, but regulatory sRNAs, capable of post-transcriptionally

regulating numerous genes especially during stress responses, are also included in this group

[115,116]. In gut pathogens, such as Salmonella, Escherichia, Vibrio, Yersinia, and respiratory

pathogens, including Staphylococcus, Streptococcus, Pseudomonas and Mycobacterium, small

RNAs and their chaperone protein, Hfq, have been implicated in virulence, suggesting that

sRNA may contribute additional regulation on many known virulence factors [114,117–127].

Recently, 17 B. pertussis sRNAs were identified using bioinformatics and Northern blotting;

several of these sRNAs were additionally identified as being BvgAS regulated, suggesting that

bordetellae virulence factor regulation is much more complex than what is currently understood

[128]. Additionally, other Gram-negative pathogens, such as Vibrio cholerae, have over 400

sRNA molecules, suggesting that there are many more bordetellae sRNA molecules to be

14 identified. Furthermore, the messenger RNA targets of these sRNA molecules and sRNA regulatory systems beyond BvgAS remain undetermined, and until these are identified, additional mechanisms behind how bordetellae regulate and cause disease will remain unknown.

Ethical implications of biomedical research.

Investigating many of these biomedical issues involves not only top tier research, but a critical assessment of the ethical issues and implications accompanying the research.

Responsible Conduct of Research (RCR) training for any undergraduate student, graduate student, or post doctoral researcher conducting NSF sponsored research was mandated in 2007, and later for the NIH [130], when section 7009 of the America Creating Opportunities to

Meaningfully Promote Excellence in Technology, Education, and Science (COMPETES) Act was passed [129]. Although RCR training is not standardized [130,131], RCR training commonly utilizes case studies to teach students about the nine core domains: human and animal experimentation, mentoring, collaboration, peer review, data management and ownership, publication practices, authorship, conflicts of interest, and research misconduct [132–134].

Additionally, some universities have established online classes or lecture series to train their researchers, and although RCR training should be greatly applauded, there are major caveats to the approaches being used today. Eliciting opinions from scientists and ethicists on proper ethical training for any biomedical researcher should create a training environment where researchers are confident in their abilities to conduct ethical, sound research and apply that knowledge to the greater good.

Recently, The Pennsylvania State University took the initiative to combine a scientist’s major area of study with a bioethics dual degree, allowing scientific researchers the freedom to

15 investigate the ethical issues behind their research and determine the impacts their findings could

have beyond their scientific community. This unique program also allows scientists to contribute

to current ethical debates, such as those regarding autism, vaccination rights, end-of-life care, personalized medicine, gene patenting, research fraud, human research subjects, stem cell research, direct-to-market advertising, public health care, globalized biomedical research, and so on, adding a scientific viewpoint to a debate commonly besieged by lawyers and politicians.

Training scientists in bioethical methodologies increases creativity, trains more prepared scientists, and allows them unique career opportunities.

Preface

The overall focus of this dissertation was to better understand how respiratory pathogens interact with the respiratory tract microbiome and to investigate the virulence factors that pathogens use to compete with host flora. Additionally, this dissertation strives to investigate the ethical implications from a scientific perspective that are associated with this work.

Chapter 2 and 3 are the first descriptions of the Type VI Secretion System (T6SS) in

Bordetella. Although the secretion system is implicated in bacteria-to-bacteria competition in other systems, first, the locus had to be functionally described and the contributions to classical disease pathogenesis determined. The T6SS was found to induce pathology, cause immunomodulation, and contribute to persistence in both the upper and lower respiratory tracts.

Persistence was found to be correlated with decreased antibody titers. We determined that the

T6SS decreases antibody production by modulating lipid antigen presentation on antigen presentation cells. Chapter 2 is currently undergoing revisions at PLoS One, and Chapter 3 is under review at the Journal of Immunology.

16 In Chapter 4, we investigate how the promiscuous mammalian pathogen B.

bronchiseptica interacts with nasal cavity microflora. We observed that this respiratory

pathogen uses a variety of strategies to displace host microorganisms from the upper respiratory

tract as a mechanism for initial host colonization. B. bronchiseptica has several innate

mechanisms that are used to directly compete with host species, and this pathogen additionally

stimulates an immune response that is unfavorable to the host flora. A manuscript describing

this research is in preparation for the Proceedings of the National Academy of Science.

Similarly, in Chapter 5 we investigated how the human pathogen, B. pertussis, interacts

with microflora, or rather, how microflora interact with B. pertussis. We observed that host

organisms actually out-compete B. pertussis, and therefore block pathogen colonization at

natural levels. We then use these organisms to investigate using murine microflora isolates as a

novel disease prophylactic. This chapter also discusses how pathogens may evolve to compete

with flora in their niche host environments. This work is currently submitted to PLoS Biology.

Chapter 6 was an independent project coordinating RNA sequencing to identify small

RNA regulatory molecules that may target known virulence factors, including the T6SS and

T3SSs. Although over 180 small RNA molecules are identified, only 18 are predicted to target

known virulence factors, suggesting we have much to learn about how Bordetella species regulate virulence and growth independent of the master regulatory system BvgAS.

Both culture methods and metagenomic sequencing are utilized in Chapter 7 to detect microorganisms present in healthy humans, as well as Chronic Obstructive Pulmonary Disease

(COPD) and asthma patients. We detect several organisms only present in diseased verses

healthy humans and hypothesize that the dysbiosis of the microbiome in diseased patients may

contribute to their symptoms. This chapter also allows us to compare human and murine

17 microflora, as well as lower and upper respiratory tract organisms, and is part of a much larger

study, correlating immune responses, microflora, and disease symptoms in patients.

Lastly, Chapters 8 and 9 are focused not on the science of this dissertation, but the ethical

implications of this research. With the recent implementation of Responsible Conduct of

Research (RCR) training by both the National Science Foundation and National Institute of

Health, finding the most effective, rigorous training methodologies has become a timely topic of

discussion. In Chapter 8, I argue that teaching foundational ethical theories to graduate student

researchers is the most effective way to not only teach RCR training, but also to train better scientists, capable of communicating their science to society and producing more applicable findings. This chapter is currently submitted to the Science Policy section of PLoS One.

In Chapter 9, I use various case studies to investigate the best possible solution to

patenting human isolated microbes. As we understand more about the human microbiome and

how it contributes to disease, we will also begin to understand how microbes may be

administered to the body to achieve a healthier microbiota. Commercialization and potentially

exploitation of these microbes will almost certainly come and determining the best legal

regulation of these measures should be discusses earlier, rather than later, by not only lawyers

and politicians, but also ethicists, doctors, and research scientists with a foundational

understanding of ethics.

18

References

1. Jones FS (1922) THE SOURCE OF THE MICROORGANISMS IN THE LUNGS OF NORMAL ANIMALS. J Exp Med 36: 317-328.

2. Efremova SA, Kurina SA, Grig MG, Sazpnova IM (1977) [Microflora of the nasal cavity and its sensitivity to antibiotics in patients with chronic rhinitis treated by cryotherapy]. Antibiotiki 22: 324-327.

3. Fagon J-Y, Chastre J, Trouillet J-L, Domart Y, Dombret M-C, et al. (1990) Characterization of Distal Bronchial Microflora during Acute Exacerbation of Chronic Bronchitis: Use of the Protected Specimen Brush Technique in 54 Mechanically Ventilated Patients. American Journal of Respiratory and Critical Care Medicine 142: 1004 -1008. doi:10.1164/ajrccm/142.5.1004

4. Tlaskalová-Hogenová H, Štěpánková R, Hudcovic T, Tučková L, Cukrowska B, et al. (2004) Commensal bacteria (normal microflora), mucosal immunity and chronic inflammatory and autoimmune diseases. Immunology Letters 93: 97-108. doi:10.1016/j.imlet.2004.02.005

5. Könönen E, Jousimies-Somer H, Bryk A, Kilp T, Kilian M (2002) Establishment of streptococci in the upper respiratory tract: longitudinal changes in the mouth and nasopharynx up to 2 years of age. J. Med. Microbiol. 51: 723-730.

6. Efremova SA, Kurina SA, Grig MG, Sazpnova IM (1977) [Microflora of the nasal cavity and its sensitivity to antibiotics in patients with chronic rhinitis treated by cryotherapy]. Antibiotiki 22: 324-327.

7. Abdulsalam A. Bakhsh Alhendi (1999) Nasal Microflora of Camels (Camelus dromedarius) Under Two Different Conditions. Pakistan Vet. J. 19.

8. Ng TFF, Wheeler E, Greig D, Waltzek TB, Gulland F, et al. (2011) Metagenomic identification of a novel anellovirus in Pacific harbor seal (Phoca vitulina richardsii) lung samples and its detection in samples from multiple years. J. Gen. Virol 92: 1318-1323. doi:10.1099/vir.0.029678-0

9. Desissa F. Gutema, Bojia E. Duguma, Ayele G. Dinka (2009) Isolation and Identification of Aerobic Bacterial Flora from the Upper Respiratory Tract of Donkeys in Central Ethiopia. Intern J Appl Res Vet Med 7: 181-189.

10. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, et al. (2007) The Human Microbiome Project. Nature 449: 804-810. doi:10.1038/nature06244

11. WHO | The world health report 2004 - changing history (2004). Available: http://www.who.int/whr/2004/en/. Accessed 12 Jan 2012.

19 12. Adams SG, Melo J, Luther M, Anzueto A (2000) Antibiotics Are Associated With Lower Relapse Rates in Outpatients With Acute Exacerbations of COPD*. Chest 117: 1345 -1352. doi:10.1378/chest.117.5.1345

13. Wilson R (2001) Bacteria, antibiotics and COPD. European Respiratory Journal 17: 995 - 1007.

14. Ko FWS, Ip M, Chan PKS, Ng SSS, Chau SS, et al. (2008) A one-year prospective study of infectious etiology in patients hospitalized with acute exacerbations of COPD and concomitant pneumonia. Respiratory Medicine 102: 1109-1116. doi:10.1016/j.rmed.2008.03.019

15. Lode H, Allewelt M, Balk S, Roux A, Mauch H, et al. (2007) A Prediction Model for Bacterial Etiology in Acute Exacerbations of COPD. Infection 35: 143-149. doi:10.1007/s15010-007-6078-z

16. De Serres G, Lampron N, La Forge J, Rouleau I, Bourbeau J, et al. (2009) Importance of viral and bacterial infections in chronic obstructive pulmonary disease exacerbations. Journal of Clinical Virology 46: 129-133. doi:10.1016/j.jcv.2009.07.010

17. Blaser MJ, Falkow S (2009) What are the consequences of the disappearing human microbiota? Nat Rev Micro 7: 887-894. doi:10.1038/nrmicro2245

18. Murphy TF (2008) The Many Faces of Pseudomonas aeruginosa in Chronic Obstructive Pulmonary Disease. Clinical Infectious Diseases 47: 1534 -1536. doi:10.1086/593187

19. Garcia-Vidal C, Almagro P, Romaní V, Rodríguez-Carballeira M, Cuchi E, et al. (2009) Pseudomonas aeruginosa in patients hospitalised for COPD exacerbation: a prospective study. European Respiratory Journal 34: 1072 -1078. doi:10.1183/09031936.00003309

20. Green RM, Gally F, Keeney JG, Alper S, Gao B, et al. (2009) Impact of Cigarette Smoke Exposure on Innate Immunity: A Caenorhabditis elegans Model. PLoS ONE 4: e6860. doi:10.1371/journal.pone.0006860

21. Read AF, Taylor LH (2001) The Ecology of Genetically Diverse Infections. Science 292: 1099 -1102. doi:10.1126/science.1059410

22. Chao L, Levin BR (1981) Structured habitats and the evolution of anticompetitor toxins in bacteria. Proceedings of the National Academy of Sciences 78: 6324 -6328.

23. Brown SP, Inglis RF, Taddei F (2009) Evolutionary ecology of microbial wars: within-host competition and (incidental) virulence. Evolutionary Applications 2: 32-39. doi:10.1111/j.1752-4571.2008.00059.x

24. Herold S, Siebert J, Huber A, Schmidt H (2005) Global Expression of Prophage Genes in Escherichia coli O157:H7 Strain EDL933 in Response to Norfloxacin. Antimicrob Agents Chemother 49: 931-944. doi:10.1128/AAC.49.3.931-944.2005

20 25. Bossi L, Fuentes JA, Mora G, Figueroa-Bossi N (2003) Prophage Contribution to Bacterial Population Dynamics. J Bacteriol 185: 6467-6471. doi:10.1128/JB.185.21.6467-6471.2003

26. Wang X, Kim Y, Ma Q, Hong SH, Pokusaeva K, et al. (2010) Cryptic prophages help bacteria cope with adverse environments. Nat Commun 1: 147. doi:10.1038/ncomms1146

27. Irie Y, O’Toole GA, Yuk MH (2005) Pseudomonas aeruginosa rhamnolipids disperse Bordetella bronchiseptica biofilms. FEMS Microbiology Letters 250: 237-243. doi:10.1016/j.femsle.2005.07.012

28. Hood RD, Singh P, Hsu F, Güvener T, Carl MA, et al. (n.d.) A Type VI Secretion System of Pseudomonas aeruginosa Targets a Toxin to Bacteria. Cell Host & Microbe 7: 25-37.

29. MacIntyre DL, Miyata ST, Kitaoka M, Pukatzki S (2010) The Vibrio cholerae type VI secretion system displays antimicrobial properties. Proceedings of the National Academy of Sciences 107: 19520 -19524. doi:10.1073/pnas.1012931107

30. Murdoch SL, Trunk K, English G, Fritsch MJ, Pourkarimi E, et al. (2011) The Opportunistic Pathogen Serratia marcescens Utilizes Type VI Secretion To Target Bacterial Competitors. J. Bacteriol. 193: 6057-6069. doi:10.1128/JB.05671-11

31. Hibbing ME, Fuqua C, Parsek MR, Peterson SB (2010) Bacterial competition: surviving and thriving in the microbial jungle. Nat Rev Micro 8: 15-25. doi:10.1038/nrmicro2259

32. Smith VH, Holt RD (1996) Resource competition and within-host disease dynamics. Trends in Ecology & Evolution 11: 386-389. doi:10.1016/0169-5347(96)20067-9

33. Griffin AS, West SA, Buckling A (2004) Cooperation and competition in pathogenic bacteria. Nature 430: 1024-1027. doi:10.1038/nature02744

34. Flint HJ, Duncan SH, Scott KP, Louis P (2007) Interactions and competition within the microbial community of the human colon: links between diet and health. Environmental Microbiology 9: 1101-1111. doi:10.1111/j.1462-2920.2007.01281.x

35. Stecher B, Robbiani R, Walker AW, Westendorf AM, Barthel M, et al. (2007) Salmonella enterica Serovar Typhimurium Exploits Inflammation to Compete with the Intestinal Microbiota. PLoS Biol 5: e244. doi:10.1371/journal.pbio.0050244

36. Lysenko ES, Ratner AJ, Nelson AL, Weiser JN (2005) The Role of Innate Immune Responses in the Outcome of Interspecies Competition for Colonization of Mucosal Surfaces. PLoS Pathog 1: e1. doi:10.1371/journal.ppat.0010001

37. Mattoo S, Cherry JD (2005) Molecular Pathogenesis, Epidemiology, and Clinical Manifestations of Respiratory Infections Due to Bordetella pertussis and Other Bordetella Subspecies. Clin. Microbiol. Rev. 18: 326-382. doi:10.1128/CMR.18.2.326-382.2005

21 38. Ko KS, Peck KR, Oh WS, Lee NY, Lee JH, et al. (2005) New Species of Bordetella, Bordetella ansorpii sp. nov., Isolated from the Purulent Exudate of an Epidermal Cyst. J Clin Microbiol 43: 2516-2519. doi:10.1128/JCM.43.5.2516-2519.2005

39. Mattoo S, Foreman-Wykert AK, Cotter PA, Miller JF (2001) Mechanisms of Bordetella pathogenesis. Front Biosci 6: E168-86.

40. Goodnow RA (1980) Biology of Bordetella bronchiseptica. Microbiol Rev 44: 722-38.

41. Diavatopoulos DA, Cummings CA, Schouls LM, Brinig MM, Relman DA, et al. (2005) Bordetella pertussis, the causative agent of whooping cough, evolved from a distinct, human-associated lineage of B. bronchiseptica. PLoS Pathog 1: e45.

42. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, et al. (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35: 32-40. doi:10.1038/ng1227 ng1227 [pii]

43. Cotter PA, Miller JF (1994) BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. Infect Immun 62: 3381- 90.

44. McMillan DJ, Shojaei M, Chhatwal GS, Guzman CA, Walker MJ (1996) Molecular analysis of the bvg-repressed urease of Bordetella bronchiseptica. Microb Pathog 21: 379- 94. doi:S0882-4010(96)90069-8 [pii] 10.1006/mpat.1996.0069

45. Center for Disease Control and Prevention (2008) Pertussis.. Available: http://www.cdc.gov/vaccines/pubs/pinkbook/downloads/pert.pdf. Accessed 12 Jan 2012.

46. Schmitt-Grohe S, Cherry JD, Heininger U, Uberall MA, Pineda E, et al. (1995) Pertussis in German adults. Clin Infect Dis 21: 860-6.

47. Heininger U, Cherry JD (2006) Pertussis immunisation in adolescents and adults-- Bordetella pertussis epidemiology should guide vaccination recommendations. Expert Opin Biol Ther 6: 685-97.

48. CDC Features - Pertussis (Whooping Cough) – What You Need To Know (n.d.). Available: http://www.cdc.gov/features/Pertussis/. Accessed 19 Jan 2012.

49. Heininger U (2001) Pertussis: an old disease that is still with us. Curr Opin Infect Dis 14: 329-35.

50. Celentano LP, Massari M, Paramatti D, Salmaso S, Tozzi AE (2005) Resurgence of pertussis in Europe. Pediatr Infect Dis J 24: 761-5. doi:00006454-200509000-00003 [pii]

51. Pam D (2002) Whooping cough makes global comeback. The Lancet Infectious Diseases 2: 322. doi:10.1016/S1473-3099(02)00308-0

22 52. Lacey BW (1960) Antigenic modulation of Bordetella pertussis. J Hyg (Lond) 58: 57-93.

53. Heikkinen E, Kallonen T, Saarinen L, Sara R, King AJ, et al. (2007) Comparative genomics of Bordetella pertussis reveals progressive gene loss in Finnish strains. PLoS ONE 2: e904. doi:10.1371/journal.pone.0000904

54. van der Zee A, Vernooij S, Peeters M, van Embden J, Mooi FR (1996) Dynamics of the population structure of Bordetella pertussis as measured by IS1002-associated RFLP: comparison of pre- and post-vaccination strains and global distribution. Microbiology 142 ( Pt 12): 3479-85.

55. Tan T, Trindade E, Skowronski D (2005) Epidemiology of Pertussis. The Pediatric Infectious Disease Journal 24. Available: http://journals.lww.com/pidj/Fulltext/2005/05001/Epidemiology_of_Pertussis.3.aspx.

56. Mielcarek N, Debrie A-S, Raze D, Bertout J, Rouanet C, et al. (2006) Live Attenuated B. pertussis as a Single-Dose Nasal Vaccine against Whooping Cough. PLoS Pathog 2: e65. doi:10.1371/journal.ppat.0020065

57. Mastrantonio P, Stefanelli P, Giuliano M, Herrera Rojas Y, Ciofi degli Atti M, et al. (1998) Bordetella parapertussis infection in children: epidemiology, clinical symptoms, and molecular characteristics of isolates. J Clin Microbiol 36: 999-1002.

58. Heininger U, Stehr K, Schmitt-Grohe S, Lorenz C, Rost R, et al. (1994) Clinical characteristics of illness caused by Bordetella parapertussis compared with illness caused by Bordetella pertussis. Pediatr Infect Dis J 13: 306-9.

59. Wolfe DN, Goebel EM, Bjornstad ON, Restif O, Harvill ET (2007) The O antigen enables Bordetella parapertussis to avoid Bordetella pertussis-induced immunity. Infect Immun 75: 4972-9. doi:IAI.00763-07 [pii] 10.1128/IAI.00763-07

60. Long GH, Karanikas AT, Harvill ET, Read AF, Hudson PJ (2010) Acellular pertussis vaccination facilitates Bordetella parapertussis infection in a rodent model of bordetellosis. Proceedings of the Royal Society B: Biological Sciences. Available: http://rspb.royalsocietypublishing.org/content/early/2010/02/26/rspb.2010.0010.abstract. Accessed 22 Jan 2012.

61. Njamkepo E, Delisle F, Hagege I, Gerbaud G, Guiso N (2000) Bordetella holmesii isolated from a patient with sickle cell anemia: analysis and comparison with other Bordetella holmesii isolates. Clin Microbiol Infect 6: 131-6.

62. Mazengia E, Silva EA, Peppe JA, Timperi R, George H (2000) Recovery of Bordetella holmesii from Patients with Pertussis-Like Symptoms: Use of Pulsed-Field Gel Electrophoresis To Characterize Circulating Strains. J Clin Microbiol 38: 2330-2333.

63. Mooi FR, Bruisten S, Linde I, Reubsaet F, Heuvelman K, et al. (n.d.) Characterization of Bordetella holmesii isolates from patients with pertussis‐like illness in the Netherlands.

23 FEMS Immunology & Medical Microbiology. Available: http://onlinelibrary.wiley.com/doi/10.1111/j.1574-695X.2011.00911.x/abstract. Accessed 22 Jan 2012.

64. Bemis DA, Carmichael LE, Appel MJ (1977) Naturally occurring respiratory disease in a kennel caused by Bordetella bronchiseptica. Cornell Vet 67: 282-293.

65. Rougier S, Galland D, Boucher S, Boussarie D, Valle M (2006) Epidemiology and susceptibility of pathogenic bacteria responsible for upper respiratory tract infections in pet rabbits. Veterinary Microbiology 115: 192.

66. Kobisch M, Pennings A (1989) An evaluation in pigs of Nobi-Vac AR and an experimental atrophic rhinitis vaccine containing P multocida DNT-toxoid and B bronchiseptica. Vet Rec 124: 57-61.

67. Dworkin MS, Sullivan PS, Buskin SE, Harrington RD, Olliffe J, et al. (1999) Bordetella bronchiseptica infection in human immunodeficiency virus-infected patients. Clin Infect Dis 28: 1095-9.

68. Baskerville M, Wood M, Baskerville A (1983) An outbreak of Bordetella bronchiseptica pneumonia in a colony of common marmosets (Callithrix jacchus). Lab. Anim. 17: 350- 355.

69. Brownstein D, Miller MA, Oates SC, Byrne BA, Jang S, et al. (2011) Antimicrobial susceptibility of bacterial isolates from sea otters (Enhydra lutris). J. Wildl. Dis. 47: 278- 292.

70. Good RC, May BD (1971) Respiratory Pathogens in Monkeys. Infection and Immunity 3: 87 -93.

71. Binns SH, Speakman AJ, Dawson S, Bennett M, Gaskell RM, et al. (1998) The use of pulsed-field gel electrophoresis to examine the epidemiology of Bordetella bronchiseptica isolated from cats and other species. Epidemiol Infect 120: 201-208.

72. Mattoo S, Cherry JD (2005) Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev 18: 326-82.

73. Ellis JA, Krakowka GS, Dayton AD, Konoby C (2002) Comparative efficacy of an injectable vaccine and an intranasal vaccine in stimulating Bordetella bronchiseptica- reactive antibody responses in seropositive dogs. J. Am. Vet. Med. Assoc. 220: 43-48.

74. Nicholson TL, Brockmeier SL, Loving CL (2009) Contribution of Bordetella bronchiseptica Filamentous Hemagglutinin and Pertactin to Respiratory Disease in Swine. Infection and Immunity 77: 2136 -2146. doi:10.1128/IAI.01379-08

24 75. Chanter N, Magyar T, Rutter JM (1989) Interactions between Bordetella bronchiseptica and toxigenic Pasteurella multocida in atrophic rhinitis of pigs. Res. Vet. Sci. 47: 48-53.

76. Vecht U, Wisselink HJ, van Dijk JE, Smith HE (1992) Virulence of Streptococcus suis type 2 strains in newborn germfree pigs depends on phenotype. Infect Immun 60: 550-556.

77. Brockmeier SL, Loving CL, Nicholson TL, Palmer MV (2008) Coinfection of pigs with porcine respiratory coronavirus and Bordetella bronchiseptica. Veterinary Microbiology 128: 36-47. doi:10.1016/j.vetmic.2007.09.025

78. Brockmeier SL, Palmer MV, Bolin SR, Rimler RB (2001) Effects of intranasal inoculation with Bordetella bronchiseptica, porcine reproductive and respiratory syndrome virus, or a combination of both organisms on subsequent infection with Pasteurella multocida in pigs. Am. J. Vet. Res. 62: 521-525.

79. Brockmeier SL (2004) Prior infection with Bordetella bronchiseptica increases nasal colonization by Haemophilus parasuis in swine. Veterinary Microbiology 99: 75-78.

80. Novotny P, Kobisch M, Cownley K, Chubb AP, Montaraz JA (1985) Evaluation of Bordetella bronchiseptica vaccines in specific-pathogen-free piglets with bacterial cell surface antigens in enzyme-linked immunosorbent assay. Infect Immun 50: 190-8.

81. Edwards JA, Groathouse NA, Boitano S (2005) Bordetella bronchiseptica Adherence to Cilia Is Mediated by Multiple Adhesin Factors and Blocked by Surfactant Protein A. Infect Immun 73: 3618-3626. doi:10.1128/IAI.73.6.3618-3626.2005

82. Tuomanen E, Weiss A (1985) Characterization of two adhesins of Bordetella pertussis for human ciliated respiratory-epithelial cells. J Infect Dis 152: 118-25.

83. Tuomanen E, Weiss A, Rich R, Zak F, Zak O (1985) Filamentous hemagglutinin and pertussis toxin promote adherence of Bordetella pertussis to cilia. Dev Biol Stand 61: 197- 204.

84. Cotter PA, Yuk MH, Mattoo S, Akerley BJ, Boschwitz J, et al. (1998) Filamentous hemagglutinin of Bordetella bronchiseptica is required for efficient establishment of tracheal colonization. Infect Immun 66: 5921-9.

85. Arico B, Rappuoli R (1987) Bordetella parapertussis and Bordetella bronchiseptica contain transcriptionally silent pertussis toxin genes. J Bacteriol 169: 2847-53.

86. Gray LS, Huber KS, Gray MC, Hewlett EL, Engelhard VH (1989) Pertussis toxin effects on T lymphocytes are mediated through CD3 and not by pertussis toxin catalyzed modification of a G protein. J Immunol 142: 1631-8.

87. Thomazzi SM, Souza MH, Melo-Filho AA, Hewlett EL, Lima AA, et al. (1995) Pertussis toxin from Bordetella pertussis blocks neutrophil migration and neutrophil-dependent edema in response to inflammation. Braz J Med Biol Res 28: 120-4.

25 88. Kirimanjeswara GS, Agosto LM, Kennett MJ, Bjornstad ON, Harvill ET (2005) Pertussis toxin inhibits neutrophil recruitment to delay antibody-mediated clearance of Bordetella pertussis. J Clin Invest 115: 3594-601.

89. Goldman WE, Klapper DG, Baseman JB (1982) Detection, isolation, and analysis of a released Bordetella pertussis product toxic to cultured tracheal cells. Infect Immun 36: 782- 794.

90. Khelef N, Zychlinsky A, Guiso N (1993) Bordetella pertussis induces apoptosis in macrophages: role of adenylate cyclase-hemolysin. Infect Immun 61: 4064-71.

91. Confer D, Eaton J (1982) Phagocyte impotence caused by an invasive bacterial adenylate cyclase. Science 217: 948 -950. doi:10.1126/science.6287574

92. Horiguchi Y, Matsuda H, Koyama H, Nakai T, Kume K (1992) Bordetella bronchiseptica dermonecrotizing toxin suppresses in vivo antibody responses in mice. FEMS Microbiol. Lett. 69: 229-234.

93. Brockmeier SL, Register KB, Magyar T, Lax AJ, Pullinger GD, et al. (2002) Role of the Dermonecrotic Toxin of Bordetella bronchiseptica in the Pathogenesis of Respiratory Disease in Swine. Infect. Immun. 70: 481-490. doi:10.1128/iai.70.2.481-490.2002

94. Anderton TL, Maskell DJ, Preston A (2004) Ciliostasis is a key early event during colonization of canine tracheal tissue by Bordetella bronchiseptica. Microbiology 150: 2843-55.

95. Finn TM, Stevens LA (1995) Tracheal colonization factor: a Bordetella pertussis secreted virulence determinant. Molecular Microbiology 16: 625-634. doi:10.1111/j.1365- 2958.1995.tb02425.x

96. Goebel EM, Wolfe DN, Elder K, Stibitz S, Harvill ET (2008) O antigen protects Bordetella parapertussis from complement. Infect Immun 76: 1774-80. doi:IAI.01629-07 [pii] 10.1128/IAI.01629-07

97. Burns VC, Pishko EJ, Preston A, Maskell DJ, Harvill ET (2003) Role of Bordetella O antigen in respiratory tract infection. Infect Immun 71: 86-94.

98. Panina EM, Mattoo S, Griffith N, Kozak NA, Yuk MH, et al. (2005) A genome-wide screen identifies a Bordetella type III secretion effector and candidate effectors in other species. Mol Microbiol 58: 267-79.

99. Stockbauer KE, Foreman-Wykert AK, Miller JF (2003) Bordetella type III secretion induces caspase 1-independent necrosis. Cell. Microbiol. 5: 123-132.

100. Yuk MH, Harvill ET, Cotter PA, Miller JF (2000) Modulation of host immune responses, induction of apoptosis and inhibition of NF-kappaB activation by the Bordetella type III secretion system. Mol Microbiol 35: 991-1004.

26 101. Carbonetti NH, Khelef N, Guiso N, Gross R (1993) A phase variant of Bordetella pertussis with a mutation in a new locus involved in the regulation of pertussis toxin and adenylate cyclase toxin expression. J Bacteriol 175: 6679-88.

102. Cotter PA, Miller JF (1997) A mutation in the Bordetella bronchiseptica bvgS gene results in reduced virulence and increased resistance to starvation, and identifies a new class of Bvg-regulated antigens. Mol Microbiol 24: 671-85.

103. Williams CL, Boucher PE, Stibitz S, Cotter PA (2005) BvgA functions as both an activator and a repressor to control Bvg phase expression of bipA in Bordetella pertussis. Mol Microbiol 56: 175-88.

104. Uhl MA, Miller JF (1994) Autophosphorylation and phosphotransfer in the Bordetella pertussis BvgAS signal transduction cascade. Proc Natl Acad Sci U S A 91: 1163-1167.

105. Melton AR, Weiss AA (1993) Characterization of environmental regulators of Bordetella pertussis. Infect Immun 61: 807-15.

106. Akerley BJ, Miller JF (1993) Flagellin gene transcription in Bordetella bronchiseptica is regulated by the BvgAS virulence control system. J Bacteriol 175: 3468-79.

107. Nakamura MM, Liew SY, Cummings CA, Brinig MM, Dieterich C, et al. (2006) Growth phase- and nutrient limitation-associated transcript abundance regulation in Bordetella pertussis. Infect Immun 74: 5537-48.

108. Stibitz S (1994) Mutations in the bvgA gene of Bordetella pertussis that differentially affect regulation of virulence determinants. J Bacteriol 176: 5615-21.

109. Akerley BJ, Monack DM, Falkow S, Miller JF (1992) The bvgAS locus negatively controls motility and synthesis of flagella in Bordetella bronchiseptica. J Bacteriol 174: 980-90.

110. Sukumar N, Mishra M, Sloan GP, Ogi T, Deora R (2007) Differential Bvg Phase- Dependent Regulation and Combinatorial Role in Pathogenesis of Two Bordetella Paralogs, BipA and BcfA. J. Bacteriol. 189: 3695-3704. doi:10.1128/jb.00009-07

111. Vergara-Irigaray N, Chavarri-Martinez A, Rodriguez-Cuesta J, Miller JF, Cotter PA, et al. (2005) Evaluation of the role of the Bvg intermediate phase in Bordetella pertussis during experimental respiratory infection. Infect Immun 73: 748-60.

112. Irie Y, Yuk MH (2007) In vivo colonization profile study of Bordetella bronchiseptica in the nasal cavity. FEMS Microbiol. Lett 275: 191-198. doi:10.1111/j.1574- 6968.2007.00852.x

113. Cotter PA, Miller JF (1997) A mutation in the Bordetella bronchiseptica bvgS gene results in reduced virulence and increased resistance to starvation, and identifies a new class of Bvg-regulated antigens. Mol Microbiol 24: 671-85.

27 114. Mraheil MA, Billion A, Kuenne C, Pischimarov J, Kreikemeyer B, et al. (2010) Comparative genome-wide analysis of small RNAs of major Gram-positive pathogens: from identification to application. Microbial Biotechnology 3: 658-676. doi:10.1111/j.1751- 7915.2010.00171.x

115. Gottesman S (2005) Micros for microbes: non-coding regulatory RNAs in bacteria. Trends Genet. 21: 399-404. doi:10.1016/j.tig.2005.05.008

116. Majdalani N, Vanderpool CK, Gottesman S (2005) Bacterial small RNA regulators. Crit. Rev. Biochem. Mol. Biol. 40: 93-113. doi:10.1080/10409230590918702

117. Perez N, Treviño J, Liu Z, Ho SCM, Babitzke P, et al. (2009) A Genome-Wide Analysis of Small Regulatory RNAs in the Human Pathogen Group A Streptococcus. PLoS ONE 4: e7668. doi:10.1371/journal.pone.0007668

118. Song T, Mika F, Lindmark B, Liu Z, Schild S, et al. (2008) A new Vibrio cholerae sRNA modulates colonization and affects release of outer membrane vesicles. Mol Microbiol 70: 100-111. doi:10.1111/j.1365-2958.2008.06392.x

119. Vogel J (2009) A rough guide to the non-coding RNA world of Salmonella. Molecular Microbiology 71: 1-11. doi:10.1111/j.1365-2958.2008.06505.x

120. Chabelskaya S, Gaillot O, Felden B (2010) A Staphylococcus aureus Small RNA Is Required for Bacterial Virulence and Regulates the Expression of an Immune-Evasion Molecule. PLoS Pathog 6: e1000927. doi:10.1371/journal.ppat.1000927

121. Pichon C (2005) From The Cover: Small RNA genes expressed from Staphylococcus aureusgenomic and pathogenicity islands with specific expression among pathogenic strains. Proceedings of the National Academy of Sciences 102: 14249-14254. doi:10.1073/pnas.0503838102

122. Livny J (2006) Identification of 17 Pseudomonas aeruginosa sRNAs and prediction of sRNA-encoding genes in 10 diverse pathogens using the bioinformatic tool sRNAPredict2. Nucleic Acids Research 34: 3484-3493. doi:10.1093/nar/gkl453

123. Arnvig KB, Young DB (2009) Identification of small RNAs in Mycobacterium tuberculosis. Molecular Microbiology 73: 397-408. doi:10.1111/j.1365-2958.2009.06777.x

124. Alejandro Toledo-Arana, Francis Repoila, Pascale Cossart (2007) Small noncoding RNAs controlling pathogenesis. Curr Opin Micro 10: 182-188.

125. Padalon-Brauch G, Hershberg R, Elgrably-Weiss M, Baruch K, Rosenshine I, et al. (2008) Small RNAs encoded within genetic islands of Salmonella typhimurium show host-induced expression and role in virulence. Nucleic Acids Research 36: 1913 -1927. doi:10.1093/nar/gkn050

28 126. Chen Y, Indurthi DC, Jones SW, Papoutsakis ET (January) Small RNAs in the Genus Clostridium. mBio 2. Available: http://mbio.asm.org/content/2/1/e00340-10.abstract. Accessed 7 Mar 2011.

127. Schiano CA, Bellows LE, Lathem WW (2010) The Small RNA Chaperone Hfq Is Required for the Virulence of Yersinia pseudotuberculosis. Infection and Immunity 78: 2034-2044. doi:10.1128/IAI.01046-09

128. Hot D, Slupek S, Wulbrecht B, D’Hondt A, Hubans C, et al. (2011) Detection of small RNAs in Bordetella pertussis and identification of a novel repeated genetic element. BMC Genomics 12: 207. doi:10.1186/1471-2164-12-207

129. 110th Congress (2007) H.R.2272 -- America Creating Opportunities to Meaningfully Promote Excellence in Technology, Education, and Science Act.. Available: http://thomas.loc.gov/cgi-bin/query/D?c110:5:./temp/~c110PswqSC::#. Accessed 21 Oct 2011.

130. DuBois JM, Schilling DA, Heitman E, Steneck NH, Kon AA (2010) Instruction in the responsible conduct of research: an inventory of programs and materials within CTSAs. Clin Transl Sci 3: 109-111. doi:10.1111/j.1752-8062.2010.00193.x

131. Kon AA, Schilling DA, Heitman E, Steneck NH, DuBois JM (2011) Content Analysis of Major Textbooks and Online Resources Used in Responsible Conduct of Research Instruction. AJOB Prim Res 2: 42-46. doi:10.1080/21507716.2011.564263

132. Horner J, Minifie FD (2011) Research ethics II: Mentoring, collaboration, peer review, and data management and ownership. J. Speech Lang. Hear. Res. 54: S330-345. doi:10.1044/1092-4388(2010/09-0264)

133. Horner J, Minifie FD (2011) Research ethics I: Responsible conduct of research (RCR)-- historical and contemporary issues pertaining to human and animal experimentation. J. Speech Lang. Hear. Res. 54: S303-329. doi:10.1044/1092-4388(2010/09-0265)

134. Horner J, Minifie FD (2011) Research Ethics III: Publication Practices and Authorship, Conflicts of Interest, and Research Misconduct. Journal of Speech, Language & Hearing Research 54: S346-S362. doi:10.1044/1092-4388(2010/09-0263)

29

Chapter 2 : A Type VI Secretion System Encoding Locus is Required for Bordetella

bronchiseptica Immunomodulation and Persistence In Vivo

30

Abstract

Type Six Secretion Systems (T6SS) have been identified in numerous Gram-negative pathogens, but a lack of natural host infection models have limited understanding its contributions to infection and pathogenesis. A 26 gene locus in the model pathogen, Bordetella bronchiseptica, contains apparent orthologs to all known T6SS core genes, as well as thirteen novel genes. Generating an in frame deletion in an ortholog of clpV, which encodes an ATPase required for protein secretion in other systems, resulted in cytotoxicity defects and increased expression of several proteins associated with apoptosis in exposed macrophages. ClpV is also required for of IL-1β, IL-6, IL-17, and IL-10 induction in infected macrophages. In wild type mice, ClpV disruption contributes to altered cytokine induction, decreased pathology, and decreased persistence in the nasal cavity and lungs. Together, these results reveal a natural host infection system in which the disruption of the clpV ortholog within an apparent T6SS locus reveals essential roles in immunomodulation and pathogenesis.

31

Introduction

Highly conserved Type VI Secretion System (T6SS) gene clusters have been recently identified in 92 different strains of bacteria [1]. T6SS loci are disproportionately associated with virulent strains, and multiple virulence-related phenotypes have been attributed to the T6SS in pathogenic bacteria, including mucosal adherence, intracellular growth within macrophages, survival within host cells, and the delivery of bacteriolytic proteins into competitor bacteria [1–

5]. In Vibrio cholerae [6], Aeromonas hydrophila [7], and Legionella pneumophila [8], T6SS activity enables macrophage cytotoxicity, while T6SSs of Salmonella typhimurium and Yersinia pseudotuberculosis facilitate HEp-2 cell invasion [9]. Abrogating T6SS functions is associated with reduced Aeromonas hydrophila virulence in a mouse model of septicemia, Pseudomonas aeruginosa in neutropenic mice [10], and Burkholderia mallei in hamsters [11]. Strikingly, disruption of the T6SS in Entero-Aggregative Escherichia coli (EAEC) does not cause an observable loss of function in a wild type murine infection model [12]. With the exception of A. hydrophila, Salmonella enterica, and Francisella tularensis, many T6SS in vitro phenotypes could not be recapitulated in vivo using wild type mice [7,13,14]. Despite evidence that the

T6SS enables virulence in multiple species, many of the discrete, in vivo interactions between the

T6SS and host immunity have not yet been determined.

This study examines the T6SS in the common respiratory pathogen, Bordetella bronchiseptica. This Gram-negative bacterium infects a wide range of mammals, including humans, and causes disease severities ranging from asymptomatic carriage to fatal pneumonia.

B. bronchiseptica commonly causes kennel cough in domesticated animals, snuffles in rabbits, and atrophic rhinitis in swine, and is considered the evolutionary progenitor-like strain of B.

32 pertussis and B. parapertussis, causative agents of whooping cough in humans [15]. B.

bronchiseptica also efficiently infects and causes disease in laboratory animals, such as mice,

rats, and rabbits, providing a natural host infection model that has been used to reveal important

interactions between bacterial virulence factors and the host immune system in vivo [16,17].

A considerable number of specific bordetellae virulence determinants, such as

autotransporters, adhesins, and toxins, require secretion through various machineries, such as the

Type I, Type II, Type III, Type IV, and Type V secretion systems (TnSS) [15,18]. These

secretion systems export factors that enable host epithelium adherence [19], disable the

mucociliary escalator [20], manipulate signaling pathways in antigen presenting cells [21,22],

and block neutrophil chemokine receptors [23]. Bordetella virulence factors, such as adenylate

cyclase toxin (ACT), pertussis toxin (PTX), fimbria, Bordetella resistance to killing protein

(BrkA), filamentous hemagglutinin (FHA), pertactin (PRN), and tracheal colonization factor

(TCF), have all been shown to require secretion systems for export [18,24–26]. Even when

many secreted factors are unknown, abrogating secretion by these systems can result in

observable effects [21,27,28]. For example, overexpression of the B. bronchiseptica T3SS locus

correlated with hypervirulence in vivo, and before a specific secreted effector was identified,

disruption of the T3SS was associated with decreased in vitro cytotoxicity and in vivo pathology

[29–32]. Although a locus homologous to known T6SSs was identified in B. bronchiseptica and

B. parapertussis genomes, its secreted effectors, function and contributions to Bordetella

pathogenesis have not yet been characterized [18,33].

To study T6SS mediated pathogenesis, we performed a functional analysis of the locus in

B. bronchiseptica strain RB50. After the 26 gene locus was fully described, an in-frame deletion of the gene encoding an apparent ortholog to a T6SS ATPase, clpV, was constructed in RB50.

33 ClpV was found to be required for macrophage cytotoxicity and induction of IL-1β, IL-6, IL-10 and IL-17 in vitro. Two dimensional gel electrophoresis of macrophage supernatant from cells exposed to either wild type RB50 or RB50ΔclpV identified several differentially regulated proteins common to know apoptotic pathways. These findings correlated with in vivo observations that ClpV is required to induce significant pathology upon infection of the lungs.

Additionally, RB50ΔclpV was rapidly cleared from the lower respiratory tract and was deficient in nasal cavity persistence in wild type mice. Our findings indicate that ClpV, a gene within an apparent T6SS encoding locus, plays an essential role in B. bronchiseptica pathogenesis in this natural host infection model.

34 Materials and Methods

Ethics Statement

This study was carried out in strict accordance with the recommendations in the Guide for the

Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee at The Pennsylvania State University at University

Park, PA (#31297 Bordetella-host Interaction). All animals were anesthetized using isoflourane or euthanized using carbon dioxide inhalation to minimize animal suffering.

Comparative protein sequence analysis.

Based on Boyer et al. analysis, there are 35 genes (BB0787-BB0821) in the B.

bronchiseptica T6SS locus. However, six genes (BB0787-BB0792) that are in the upstream of

BB0793 were previously annotated as a possible T2SS locus in RB50 by Parkhill et al, and there

are only three predicted operons (BB0793-BB0810, BB0811-BB0812, and BB0813-BB0818)

within this locus based on OperonDB (http://operondb.cbcb.umd.edu/cgi-

bin/operondb/pairs.cgi?genome_id=120). Thus, 26 genes (BB0793-BB0818) were defined as the

T6SS locus. The DNA and protein sequences corresponding to all the genes present in the T6SS

locus of B. bronchiseptica strain RB50 were obtained online (http://www.ncbi.nlm.nih.gov); the

orthologous genes in P. aeruginosa, S. enterica, and V. cholerae were located via KEGG

ortholog database (http://www.genome.jp/kegg/genes.html). The amino acid sequence similarity

was determined by comparing RB50 genes to orthologous genes in P. aeruginosa, S. enterica,

and V. cholerae using the online NCBI protein BLAST search

(http://www.ncbi.nlm.nih.gov/BLAST).

Bacterial strains and growth.

35 B. bronchiseptica strain RB50 and strain 1289 have been described elsewhere [29,34].

Bacteria were maintained on Bordet-Gengou agar (Difco) supplemented with 10% sheep blood

(Hema Resources) with 20 µg/ml streptomycin (Sigma). Bacteria were grown in liquid culture

to mid-log phase while shaking in Stainer-Scholte (SS) broth [35] overnight at 37°C.

Construction of RB50∆clpV and 1289∆clpV strains.

The RB50ΔclpV strain was constructed using an allelic exchange strategy as previously described [29]. The first three codons of clpV (BB0810) and the 630 base pairs (bp) upstream were amplified via PCR using primers flanked with EcoRI on the 5' end and HindIII on the 3' end

(Supplementary Table 1, 5’F and 5’R). The last eight codons of clpV and the 432 bp downstream

were amplified via polymerase chain reaction using primers flanked with HindIII on the 5' end

and EcoRI on the 3' end (Supplementary Table 1, 3’F and 3’R). These PCR amplified fragments

were purified (Qiagen, Valencia, CA), BamHI digested (New England Biolabs), gel purified

(Qiagen, Valencia, CA), and ligated overnight at 4°C (New England Biolabs), and amplified with

the 5' F and 3' R primers as described above. The 1,280 bp knock-out construct was then ligated

into the TOPO-TA vector, transformed into Mach1 DH5α cells (Invitrogen), and verified by

sequencing. The 1280-bp construct was digested from TOPO-TA, gel purified, and ligated

overnight into the EcoRI-digested pSS4245, a Bordetella allelic exchange vector (courtesy of S.

Stibitz). Triparental mating with DH5α harboring pSS4545 ΔclpV, DH5α containing pSS1827,

– and B. bronchiseptica strain grown under Bvg conditions by growth on BG plus 50 mM MgSO4

was done for 4 h on a BG-10 mM MgCl2-50 mM MgSO4 plate at 37°C. Then, B. bronchiseptica

containing pSS4245 ΔclpV was positively selected for by growth on BG-streptomycin-

kanamycin-50 mM MgSO4 plates and incubated for 2 days at 37°C; this step was repeated to

ensure purity. The resulting colonies were streaked onto BG plates and incubated for 2 days at

36 37°C, which resulted in colonies lacking pSS4245 and containing either the wild-type or

knockout gene. Colonies were then screened for the presence of either the wild-type or knockout

gene by using screening primers (Supplementary Table 1) which detected either the wild-type

clpV (2,003 bp) or the ΔclpV deletion (1,280 bp) with PCR. The absence of pSS4245 was

confirmed by growth on BG-streptomycin plates and lack of growth on BG-kanamycin plates.

qRT-PCR.

Quantitative reverse transcription PCR (qRT-PCR) was preformed as previously

described [29,36,37]. Briefly, bacteria grown in SS broth until an OD600 of 0.2 were subcultured

8 into four, independent five mL cultures until OD600 reached 0.8; 10 cells were immediately

pelleted by centrifugation at 4°C 8,000 RPM for five minutes. Total RNA was extracted with

Trizol (Invitrogen), treated with RNase-free DNase I (Invitrogen), and purified using RNeasy columns (Qiagen) according to the instructions of the manufacturer. One microgram of RNA from each biological replicate was reverse transcribed using 300 ng of random oligonucleotide hexamers and SuperScript III RTase (Invitrogen). The resulting cDNA was then diluted 1:100, and 1 µL aliquots were used for qRT-PCR. 300nM of primers (Supplemental Table 1B) designed using IDT DNA software (http://www.idtdna.com) were used in conjunction with 2X

SYBR green PCR master mix (Applied Biosystems). Control samples of reaction mixtures excluding reverse transcriptase were included to confirm the absence of DNA contamination; amplification of the 16S RNA amplicon was used as an internal qRT-PCR control. Dissociation curve analysis was preformed to confirm sample homogeneity. Threshold fluorescence was established within the geometric phase of exponential amplification, and the cycle threshold (CT) was determined for each sample. The CT from each replicate was averaged, and the 16S RNA

amplicon was used as internal control for data normalization. The change in transcript level was

37 determined using the relative quantitative CT method (ΔΔCT) [38]. All primers used in qRT-PCR

analysis can be found in Supplemental Table 1B.

Cytotoxicity assay.

Cytotoxicity assays were preformed as previously described [29,32,40]. Briefly, J774A.1

murine macrophages cells (ATCC TIB-67) were cultured in Dulbecco’s modified Eagle’s

medium (DMEM, Difco) supplemented with 10% fetal bovine serum, 1% penicillin-

streptomycin, 1% nonessential amino acids, and 1% sodium pyruvate. The cells were grown to

85% confluency in 5% CO2 in 96-well plates (Greiner Bio-One) at 37°C. At least one hour prior to the addition of bacteria, DMEM was replaced with RPMI medium lacking phenol red with 5% fetal bovine serum, 1% L-glutamine, 1% nonessential amino acids, and 1% sodium pyruvate.

Bacteria diluted in RPMI at multiplicities of infection (MOI) of 0.1, 1, and 10 were centrifuged onto the macrophages at 300 x g for 5 minutes and incubated in 5% CO2 at 37°C for 2, 4, and 6

hours. The cell culture supernatants were collected, and lactate dehydrogenase (LDH) release, a

measure of cytotoxicity, was analyzed using a Cytotox96 kit (Promega) according to the

instructions of the manufacturer.

Protein Extraction and 2D Gel Electrophoresis.

RAW 264.7 cells obtained from ATCC were grown in DMEM supplemented with 10%

FBS in a 5% CO2 incubator at 37˚C. The cells were grown in a monolayer in 6 well cell culture

plates (70% confluency), and serum free media was applied 3 hours before the beginning of the

assay. Six wells of monocyte cells were treated with media alone, B. bronchiseptica strain

RB50, or RB50ΔclpV at an MOI of 10, centrifuged at 250 x g for 5 minutes, and incubated at

37˚C and 5% CO2 for 2 hours. Cellular protein was extracted using previously established

methods [39]. Briefly, the supernatant was removed from each well, and the cultured cells were

38 washed twice in phosphate-buffered saline (PBS). The cells were then harvested by scraping into

3 mL cold buffer containing 50mM Tris pH 8.6,10 mM EDTA, 65 mM DTT, protease-inhibitor cocktail (Pierce), 2000 U/mL DNase I (Ambion) and 2.5 mg/mL RNase A (Qiagen), and the cellular suspensions were pooled. The cells were lysed using a homogenizer at 4˚C and centrifuged at 1000 x g to remove membranes. The protein concentration was determined using the Pierce 660nm assay (Thermo Scientific), as per manufacturer’s instructions. Two- dimensional (2D) electrophoresis was performed using the Ready-Prep 2D Starter Kit (Bio-Rad) using IPG strips with pH range 3 -10 (Bio-Rad) for the first dimension and Criterion 12.5% Tris

HCL precast gels (Bio-Rad). 500 μg of protein were loaded for each sample, and the gels were stained with Gelcode Blue reagent (Pierce). The gels were analyzed using PDQuest software

(Biorad). Protein spots were excised and trypsin digested for analysis using nano-LC MS/MS

(Waters QTOF Premier). The proteins were identified using MASCOT software (Matrix).

Intracellular staining.

Intracellular staining of J774 murine macrophages was performed as previously described

[41]. Briefly, cells grown on coverslips were washing three times with PBS and fixed in 4% paraformaldehyde in phosphate-buffered saline buffer (PBS) (Omnipur) for ten minutes. Cells

were then again washed three times with PBS and blocked with 3% bovine serum albumin in

PBS for 30 minutes. The primary antibody, Annexin V-FITC (BD Pharmingen), was diluted in

3% BSA and PBS and incubated with the cells for 1 hour at room temperature. After three

washes in PBS, the cells were stained with DAPI/PBS for 10 minutes at room temperature. Cells

were then mounted onto glass slides in Vectashield (Vector Laboratories, Inc., Burlingame, CA)

and examined using a materials microscope (Olympus BX61) at the Cytometry Core Facility at

39 University Park, PA. All images were saved as TIFF files and processed in Microsoft

Powerpoint.

Cytokine detection.

Cytokine analysis was preformed as previously described [42,43]. Briefly, cell culture supernatants were collected from J774 macrophages that were stimulated with RB50 or

RB50ΔclpV at an MOI of 0.1 for 2, 4, 6, or 24 hours or murine lung homogenates that were used for bacterial quantification and frozen at -80°C until assayed were collected. Interleukin-1β (IL-

1β), interleukin-6 (IL-6), interleukin-10 (IL-10), interleukin-17 (IL-17), Interferon-γ (IFN-γ), and tumor necrosis factor α (TNFα) concentrations were determined via ELISA in accordance with the supplier’s protocols (R&D Systems).

Animal experiments.

Wild type C57BL/6 mice were obtained from Jackson Laboratories, Bar Harbor, ME.

Mice were bred and maintained in a specific pathogen-free facility at The Pennsylvania State

University, University Park, PA, and all experiments were carried out in accordance with all institutional guidelines. All animal experiments were performed as previously described

[17,29,44]. Briefly, the number of bacterial colony forming units (CFUs) in liquid cultures was calculated based on the optical density measured by absorbance of light at 600 nm. Bacteria were then diluted to 107 CFU/ml in sterile PBS. Inocula were confirmed by plating dilutions on

BG agar and counting the resulting colonies after two days of growth at 37°C. For inoculation,

mice were sedated with 5% isoflurane (IsoFlo, Abbott Laboratories) in oxygen and inoculated by

gently pipetting 50µl PBS containing the indicated CFU of bacteria onto the external nares. For

quantification of bacterial numbers, mice were euthanized with CO2 inhalation and the indicated

organs excised. Tissues were homogenized in PBS, serially diluted and plated onto BG agar

40 plates with 20 µg/mL streptomycin, and colonies were counted after 2 days of growth at 37°C.

Survival curves were generated as previously described [29]. Mice were observed over a 28 day

period; any mouse exhibiting lethal bordetellosis, indicated by ruffled fur, labored breathing, and

diminished responsiveness, was euthanized immediately to prevent unnecessary suffering

[17,45].

Lung pathology.

Three days following inoculation with either RB50 or RB50ΔclpV, the mice were euthanized and the trachea and lungs were inflated with 1.5 ml of 10% formalin in PBS. The tissues were processed and stained with hematoxylin and eosin (H&E) at the Animal Diagnostic

Laboratory at The Pennsylvania State University, in University Park, PA. Sections were analyzed and scored on a qualitative scale as previously described [45]. An assessment of

microscopic lesions was made by one of the authors (M. J. Kennett) experienced in rodent

pathology and blinded to experimental treatment. Descriptive evaluations of the lesions were

recorded, and lung lesions were graded by using a scale of 0 to 5. Sections with no lesions and

no inflammation were given a score of 0, a score of 1 indicated slight inflammation with few or

scattered lesions and fewer than 10% of lung fields affected, a score of 2 indicated mild lesions

with 10 to 20% of lung fields affected, a score of 3 indicated moderate lesions with 20 to 30% of

the lung fields affected, and those given a score of 4 were characterized by extensive lesions,

marked inflammation, and 31 to 50% of the lung was affected. A score of 5 indicated there were

extensive lesions with >50% of the lung fields affected.

41

Results

T6SS locus in Bordetella bronchiseptica strain RB50.

In 2008, Bingle et al. used comparative sequence analysis to first identify the T6SS associated genes present in the published genome of B. bronchiseptica [33]. Four of these proteins in B. bronchiseptica shared homology to proteins that have been dubbed the ‘core components’ of the T6SS machinery: ClpV, IcmF, Hcp, and VgrS [1,33]. In this study, we

Figure 2-1 Genetic comparison of B. bronchiseptica T6SS locus to known T6SS loci. The comparison of T6SS locus among B. bronchiseptica, P. aeruginosa, S. enterica, and V. cholerae. Homologous genes are indicated with the same color, while genes with no homologues are indicated with white color. The numbers in the arrows indicate the percentages of amino acid sequence similarity compared to B. bronchiseptica. The length of arrows is relative to the length of the gene. * Indicates the gene targeted for deletion in B. bronchiseptica and its homologues. C. Chen and J. Park completed the locus analysis and figure construction.

42 characterize a contiguous 26 gene locus in B. bronchiseptica strain RB50 predicted to encode

proteins sharing high amino acid sequence similarity with highly conserved T6SS proteins found in V. cholerae, S. enterica and P. aeruginosa (Figure 1). We have retained the names of the

T6SS ‘core component’ genes, while naming the unique T6SS genes of Bordetella tssA-V,

representing “Type Six Secretion” (Figure 1). Amino acid sequence motif analysis of the

predicted protein ClpV, a putative AAA+ ATPase, revealed two ATP binding sites with Walker

A and B motifs [46] (J. Park and E.T. Harvill, unpublished data), suggesting that it may enable effector molecule binding or provide energy for protein translocation, as observed in V. cholerae,

P. aeruginosa, E. coli, and other bacteria [33,47]. The predicted IcmF protein identified in

bordetellae is similar to the IcmF- and IcmH-like proteins found in the T4SS. Yeast two hybrid assays in Edwardsiella tarda suggests that IcmF- and IcmH-like proteins may form a transport apparatus and act synergistically in translocating substrates [48]. A gene sharing 20% amino acid sequence to the Hcp encoding gene in V. cholerae and 40% identity with that of P. aeruginosa was identified in B. bronchiseptica. Hcp may function as an effector and/or assemble into a hexameric ring structure that forms a channel or pilus for conduction of other effectors through the cell membrane, which has been shown to be essential for T6SS mediated virulence in Vibrio and Pseudomonas [49]. A homolog of vgrS was identified in B. bronchiseptica, possessing 34% and 41% sequence identity with V. cholerae and P. aeruginosa

vgrS genes, respectively, and contains the GP5 region predicted to form the base of the needle apparatus. VgrS of V. cholerae contains regions of homology to the actin cross-linking domain of an RtxA toxin [50], and more recently has been shown to share homology with the gp27 of T4 bacteriophage, which forms part of a tail spike apparatus for membrane penetration [51,52]. Both

Hcp and VgrG are found in the secretomes of most bacteria possessing a functional T6SS, even

43 though they lack an export signal peptide [4,6,33]. These same four ‘core’ components of the

T6SS machinery were also identified in B. parapertussis strain 12822; however, further DNA

sequence analysis of this locus identified frameshift mutations in upstream genes and a

pseudogene that replaced vgrG (J. Park and E.T. Harvill, unpublished data), suggesting it may be

either defective or functionally different from the locus found in B. bronchiseptica. No T6SS

genes were found in the sequenced genome of B. pertussis strain Tohoma I, indicating that this

locus may have been lost through genome degradation in the course of B. pertussis evolution

[53].

Deletion of clpV from B. bronchiseptica strain RB50.

ClpV has been shown to be required for

translocation of T6SS effector proteins essential for

virulence in V. cholerae, P. aeruginosa, and E. coli

[3,4,54]. To investigate the role of T6SS in B.

bronchiseptica pathogenesis, we constructed an in-

frame deletion of clpV (BB0810) in the genome of

B. bronchiseptica strain RB50 (RB50ΔclpV) by

utilizing the bordetellae allelic exchange vector

pSS4245 [55]. The wild type gene produced a 2,003

bp PCR product, whereas the clpV mutant region

resulted in a product of 1,280 bp, as expected Figure 2-2 Confirmation of RB50ΔclpV construction and T6SS expression. (Figure 2A). Following deletion of clpV, expression A. PCR analysis of clpV in RB50 (left) and RB50ΔclpV (right). B. RT-PCR analysis of relative of the four core T6SS transcripts, icmF, hcp, vgrG, expression of vgrG, hcp, clpV, icmF, bvgS, and fhaB in RB50ΔclpV relative to RB50 expressed as mean ± standard deviation. Each gene was normalized to the expression of 16sRNA. Mutant construction was completed by N. Spidale. 44 and clpV in RB50ΔclpV was compared with expression in the parental strain by qRT-PCR.

Expression of icmF, hcp, and vgrG in the mutant strain remained comparable to that of RB50

while clpV expression was undetectable, suggesting that this gene has been effectively disrupted

without altering the expression of neighboring genes (Figure 2B). The growth rate of

RB50ΔclpV in SS broth at 37oC was not different from that of parental strain RB50 (data not

shown), further suggesting that clpV was successfully disrupted without causing additional

defects.

T6SS contributes to cytotoxicity of murine macrophages.

In other bacterial systems, the T6SS has been found to mediate interactions between

bacteria and phagocytic cells, including

protection against amoeba predation,

enhanced intracellular survival within

macrophages, and the ability to directly

kill macrophages in vitro [4,6,7,13].

Additionally, because B. bronchiseptica

is known to be highly cytotoxic to cell

cultured macrophages, we hypothesized

that the T6SS may play a role in

cytotoxicity. Using a lactate

dehydrogenase (LDH) release assay, we

measured the cytotoxic effects of B.

Figure 2-3 T6SS mediated cytotoxicity of murine macrophages. A. LDH release assay monitoring cytotoxicity of J774 murine macrophages at an MOI of 1 for 2, 4, and 6 hour incubations with RB50, RB50ΔclpV, 1289, or 1289ΔclpV. B. J774 macrophages stained with Annexin V (green) and DAPI (blue) after incubation with RB50, RB50ΔclpV, or media alone for three hours. * denotes p value < 0.01. S. E. Hester contributed to the cytotoxicity assay.

45 bronchiseptica on murine macrophages. When macrophages were infected with RB50 at a MOI of 1, we observed increasing levels of cytotoxicity over the first 6 hours of incubation (Figure

3A). 71% of macrophages were lysed by RB50 by six hours, as previously seen [29,56,57].

Minimal cytotoxicity at any point throughout the 6 hour incubation was observed in macrophages exposed to RB50ΔclpV (Figure 3A). Because complementation by plasmid expression of clpV was unsuccessful after multiple attempts, clpV was deleted from RB50 three times, independently, with identical effects on cytotoxicity (data not shown). Further, we deleted clpV from another wild-type B. bronchiseptica isolate, strain 1289, known to exhibit hypervirulence and increased cytotoxicity associated with T3SS over-expression. As expected, wild type 1289 induced 92% cytotoxicity after six hours incubation, more than RB50 (Figure

3A). However, the clpV mutant of 1289 induced minimal cytotoxicity at all time points, similar to RB50ΔclpV, suggesting that ClpV is required for macrophage killing in multiple B. bronchiseptica strains. Not surprisingly, a mutant lacking hcp, which encodes a putative T6SS structural component, Hcp, has been constructed in our laboratory and also does not induce macrophage cytotoxicity (S. J. Muse and E. T. Harvill, unpublished data), suggesting that the B. bronchiseptica T6SS is required for macrophage cell death.

To examine the ClpV-mediated mechanism of LDH release, J774 cells were stained for the presence of Annexin V after being incubated with either strain RB50 or RB50ΔclpV at an

MOI of 100 for 2 hours. Significantly more Annexin V positive macrophages were observed when cells were incubated with RB50 compared to RB50ΔclpV (Figure 3B). Additionally,

RB50ΔclpV stimulated small Annexin V puncta near the membrane without stimulating full

46 Annexin V membrane staining. Together, these results show that the T6SS contributes to the apoptotic death of macrophages in vitro.

Macrophage proteome changes in response to T6SS.

To investigate which cell signaling pathways were associated with T6SS mediated macrophage cell death, we analyzed the proteome of murine macrophages exposed to either

RB50 or RB50ΔclpV. Murine J774 macrophages were stimulated with either RB50 or

RB50ΔclpV at an

MOI of 10 for 2

hours, lysed, and

analyzed by two

dimensional gel

electrophoresis. This

approach enables

Figure 2-4 T6SS-mediated proteomic changes during in vitro macrophage infection. identification of Two dimensional gel electrophoresis was performed on whole cell extract from macrophages infected with either RB50 or RB50ΔclpV at an MOI of 10 for 2 hours. A bacterial proteins composite image of proteins identified as increased during RB50 compared to RB50DclpV infection was created using PDQuest software, and the proteins selected for mass spectrometry analysis are circled in red and numbered 1 to 6. 2D gel secreted into electrophoresis and analysis was completed by S. J. Muse. eukaryotic cells in a contact dependent manner [4]. This approach also enables detection of proteins that are differentially produced by macrophages in response to a functional T6SS. A total of 431 different proteins were visualized by two-dimensional gel electrophoresis, and 283 proteins differed between RB50 or RB50ΔclpV infected macrophages (Figure 4). Six of the most prominent proteins that were observed only in RB50 infection compared to RB50ΔclpV infection

47 (proteins 1-6) were identified via mass spectrophotometry. Murine macrophage proteins identified included pyruvate kinase isoxyme M1 M2, transcription factor E2F7, isocitrate dehydrogenase NADP Fragment, voltage dependent anion selective channel protein 2, and guanine nucleotide binding protein subunit beta 2 like 1. NADH-quinone oxidoreductase subunit C was the only bacterial protein identified. Together, these data suggest that deletion of clpV abrogates T6SS secretion into murine macrophages and changes the cellular response

produced during infection in vitro.

T6SS stimulates IL-1β and IL-6

production in vitro.

To determine how ClpV-mediated

cytotoxicity and differential proteome

regulation affected the cytokine output, we

assayed production of multiple cytokines

by macrophages exposed to either RB50 or

RB50ΔclpV. IL-1β, IL-6, IL-10, IL-17,

IFN-γ, and TNFα were analyzed since

these cytokines are known to contribute to

host immunity or pathogenesis [42,45].

No changes were observed in TNFα

production as early as 2 hours post

inoculation (data not shown). Even after

Figure 2-5 T6SS-mediate cytokine production in vitro. A-C. Supernatants from infected murine macrophages were recovered after 24 hours at an MOI of 0.1 of RB50 (dark bars) or RB50ΔclpV (light bars) and assayed for TNFα (A), IL-6 (B), IL-1β (C), IL-17 (D), IL-10 (E) and IFN-γ (F). * denotes p value < 0.05.

48 24 hours of stimulation, cells stimulated with either wild type bacteria or the mutant produced

similar amounts of TNFα (Figure 5A). Macrophages exposed to RB50 produced more IL-6 and

IL-1β than those exposed to RB50ΔclpV (Figure 5B-C), suggesting that the T6SS may stimulate

IL-6 and IL-1β production independently of TNFα induction. IL-17, known to be downstream of

IL-1β, was also up-regulated in RB50 stimulated macrophages compared to macrophages stimulated by the clpV mutant (Figure 5D). Stimulation of IL-6, IL-1β, and IL-17, independent of ClpV-mediated TNFα production suggests that the T6SS may play an essential role in recruiting immune cells to the site of infection. We also investigated production of the anti- inflammatory cytokine IL-10 and its balanced counterpart IFN-γ. We observed that wild type bacteria induced more IL-10 production than the clpV mutant and very low levels of IFN-γ, suggesting that a mutant lacking clpV may have alternative cell recruitment mechanisms by altering IL-10 production (Figure 5E-F). Together, these data suggests that the T6SS may alter cytokines that regulate inflammation initiation, cell recruitment, and affect downstream adaptive immune response pathways.

T6SS is required for persistence in the murine respiratory tract.

To determine whether ClpV-dependent effects on cytokine production in vitro contribute to in vivo colonization and persistence of Bordetella, we used a well established murine model of infection [58,59]. We inoculated C57BL/6 mice with 5×105 CFU of B. bronchiseptica strain

RB50 or RB50ΔclpV, and bacterial numbers were determined in the nasal cavity, trachea, and lungs at 0, 3, 7, 14, 28 and 49 days post-inoculation (Figures 6A-C). RB50ΔclpV colonized the respiratory tract similarly to RB50 for the first three days post-inoculation. However, by day 7

RB50ΔclpV numbers in the lungs were one tenth that of wild type (Figure 6C). Compared to wild type, by day 14 fewer RB50ΔclpV bacteria were reported in the trachea, and by day 28

49 numbers of the mutant were lowers in the nasal cavity (Figures 6A and 6B). The bacterial load

of both the mutant and wild type decline over time in the nasal cavity, trachea, and lungs;

however, RB50ΔclpV was cleared from the trachea and lungs 28 days post-inoculation, while

significant numbers of RB50 could still be detected in the lungs at 49 days post-inoculation.

Wild-type B. bronchiseptica is known to persist indefinitely (>150 days) in the nasal cavities of

Figure 2-6 Colonization and histopathology of RB50ΔclpV compared to wild type. A-C. Colonization of RB50 verses RB50ΔclpV in C57BL/6 mice at an inoculation dose of 5×105 CFU in 50 µL in the nasal cavity (A), trachea (B), and lung (C). D. Representative H&E lung sections from C57BL/6 mice on day 3 post-inoculation and their average pathology scores. Colonization data was collected and analyzed by O. Y. Rolin, and histopathology was analyzed by M. J. Kennett.

50 laboratory mice at levels greater than 103 CFU [60]. While RB50ΔclpV was still present in the

nasal cavity after 49 days, it was reduced to approximately 102 CFU (Figure 6A). Together, these data indicated that the T6SS may contribute to the ability of B. bronchiseptica to persist in the murine respiratory tract.

T6SS mediates increased pathology and cell recruitment in vivo.

When mice were dissected three days following infection with RB50, their lungs were visibly inflamed and erythematous. In contrast, the lungs of mice infected with RB50ΔclpV appeared healthy (data not shown). We hypothesized that although bacterial loads recovered from each group were comparable at this time point, T6SS activity was causing enhanced leukocyte recruitment into the lungs and increased tissue damage and host cell necrosis.

Histological analysis of lungs stained with H&E revealed significantly attenuated inflammatory pathology in mice infected with RB50ΔclpV (2.3 score) relative to those infected with RB50 (3.7 score) (Figure 6D). In the lungs of RB50 infected mice, we observed a robust accumulation of polymorphonuclear cells (PMNs) cuffing the perivascular spaces, infiltrating into the connective tissue underlying the respiratory epithelium of the bronchioles, and collecting within alveolar spaces. In comparison, lungs from mice infected with RB50ΔclpV had visibly reduced cellular infiltration into perivascular spaces and very little infiltrate in the alveolar spaces. Despite significantly higher pathology scores, a similar amount of necrotic cell death was observed in both the RB50 and RB50ΔclpV infected lungs (Figure 3D). The LD50 of wild type RB50 in

C57BL/6 mice is approximately 106.3 CFU. Fatality from this dose occurs within three days of

inoculation. [29]. C57BL/6 mice inoculated with up to 108.1 CFU of RB50ΔclpV survived for at

least 90 days, indicating the virulence of B. bronchiseptica requires the T6SS (data not shown).

51 These data suggest that the T6SS contributes to B. bronchiseptica-mediated pathology and decreases the mean lethal dose of B. bronchiseptica strain RB50.

T6SS modulates a Th1 immune response.

We observed decreased cytokine production in vitro and decreased cell recruitment in vivo. Therefore, we measured cytokines at the infection site to determine whether a functional

T6SS alters cytokine production in ways that might cause skewing of the T helper cytokine

response profile. When directly assaying lung

homogenates from day 7 and day 28 post inoculation

for the presence of cytokines, we found that mice

infected with RB50ΔclpV had lower levels of IL-6 and

IL-17, cytokines associated with Th17 responses, and

significantly higher levels of IFN-γ, a Th1 cytokine

when compared to RB50 infected mice (Figure 7A-C).

No differences were observed between RB50 or

RB50ΔclpV infection in TNFα, IL-1β, and IL-10 (data

not shown). Th1 responses have been shown to be

critical for immune mediated clearance of B.

bronchiseptica, while Th17 cells have been shown to

contribute to the clearance of closely related pathogen,

B. pertussis [61]. These findings suggest that the

T6SS may delay immune mediated clearance by

Figure 2-7 Effects of the T6SS on cytokine production in vivo. A-C. Cytokines recovered from lungs of mice infected with RB50 (dark bars) or RB50ΔclpV (light bars) for IL-6 (A), IL-17 (B), and IFN-γ (C) on days 7 and 28 post-inoculation. * denotes p value < 0.05. Cytokine experiments and analysis was completed by O. Y. Rolin.

52 shifting the immune reaction to a Th17 response and preventing the development of critical Th1 responses.

53

Discussion

This work represents the first investigation of Bordetella T6SS function and describes a

robust natural host infection model in which the subtleties of complex immune interactions with

the host can be dissected. As with multiple other pathogens, we find that mutation of clpV

affects macrophages cytotoxicity in vitro; however, we were able to identify several

macrophages proteins that change in response to ClpV. Additionally, we made the novel

observation that ClpV contributes to IL-1β, IL-6, IL-10, and IL-17 production in vitro, and using natural host infection, confirmed effects on cytokine production in vivo. Furthermore, we were able to determine that ClpV is required to induce enhanced immunopathology and facilitate murine respiratory tract persistence. Our results suggest that the T6SS is a novel virulence factor that significantly contributes to the virulence of the respiratory pathogen B. bronchiseptica.

Numerous Gram-negative pathogens, such as Vibrio choleraee [6], Aeromonas

hydrophila [7], Legionella pneumophila [8], Salmonella typhimurium, and Yersinia pseudotuberculosis [62] possess a T6SS that contributes to virulence during in vitro infections.

This study shows that the B. bronchiseptica T6SS gene clpV is required for macrophage cytotoxicity and identifies proteomic changes associated with T6SS-mediated cellular apoptosis, including proteins associated with intracellular bacterial survival (pyruvate kinase isozymes

M1/M2) [63], eukaryotic intracellular signaling (voltage dependent anion selection channel)

[64], structural mimicry (guanine nucleotide binding protein subunit beta 2) [65], and apoptosis

(voltage dependent anion selection channel, transcription factor E2F7 and isocitrate dehydrogenase NADP fragment) [66–68]. For example, although the E2F family broadly contributes to eukaryotic cell cycle regulation, altered regulation of E2F7 and E2F8 can lead to

54 apoptosis through an E2F1-dependent manner [66]. Additionally, the B. bronchiseptica T6SS could be affecting host cellular structure by targeting GTPases, as recently observed in

Burkholderia cenocepacia [69]. It is currently unclear if the eukaryotic proteins identified here are directly controlled by bacterial factors, or whether or not they are secondary effects. The only bacterial protein identified in macrophages following RB50 infection was a NADH-quinone oxidoreductase (NQO) subunit C. In other Gram-negative pathogens, a six subunit complex including an NQO is used to transport sodium, catalyzing electron transfer from NADH to quinone [70–72]. Pathogens, such as Helicobacter pylori, have been shown to require similar reductases to withstand oxidative stress while inside phagocytic cells [73]. Alternatively, there is speculation that deregulation of NADH may be involved in eukaryotic programmed cell death, suggesting a pathogenic mechanism for this protein [74,75]. Further research examining these discrete protein interactions will help to better elucidate the specific T6SS pathogenesis mechanisms.

It is likely that interactions between the B. bronchiseptica T6SS and macrophages mediate critical activities very early in the course of infection. Within the first three days of

RB50 infection, although the colonization burden of mutant and wild type is equal, there is severe T6SS-dependent immune-pathology and cell recruitment. Both in vitro and in vivo, ClpV was observed to stimulate IL-1β, IL-6, and IL-17 production, potentially explaining the heightened cellular recruitment to the lungs. The increased lung leukocytes recruitment in mice infected with RB50 might be predicted to lead to more rapid clearance of RB50; however, reduced numbers of RB50ΔclpV were recovered from lungs as early as 7 days post-inoculation.

This increased immunopathology correlates with lower in vivo production of T-helper 1 (Th1) cytokines, such as IFN-γ. Antibody production and Th1 responses have been shown to be

55 essential for B. bronchiseptica clearance in vivo, and its hypothesized that B. bronchiseptica has evolved to stimulate IL-10 production to evade clearance [76,77]. RB50ΔclpV stimulates a robust Th1 response, which likely contributes to its increased clearance from the lower respiratory tract. The exact mechanisms behind B. bronchiseptica-induced pathology remain unclear, although this could be attributed to a T6SS mediated cytotoxicity toward macrophages and a subsequent inflammatory response. Although not fully understood, heightened immune- pathology and increased bacteria numbers may enable bacteria to cause disease symptoms, such as coughing that may enhance transmission. Alternatively, localized pathology may facilitate initial colonization by inducing inflammation that disrupts mucociliary clearance mechanisms or resident host microflora.

B. bronchiseptica has previously been observed to persist indefinitely in the nasal cavity of experimental mice; however the clpV mutant persists at much lower levels [29,59,78]. It is unclear whether the T6SS enables long-term persistence at higher numbers in the nasal cavity by modulating key early immune interactions that subvert productive adaptive immune responses or whether T6SS mediates ongoing resistance to opsonophagocytic clearance. Recent work showed that P. aeuroginosa toxin, Tse2, part of a toxin-immunity system secreted through the T6SS, mediates killing of other prokaryotic organisms, but not eukaryotic organisms [52]. Although we have not been able to identify any obvious Tse2 homologs in the B. bronchiseptica genome to date, the T6SS may mediate protection or confer an advantage over host nasal microflora, preventing its displacement by competitor species.

Of the three classic Bordetella strains that have had their genomes sequenced, B. bronchiseptica strain RB50, B. parapertussis strain 12822 and B. pertussis strain Tahoma I,

56 RB50 is the only strain whose T6SS is predicted to be functional. Previous work has shown that

RB50 is also the only one of these strains known to be cytotoxic to macrophages; B. pertussis

and B. parapertussis are shown to be non-cytotoxic for up to six hours in vitro [18]. Strikingly,

RB50ΔclpV cytotoxicity is similar to observed levels of B. pertussis and B. parapertussis

cytotoxicity. Although Bordetella cytotoxicity has been attributed to ACT and the T3SS, the loss

of T6SS function may explain why B. pertussis and B. parapertussis strains do not kill

macrophages even though they express ACT and, in some cases, have a functional T3SS

[21,79,80]. Surprisingly, RB50ΔclpV, which would be expected to retain T3SS-mediated

cytotoxicity, killed less than 10% of macrophages even after six hours. Notably, the T3SS and

T6SS have been shown to be expressed at alternate times during infection in Pseudomonas and

Salmonella [13,54,81]. Both secretion systems appear to be required for the observed

cytotoxicity of RB50, and there is significant overlap in the phenotypes of T3SS and T6SS

mutants. Both mutants display overall decreased pathology, shortened duration of colonization,

and attenuated virulence in vivo [30,32,82,83]. However, only the T6SS appears to be required

for IL-6 production and nasal cavity persistence. Further work is necessary to understand the interactions between these secretion systems and how they independently and cooperatively affect the infection process.

57 Authors and Contributions

Authors: Laura S. Weyrich1,3†, Olivier Y. Rolin1,4†, Sarah J. Muse1,3, Jihye Park1,5,

Nicholas Spidale2, Mary J. Kennett1, Sara E. Hester1,3, Chun Chen2, Edward G. Dudley2, and

Eric T. Harvill1

1Department of Veterinary and Biomedical Sciences, The Pennsylvania State University

115 Henning Building, University Park, PA 16802

2Department of Food Science, The Pennsylvania State University, 202 Food Science

Building, University Park, PA, 16802.

3Graduate Program in Biochemistry, Microbiology, and Molecular Biology

4 Graduate Program in Immunology and Infectious Disease

5Graduate Program in Bioinformatics

† Authors contributed equally.

Conceived and designed experiments: LSW, OYR, NS, SHE, EGD, ETH

Preformed experiments: LSW, OYR, SJM, SEH

Analyzed data: LSW, OYR, MJK, JP, CC, EGD, ETH

Wrote manuscript: LSW, OYR, ETH

58

References

1. Boyer F, Fichant G, Berthod J, Vandenbrouck Y, Attree I (2009) Dissecting the bacterial type VI secretion system by a genome wide in silico analysis: what can be learned from available microbial genomic resources? BMC Genomics 10: 104.

2. Boyd EF, Cohen A, Naughton L, Ussery D, Binnewies T, et al. (2008) Molecular analysis of the emergence of pandemic Vibrio parahaemolyticus. BMC Microbiology 8: 110.

3. Filloux A, Hachani A, Bleves S (2008) The bacterial type VI secretion machine: yet another player for protein transport across membranes. Microbiology 154: 1570-1583. doi:10.1099/mic.0.2008/016840-0

4. Cascales E (2008) The type VI secretion toolkit. EMBO Rep 9: 735-41. doi:embor2008131 [pii] 10.1038/embor.2008.131

5. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, et al. (2011) Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475: 343-347. doi:10.1038/nature10244

6. Pukatzki S, Ma AT, Sturtevant D, Krastins B, Sarracino D, et al. (2006) Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system. Proceedings of the National Academy of Sciences of the United States of America 103: 1528-1533. doi:10.1073/pnas.0510322103

7. Suarez G, Sierra JC, Sha J, Wang S, Erova TE, et al. (2008) Molecular characterization of a functional type VI secretion system from a clinical isolate of Aeromonas hydrophila. Microbial Pathogenesis 44: 344-361.

8. Purcell M, Shuman HA (1998) The Legionella pneumophila icmGCDJBF Genes Are Required for Killing of Human Macrophages. Infect. Immun. 66: 2245-2255.

9. Schlieker C, Zentgraf H, Dersch P, Mogk A (2005) ClpV, a unique Hsp100/Clp member of pathogenic proteobacteria. Biological Chemistry 386: 1115-1127. doi:10.1515/BC.2005.128

10. Wang J, Li C, Yang H, Mushegian A, Jin S (1998) A novel serine/threonine protein kinase homologue of Pseudomonas aeruginosa is specifically inducible within the host infection site and is required for full virulence in neutropenic mice. J. Bacteriol 180: 6764-6768.

11. Schell MA, Ulrich RL, Ribot WJ, Brueggemann EE, Hines HB, et al. (2007) Type VI secretion is a major virulence determinant in Burkholderia mallei. Molecular Microbiology 64: 1466-1485. doi:10.1111/j.1365-2958.2007.05734.x

59 12. Dudley EG, Thomson NR, Parkhill J, Morin NP, Nataro JP (2006) Proteomic and microarray characterization of the AggR regulon identifies a pheU pathogenicity island in enteroaggregative Escherichia coli. Molecular Microbiology 61: 1267-1282.

13. Parsons DA, Heffron F (2005) sciS, an icmF Homolog in Salmonella enterica Serovar Typhimurium, Limits Intracellular Replication and Decreases Virulence. Infect. Immun. 73: 4338-4345.

14. Broms JE, Lavander M, Sjostedt A (2009) A Conserved {alpha}-Helix Essential for a Type VI Secretion-Like System of Francisella tularensis. J. Bacteriol. 191: 2431-2446. doi:10.1128/jb.01759-08

15. Mattoo S, Cherry JD (2005) Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev 18: 326-82.

16. Goodnow RA (1980) Biology of Bordetella bronchiseptica. Microbiol Rev 44: 722-38.

17. Harvill ET, Cotter PA, Yuk MH, Miller JF (1999) Probing the function of Bordetella bronchiseptica adenylate cyclase toxin by manipulating host immunity. Infect Immun 67: 1493-500.

18. Shrivastava R, Miller JF (2009) Virulence factor secretion and translocation by Bordetella species. Current Opinion in Microbiology 12: 88-93.

19. Mazar J, Cotter PA (2006) Topology and maturation of filamentous haemagglutinin suggest a new model for two-partner secretion. Mol Microbiol 62: 641-54.

20. Finn TM, Stevens LA (1995) Tracheal colonization factor: a Bordetella pertussis secreted virulence determinant. Molecular Microbiology 16: 625-634. doi:10.1111/j.1365- 2958.1995.tb02425.x

21. Yuk MH, Harvill ET, Cotter PA, Miller JF (2000) Modulation of host immune responses, induction of apoptosis and inhibition of NF-kappaB activation by the Bordetella type III secretion system. Mol Microbiol 35: 991-1004.

22. Hickey FB, Brereton CF, Mills KHG (2008) Adenylate cycalse toxin of Bordetella pertussis inhibits TLR-induced IRF-1 and IRF-8 activation and IL-12 production and enhances IL-10 through MAPK activation in dendritic cells. J Leukoc Biol 84: 234-243. doi:10.1189/jlb.0208113

23. Kirimanjeswara GS, Agosto LM, Kennett MJ, Bjornstad ON, Harvill ET (2005) Pertussis toxin inhibits neutrophil recruitment to delay antibody-mediated clearance of Bordetella pertussis. J Clin Invest 115: 3594-601.

60 24. Glaser P, Sakamoto H, Bellalou J, Ullmann A, Danchin A (1988) Secretion of cyclolysin, the calmodulin-sensitive adenylate cyclase-haemolysin bifunctional protein of Bordetella pertussis. Embo J 7: 3997-4004.

25. Hodak H, Clantin B, Willery E, Villeret V, Locht C, et al. (2006) Secretion signal of the filamentous haemagglutinin, a model two-partner secretion substrate. Mol. Microbiol. 61: 368-382. doi:10.1111/j.1365-2958.2006.05242.x

26. Weiss AA, Johnson FD, Burns DL (1993) Molecular characterization of an operon required for pertussis toxin secretion. Proc Natl Acad Sci U S A 90: 2970-4.

27. Skinner JA, Pilione MR, Shen H, Harvill ET, Yuk MH (2005) Bordetella type III secretion modulates dendritic cell migration resulting in immunosuppression and bacterial persistence. J Immunol 175: 4647-52.

28. Stockbauer KE, Foreman-Wykert AK, Miller JF (2003) Bordetella type III secretion induces caspase 1-independent necrosis. Cell. Microbiol. 5: 123-132.

29. Buboltz AMNTL (2008) Enhanced TTSS function contributes to the hypervirulence of a Bordetella bronchiseptica lineage. Journal of Bacteriology Submitted.

30. Medhekar B, Shrivastava R, Mattoo S, Gingery M, Miller JF (2009) Bordetella Bsp22 forms a filamentous type III secretion system tip complex and is immunoprotective in vitro and in vivo. Molecular Microbiology 71: 492-504. doi:10.1111/j.1365-2958.2008.06543.x

31. Panina EM, Mattoo S, Griffith N, Kozak NA, Yuk MH, et al. (2005) A genome-wide screen identifies a Bordetella type III secretion effector and candidate effectors in other species. Mol Microbiol 58: 267-79.

32. Yuk MH, Harvill ET, Miller JF (1998) The BvgAS virulence control system regulates type III secretion in Bordetella bronchiseptica. Mol Microbiol 28: 945-59.

33. Bingle LEH, Bailey CM, Pallen MJ (2008) Type VI secretion: a beginner’s guide. Current Opinion in Microbiology 11: 3-8.

34. Cotter PA, Miller JF (1994) BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. Infect Immun 62: 3381- 90.

35. Stainer DW, Scholte MJ (1970) A simple chemically defined medium for the production of phase I Bordetellla pertussis. J. Gen. Microbiol. 63: 211-220.

36. Buboltz AM, Nicholson TL, Parette MR, Hester SE, Parkhill J, et al. (2008) Replacement of Adenylate Cyclase Toxin in a Lineage of Bordetella bronchiseptica. J. Bacteriol.: JB.00226-08. doi:10.1128/jb.00226-08

61 37. Nicholson TL (2007) Construction and validation of a first-generation Bordetella bronchiseptica long-oligonulceotide microarray by transcriptional profiling the Bvg regulon. BMC Genomics 8: 220.

38. Saeed AI, Sharov V, White J, Li J, Liang W, et al. (2003) TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34: 374-8.

39. Dupont A, Tokarski C, Dekeyzer O, Guihot A-L, Amouyel P, et al. (2004) Two- dimensional maps and databases of the human macrophage proteome and secretome. Proteomics 4: 1761-1778. doi:10.1002/pmic.200300691

40. Mattoo S, Yuk MH, Huang LL, Miller JF (2004) Regulation of type III secretion in Bordetella. Mol Microbiol 52: 1201-14.

41. Manna D, Aligo J, Xu C, Park WS, Koc H, et al. (n.d.) Endocytic Rab proteins are required for hepatitis C virus replication complex formation. Virology 398: 21-37.

42. Mann PB, Kennett MJ, Harvill ET (2004) Toll-like receptor 4 is critical to innate host defense in a murine model of bordetellosis. J Infect Dis 189: 833-6.

43. Zhang X, Hester SE, Kennett MJ, Karanikas AT, Bendor L, et al. (2011) Interleukin-1 receptor signaling is required to overcome the effects of pertussis toxin and for efficient infection- or vaccination-induced immunity against Bordetella pertussis. Infect. Immun. 79: 527-541. doi:10.1128/IAI.00590-10

44. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of antibodies in immunity to Bordetella infections. Infect Immun 71: 1719-24.

45. Mann PB, Elder KD, Kennett MJ, Harvill ET (2004) Toll-like receptor 4-dependent early elicited tumor necrosis factor alpha expression is critical for innate host defense against Bordetella bronchiseptica. Infect Immun 72: 6650-8.

46. Hanson PI, Whiteheart SW (2005) AAA+ proteins: have engine, will work. Nat Rev Mol Cell Biol 6: 519-529. doi:10.1038/nrm1684

47. Bönemann G, Pietrosiuk A, Diemand A, Zentgraf H, Mogk A (2009) Remodelling of VipA/VipB tubules by ClpV-mediated threading is crucial for type VI protein secretion. EMBO J 28: 315-325. doi:10.1038/emboj.2008.269

48. Zheng J, Leung KY (2007) Dissection of a type VI secretion system in Edwardsiella tarda. Molecular Microbiology 66: 1192-1206. doi:10.1111/j.1365-2958.2007.05993.x

49. Filloux A (2009) The type VI secretion system: a tubular story. EMBO J 28: 309-310.

50. Sheahan K-L, Cordero CL, Fullner Satchell KJ (2004) Identification of a domain within the multifunctional Vibrio cholerae RTX toxin that covalently cross-links actin. Proceedings of

62 the National Academy of Sciences of the United States of America 101: 9798-9803. doi:10.1073/pnas.0401104101

51. Pukatzki S, McAuley SB, Miyata ST (2009) The type VI secretion system: translocation of effectors and effector-domains. Current Opinion in Microbiology 12: 11-17.

52. Hood RD, Singh P, Hsu F, Güvener T, Carl MA, et al. (2010) A type VI secretion system of Pseudomonas aeruginosa targets a toxin to bacteria. Cell Host Microbe 7: 25-37. doi:10.1016/j.chom.2009.12.007

53. Diavatopoulos DA, Cummings CA, Schouls LM, Brinig MM, Relman DA, et al. (2005) Bordetella pertussis, the causative agent of whooping cough, evolved from a distinct, human-associated lineage of B. bronchiseptica. PLoS Pathog 1: e45.

54. Mougous JD, Cuff ME, Raunser S, Shen A, Zhou M, et al. (2006) A virulence locus of Pseudomonas aeruginosa encodes a protein secretion apparatus. Science 312: 1526-1530. doi:10.1126/science.1128393

55. Buboltz AM, Nicholson TL, Weyrich LS, Harvill ET (2009) Role of the type III secretion system in a hypervirulent lineage of Bordetella bronchiseptica. Infect. Immun 77: 3969- 3977. doi:10.1128/IAI.01362-08

56. Heininger U, Cotter PA, Fescemyer HW, Martinez de Tejada G, Yuk MH, et al. (2002) Comparative phenotypic analysis of the Bordetella parapertussis isolate chosen for genomic sequencing. Infect Immun 70: 3777-84.

57. Mann P, Goebel E, Barbarich J, Pilione M, Kennett M, et al. (2007) Use of a Genetically Defined Double Mutant Strain of Bordetella bronchiseptica Lacking Adenylate Cyclase and Type III Secretion as a Live Vaccine. Infect Immun 75: 3665-72.

58. BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. (1994). Available: http://medjournal.hmc.psu.edu:2060/pmc/articles/PMC302969/?tool=pubmed. Accessed 24 Jul 2011.

59. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of Antibodies in Immunity to Bordetella Infections. Infect. Immun. 71: 1719-1724. doi:10.1128/iai.71.4.1719-1724.2003

60. Pishko EJ, Kirimanjeswara GS, Pilione MR, Gopinathan L, Kennett MJ, et al. (2004) Antibody-mediated bacterial clearance from the lower respiratory tract of mice requires complement component C3. Eur J Immunol 34: 184-93.

61. Andreasen C, Powell DA, Carbonetti NH (2009) Pertussis Toxin Stimulates IL-17 Production in Response to Bordetella pertussis Infection in Mice. PLoS ONE 4: e7079. doi:10.1371/journal.pone.0007079

63 62. Schlieker C, Zentgraf H, Dersch P, Mogk A (2005) ClpV, a unique Hsp100/Clp member of pathogenic proteobacteria. Biological Chemistry 386: 1115-1127. doi:doi: 10.1515/BC.2005.128

63. Williams JM, Chen G, Zhu L, Rest RF (1998) Using the yeast two‐hybrid system to identify human epithelial cell proteins that bind gonococcal Opa proteins: intracellular gonococci bind pyruvate kinase via their Opa proteins and require host pyruvate for growth. Molecular Microbiology 27: 171-186. doi:10.1046/j.1365-2958.1998.00670.x

64. Shoshan-Barmatz V, Israelson A, Brdiczka D, Sheu SS (2006) The Voltage-Dependent Anion Channel (VDAC): Function in Intracellular Signalling, Cell Life and Cell Death. Current Pharmaceutical Design 12: 2249-2270. doi:10.2174/138161206777585111

65. Stebbins CE, Galán JE (2001) Structural mimicry in bacterial virulence. Nature 412: 701- 705. doi:10.1038/35089000

66. Li J, Ran C, Li E, Gordon F, Comstock G, et al. (2008) Synergistic Function of E2F7 and E2F8 Is Essential for Cell Survival and Embryonic Development. Developmental Cell 14: 62-75. doi:10.1016/j.devcel.2007.10.017

67. Kim SY, Lee SM, Tak JK, Choi KS, Kwon TK, et al. (2007) Regulation of singlet oxygen- induced apoptosis by cytosolic NADP+-dependent isocitrate dehydrogenase. Molecular and Cellular Biochemistry 302: 27-34. doi:10.1007/s11010-007-9421-x

68. Lang F, Lang PA, Lang KS, Brand V, Tanneur V, et al. (2004) Channel-induced apoptosis of infected host cells?the case of malaria. Pfl�gers Archiv European Journal of Physiology 448: 319-324. doi:10.1007/s00424-004-1254-9

69. Rosales-Reyes R, Skeldon AM, Aubert DF, Valvano MA (2011) The Type VI secretion system of Burkholderia cenocepacia targets multiple Rho family GTPases disrupting the actin cytoskeleton and the assembly of NADPH oxidase complex in macrophages. Cellular Microbiology. Available: http://www.ncbi.nlm.nih.gov/pubmed/22023353. Accessed 8 Nov 2011.

70. Brandt U (2006) Energy Converting NADH: Quinone Oxidoreductase (Complex I). Annual Review of Biochemistry 75: 69-92. doi:10.1146/annurev.biochem.75.103004.142539

71. Steuber J, Schmid C, Rufibach M, Dimroth P (2000) Na+ translocation by complex I (NADH:quinone oxidoreductase) of Escherichia coli. Molecular Microbiology 35: 428-434. doi:10.1046/j.1365-2958.2000.01712.x

72. Yagi T (1991) Bacterial NADH-quinone oxidoreductases. Journal of Bioenergetics and Biomembranes 23: 211-225. doi:10.1007/BF00762218

64 73. Wang G, Maier RJ (2004) An NADPH Quinone Reductase of Helicobacter pylori Plays an Important Role in Oxidative Stress Resistance and Host Colonization. Infect. Immun. 72: 1391-1396. doi:10.1128/IAI.72.3.1391-1396.2004

74. Petrussa E, Bertolini A, Casolo V, Krajňáková J, Macrì F, et al. (2009) Mitochondrial bioenergetics linked to the manifestation of programmed cell death during somatic embryogenesis of Abies alba. Planta 231: 93-107. doi:10.1007/s00425-009-1028-x

75. Chomova M, Racay P (2010) Mitochondrial complex I in the network of known and unknown facts. General Physiology and Biophysics 29: 3-11. doi:10.4149/gpb_2010_01_3

76. Pilione MR, Harvill ET (2006) The Bordetella bronchiseptica type III secretion system inhibits gamma interferon production that is required for efficient antibody-mediated bacterial clearance. Infect Immun 74: 1043-9.

77. Kirimanjeswara GS, Mann PB, Pilione M, Kennett MJ, Harvill ET (2005) The complex mechanism of antibody-mediated clearance of Bordetella from the lungs requires TLR4. J Immunol 175: 7504-11.

78. Cotter PA, Yuk MH, Mattoo S, Akerley BJ, Boschwitz J, et al. (1998) Filamentous hemagglutinin of Bordetella bronchiseptica is required for efficient establishment of tracheal colonization. Infect Immun 66: 5921-9.

79. Stockbauer KE, Foreman-Wykert AK, Miller JF (2003) Bordetella type III secretion induces caspase 1-independent necrosis. Cellular Microbiology 5: 123-132. doi:doi:10.1046/j.1462-5822.2003.00260.x

80. Fennelly NK, Sisti F, Higgins SC, Ross PJ, van der Heide H, et al. (2008) Bordetella pertussis Expresses a Functional Type III Secretion System That Subverts Protective Innate and Adaptive Immune Responses. Infect. Immun. 76: 1257-1266. doi:10.1128/iai.00836-07

81. Suarez G, Sierra JC, Erova TE, Sha J, Horneman AJ, et al. (2010) A Type VI Secretion System Effector Protein, VgrG1, from Aeromonas hydrophila That Induces Host Cell Toxicity by ADP Ribosylation of Actin. J. Bacteriol. 192: 155-168. doi:10.1128/JB.01260- 09

82. French CT, Panina EM, Yeh SH, Griffith N, Arambula DG, et al. (2009) The Bordetella T3SS effector BteA contains a conserved N-terminal motif that guides bacterial virulence factors to lipid rafts. Cell Microbiol 11: 1735-1749. doi:10.1111/j.1462-5822.2009.01361.x

83. Ming Huam Y, Eric TH, Peggy AC, Jeff FM (2000) Modulation of host immune responses, induction of apoptosis and inhibition of NF-κB activation by the Bordetella type III secretion system. Molecular Microbiology 35: 991-1004.

65

Chapter 3 : CD1d Manipulation by the Bordetella bronchiseptica Type VI Secretion

System Modulates Antibody Production and Persistence

66

Abstract

Manipulating protein antigen presentation appears to be a common virulence mechanism

amongst bacterial pathogens, but modulating lipid antigen presentation by a bacterial pathogen

has not been documented. Altering communication between antigen presentation cells (APCs)

and adaptive immune cells can have significant effects on bacterial clearance, persistence, and

the overall immune response. Recently, the B. bronchiseptica Type VI Secretion System (T6SS) was found to contribute to persistence during murine respiratory tract infections. Using a mutant deficient in the T6SS, we correlated lower respiratory tract persistence with the ability to dampen both T cell dependent and independent antibody production. By monitoring dendritic cell, macrophage, and B cell accumulation over time, we determined that the T6SS was not responsible for cell trafficking modulation. However, MHC-II and CD1d surface expression in bone marrow derived dendritic cells was dampened by the T6SS in vitro, and dendritic cell,

macrophage, and B cell CD1d expression was found to be decreased by the T6SS in vivo.

Natural killer T cell accumulation in the lungs was also dampened by the T6SS, and CD1d was

required for the T6SS-mediated dampening of antibody production. Lastly, we determined that

the T6SS is required for dendritic cell death, suggesting that the T6SS dampens CD1d expression

on APCs by mediating more cell death, resulting in decreased NK T cell accumulation at the site

of the infection and overall decreased antibody titers. This is the first work that examines CD1d

manipulation by a bacterial pathogen and is the first example of antigen presentation modulation

by the T6SS.

67 Introduction

Bordetella bronchiseptica is a respiratory pathogen that causes disease in a wide range of both human and nonhuman mammals and is closely related to Bordetella pertussis and

Bordetella parapertussis, the causative agents of human whooping cough [1]. B. bronchiseptica provides an excellent natural host infection system in which to elucidate interactions between pathogens and host immunity [2–4]. Our laboratory has previously shown that efficient antibody production and antibody-mediated phagocytosis are required to clear B. pertussis, B. parapertussis, and B. bronchiseptica from the lungs [5–7].

The three classical bordetellae share multiple virulence determinants that modulate antibody responses, including dermonecrotic toxin (Dnt), adenylate cyclase toxin (CyaA), and a

Type III Secretion System (T3SS) (Bp and Bb only), although much of the work investigating adaptive immune response modulation has been done in B. bronchiseptica [8,9]. Purified B. bronchiseptica Dnt has been shown to decrease antibody production by inducing spleenic atrophy [10]. CyaA alone can inhibit macrophage activation by modulating CD40 and CD86 surface expression, leading to downstream adaptive immune effects, and was found to act with the T3SS to modulate a macrophage response that inhibits CD4+ T cell activation [11]. The B. bronchiseptica T3SS was found to affect dendritic cell maturation and migration and decrease antibody production, leading to increased clearance from the lower respiratory tract [12,13].

Furthermore, CyaA, T3SS, and the newly discovered Type VI Secretion System (T6SS) have been found to cause cell death in antigen presentation cells, further affecting how well innate cells can communicate with the adaptive immune system during infection [14–16]. However, the mechanisms behind how individual virulence factors directly manipulate antibody production

68 or how phagocytic cell modification leads to changes in antibody production have not been

determined [17].

Numerous pathogens have been found to regulate antigen presentation, activation, and cell maturation as a means of curbing bacterial clearance. Modulation of Major

Histocompatibility Complex I or II (MHC-I or MHC-II) by numerous bacterial pathogens is a common means of manipulating host cell activation. For example, Salmonella, Yersinia, and

Chlamydia have all be shown to manipulate MHC-I, by either impairing synthesis or expression of MHC-I [18–20]. Francisella, Mycobacterium, Chlamydia, Escherichia, Vibrio, and

Salmonella have all been found to modulate MHC-II expression, cell surface trafficking, or synthesis [21–23]. In Salmonella, Type III Secretion System (T3SS) effectors have been implicated in MHC II regulation through ubiquitination [21,24]. By regulating antigen presentation, pathogens attenuate T-helper cell activation, decreasing T cell dependent antibody production and interfering with T-helper cell-mediated activation of macrophages required for pathogen clearance. Although purified CyaA can modulate surface expression of activation makers, such as CD40 and CD86, Bordetella have not been shown to modulate antigen presentation.

Similar to peptide presentation by classical MHC molecules, lipid antigen is presented to natural killer T cells (NK T cells) via surface CD1d on APCs. NK T cells produce both anti- and

pro-inflammatory cytokines, including Interferon-γ (IFN-γ) [25–27]. Interaction between CD1d

and NK T cells can enhance B cell activation, both directly and indirectly by increasing co-

stimulatory activation of APCs, often increasing antibody production [28]. CD1d has been

shown to be required for efficient antibody production against pneumoncoccal polysaccharides

and during malaria infection [28,29]. Additionally, B cell CD1d expression was found to be

69 required for the NKT cell mediated enhancement of antibody production, suggesting that surface

expression of CD1d on both APCs and B cells is required for efficient antibody production [30].

To our knowledge, CD1d modulation by a bacterial pathogen and its contributions to Bordetella-

specific antibody production has not yet been observed. Here, we hypothesized that B.

bronchiseptica may manipulate antigen presentation and ultimately modulate the ability of

phagocytic cells to communicate with adaptive immunity.

In this study, we investigate the mechanism behind B. bronchiseptica T6SS-dependent

persistence in the lower respiratory tract. Using a B. bronchiseptica mutant lacking clpV,

RB50ΔclpV, which encodes the apparent T6SS ATPase that is required for secretion in well characterized T6SSs, we observed that the ClpV is required to dampen T cell-dependent and - independent production of antibodies, which are required for clearance. Because APC recruitment and accumulation did not appear to be affected by ClpV, we investigated T6SS- dependent effects on antigen presentation machinery as a probable mechanism to dampen antibody responses. ClpV was required to block MHC-II surface expression in vitro and late

CD1d surface expression both in vitro and in vivo, as well as dendritic cell death. Additionally,

CD1d expression modulation correlated with decreased numbers of NK T cells at the site of infection, and NK T cells were found to be required for ClpV-dependent manipulation of antibody production. Together, these data indicate that the T6SS is required to manipulate CD1d expression on APCs, significantly effecting antibody production and bacterial persistence.

70 Materials and Methods

Bacterial strains and growth.

B. bronchiseptica strain RB50 and RB50ΔclpV have been described elsewhere [31,32].

Bacteria were maintained on Bordet-Gengou agar (Difco) supplemented with 10% sheep blood

(Hema Resources) with 20 µg/ml streptomycin (Sigma). Bacteria were grown in Stainer-Scholte

(SS) broth at 37°C while shaking in liquid culture to mid-log phase [33].

Animal experiments.

Wild type C57BL/6 mice were obtained from Jackson Laboratories, Bar Harbor, ME.

Mice deficient in both NK T cells and CD1d (CD1d-/-) mice were a kind gift from Dr. Mark

Exley at Harvard University [34]. Mice were bred and maintained at The Pennsylvania State

University, University Park, PA in a specific pathogen-free facility, and all experiments were

carried out in accordance with institutional guidelines. Animal experiments were done as

previously described [5,31,35]. Briefly, the number of bacterial colony forming units (CFU) in

liquid cultures was calculated based on the optical density measured by absorbance of light at

600 nm. Bacteria were then diluted to 107 CFU/ml in sterile phosphate buffered saline (PBS).

Inocula were confirmed by plating dilutions on BG agar and counting the resulting colonies after

two days of growth at 37°C. For inoculation, mice were sedated with 5% isoflurane (IsoFlo,

Abbott Laboratories) in oxygen, and 50µl PBS containing the indicated CFU of bacteria was

gently pipetted onto the external nares. For bacterial quantification, mice were euthanized via

CO2 inhalation, and the designated organs were excised. Blood was obtained by cardiac

puncture. Tissues were homogenized in PBS, serially diluted, and plated onto BG agar plates

with 20 µg/mL streptomycin, and colonies were counted after 2 days of growth at 37°C.

Serum Antibody Titers.

71 Serum antibody titers were measured by indirect ELISA, as previously described [36].

Briefly, 96 well titer plates were coated with 5x105 CFU heat killed B. bronchiseptica per well

o and fixed overnight at 4 C in buffer containing 5.3g Na2CO3 , 4.2g NaHCO3 and 1g NaN3 per liter at pH 9.6. Serum obtained from infected mice was added to titer plates at 50:1 dilution in

PBS-T containing 5% non-fat dry milk. Serial 1:2 dilutions were made and antibody bound to antigen was detected by horseradish peroxidase (HRP) conguated anti- Ig, IgG1, IgG2a, IgG2b, or IgM antibodies. 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) in phosphate

citrate buffer and H2O2 was used to determine the concentration of secondary antibody present in

the wells. OD405 was measured 15 minutes after addition of chromogen. Cutoff values for a positive reaction were determined as values greater than the mean + 3 x standard deviation of control wells not probed with serum. An endpoint titer was estimated using a point slope formula as follows: Dilution factor below cutoff x (Dilution factor below cutoff - Dilution factor above cutoff)/(OD Above cutoff – OD below cutoff)x(ODCutoff - OD below cutoff).

Flow cytometry.

Approximately 2 million lymph node and lung cells were prepared by mechanical separation and centrifugation, as previously described [37,38]. Briefly, washed cells in PBS containing 2% FBS were added to 96 well titer plates. Fc block was performed using Rat anti-

Mouse anti-CD32/CD16 ( BD Bioscience) at a 200:1 dilution in 50µl for 15mins at 4oC. Next,

cells were stained with CD11c – perCp+ (200:1), B220- FITC+ (200:1)+, Pe – MHCII+ (4000:1),

CD45-V500+ (200:1), CD40-APC+ (500:1), CD1d-APC+ (800:1), CD1d-Tetramer-Pe (1000:1),

CD86-AF700* (100:1), CD70-Biotin* (400:1), and Streptavidin-Pe-Cy7* (500:1) (+ =

BDbiosciences; *=Biolegend) in 50µl for 30 minutes at 4oC. Cells were washed and fixed in 2%

paraformaldehyde with 1% FBS for 20 minutes and then analyzed on a BD Fortessa Flow

72 Cytometer at The Pennsylvania State University Microscopy and Cytometry Facility in

University Park, PA.

Bone Marrow Derived Dendritic Cells Stimulation.

Bone marrow derived dendritic cells were isolated from wild type CD57BL/6 mice as

previously described [39]. Briefly, CD57BL/6 bone marrow was extracted and cultured in vitro

for eight days at 37°C in 5% CO2. Next, cells were stimulated with either RB50 or RB50ΔclpV in PBS for 2 or 6 hours at multiplicities of infection (MOI) of 0.1, 1 or 10. After incubation, cells were either analyzed for cytotoxicity or stained for flow cytometry analysis.

Cytotoxicity assay.

Cytotoxicity assays were preformed as previously described [31,40,41]. Briefly, J774.2 murine macrophage cells (ATCC) were cultured in Dulbecco’s modified Eagle’s medium

(DMEM, Difco) supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, 1% nonessential amino acids, and 1% sodium pyruvate. The cells were grown in 96-well plates

(Greiner Bio-One) to 85% confluency in 5% CO2 at 37°C. RPMI media lacking phenol red with

5% fetal bovine serum, 1% L-glutamine, 1% nonessential amino acids, and 1% sodium pyruvate

was used to replace DMEM at least one hour prior to the experiment. Bacteria diluted in PBS at

MOIs of 0.1, 1, and 10 were added to the macrophages, and then the cells and bacteria were

centrifuged at 300 x g for 5 minutes and incubated in 5% CO2 at 37°C for 2, 4, or 6 hours. After

incubation, lactate dehydrogenase (LDH) release, a measure of cytotoxicity, was analyzed by

collecting cell supernatants and using a Cytotox96 kit (Promega) according to the instructions of

the manufacturer.

73 Results

T6SS enables persistence and dampens antibody production in vivo.

Our lab previously found that the B. bronchiseptica T6SS contributes to macrophage cell

death, cytokine production, and persistence in vivo (Chapter 2). Because a mutant lacking T6SS

function, RB50ΔclpV, was cleared more quickly from

the lower respiratory tract than the parental wild type

strain, RB50, we hypothesized that the T6SS mutant

may induce higher antibody titers, similar to what was

previously observed with the T3SS mutant [17]. To

investigate this hypothesis, wild type C57BL/6 mice

were inoculated with 5×105 CFU of RB50 or

RB50ΔclpV. By day 7 post-inoculation, numbers of

wild type RB50 recovered from the lungs were ~100-

fold higher than that of RB50ΔclpV, and the mutant

was ultimately cleared from the lower respiratory tract

by day 28 post-inoculation, as previously observed

Figure 3-1 Lung colonization and antibody (Figure 1A) [42]. To determine if increased antibody titers in RB50 vs RB50ΔclpV infected mice. A-B. Mice were inoculated with 5×105 CFU of RB50 vs RB50ΔclpV, and the lungs and blood production contributed to the rapid clearance of were excised at designated time points to enumerate the bacteria. A) RB50 (♦) vs RB50ΔclpV, blood was obtained from each animal RB50ΔclpV(▲) colonization was monitored on days 3, 7, 14, and 28 post-inoculation. B.) Anti- throughout the time course, and anti-B. RB50 serum IgG titers were determined by ELISA from mice infected with either RB50 (dark bars) or RB50ΔclpV (light bars) on days bronchiseptica serum IgG antibody titers were 14 and 28 post-inoculation. Statistical significance was determine by comparing each analyzed by ELISA. Although there were fewer time point independently via a Student’s t-test, and * denotes p value greater than 0.05. O. Y. Rolin completed the titer ELISAs.

74 RB50ΔclpV bacteria in the lower respiratory tract, the mutant induced much higher anti-B.

bronchiseptica IgG titers than RB50 infected mice on days 14 and 28 post inoculation (Figure

1B). These results suggest that the T6SS is capable of dampening antibody production in vivo as

a mechanism to facilitate persistence in the lower respiratory tract.

T6SS dampens both T cell-independent and -dependent antibody production.

To determine which step in the pathway to antibody production the T6SS modulates, we

analyzed serum titers to determine if T cell-

independent (IgM) or -dependent antibody

(IgG2a) isotypes were differentially affected.

* Both IgG2a and IgM titers were higher in mice

inoculated with RB50ΔclpV compared to RB50

at days 14 and 28 post-inoculation, indicating

that the T6SS affects both T cell-independent

and -dependent antibody production (Figure

2A and 2B). These data suggest that the T6SS * may modulate part of the innate immune

system, such as macrophages or dendritic cells,

Figure 3-2 IgM and IgG2a titers from mice infected which then can have downstream effects on with either RB50 or RB50ΔclpV. A-B. Serum was taken from mice inoculated with either both T cells and B cells. 5×105 CFU of RB50 (♦) or RB50ΔclpV (■), and serum IgM (A) or IgG2a (B) was determined for mice infected with at 14 and 28 days post-inoculation. Statistical T6SS does not affect APC recruitment or B significance was determine by comparing each time point independently via a Student’s t-test, and * denotes p value greater than 0.05. O. Y. Rolin completed these cell expansion. experiments.

One potential mechanism to modulate

75 antibody production is to manipulate delivery of bacterial antigen to lymph nodes, where T cell

and B cell activation occurs. To investigate whether or not the T6SS can modulate recruitment of

infected APCs carrying antigen to the

Figure 3-3 Immune cells during RB50 or RB50ΔclpV infection over time in the cervical lymph node, mediastinal lymph node, or lungs. A-I. Dendritic cells (CD11c+), macrophages (F4/80+), or B cells (B220+) were identified by flow cytometry from the cervical lymph node (CLN) , mediastinal lymph node (MLN), and lungs in mice infected with either RB50 (♦) or RB50ΔclpV(■). In the CLN, MLN, and lungs, dendritic cells (A-C), macrophages (D-F), and B cells (G-I) were observed on days 2, 7, and 14 post-inoculation. * denotes p value greater than 0.05. O. Y. Rolin completed these experiments.

draining lymphoid tissues, we inoculated C57BL/6 mice with either RB50 or RB50ΔclpV and

monitored cell expansion and lymphoid recruitment of alveolar macrophages, dendritic cells, and

76 B cells on days 2, 7, and 14 post-inoculation. Mice inoculated with RB50 and RB50ΔclpV had

similar numbers of CD11c+ cells in the lungs, cervical lymph nodes, and mediastinal lymph

nodes (Figure 3A-C). Macrophages (F480+) recovered from RB50-infected lymphoid tissue generally decreased over time, similar to RB50ΔclpV, from the lymphoid tissues (Figure 3D-E),

but were higher in the RB50 infected lungs on day 7 post-inoculation in comparison to

RB50ΔclpV (Figure 3F). These observations demonstrate that the T6SS does not affect APC

accumulation over time. B cell (B220+) numbers generally increased overtime in the lymphoid

tissue and lungs (Figure 3G-I). Interestingly, B cells increased in cervical lymph nodes of

RB50ΔclpV infected mice compared to wild type on day 7 post-inoculation, but this increase in

B cell numbers was not observed at any time point or in the lungs or MLNs, indicating that the

T6SS does not affect general B cell proliferation (Figure 3G). From these data, we conclude that the T6SS does not inhibit antibody production by blocking APC trafficking to lymphoid tissues or B cell expansion during infection.

T6SS modulates MHC-II and CD1d expression on CD11c+ cells in vitro.

77 Because the T6SS did not affect APC trafficking is not affected by the T6SS, we hypothesized that B. bronchiseptica may use this secretion system to modify peptide (MHC-II)

or lipid (CD1d)

presentation machinery.

To test this hypothesis,

bone marrow derived

dendritic cells (BMDCs)

were exposed to either

RB50 or RB50ΔclpV for

Figure 3-4 MHC-II and CD1d surface expression on BMDCs at 2 and 6 hours. either 2 or 6 hours, and A-D. BMDCs obtained from C57BL/6 mice were cultured and inoculated with either RB50 (dark grey), RB50ΔclpV (light grey), or PBS (black) as a control. MHC-II or CD1d surface MHC-II mean fluorescence intensity (MFI) was analyzed by flow cytometry on CD11c+ cells after two hours incubation (A) or six hours incubation (C). CD1d MFI was additionally analyzed after two hours (B) or six hours (D) incubation on expression was CD11c+ cells. Statistical significance was determine by comparing each time point independently via a Student’s t-test, and * denotes p value greater than 0.05. monitored as a function of mean fluorescence intensity (MFI). Although, neither RB50 nor RB50ΔclpV was able to stimulate MHC-II surface expression above that of naïve BMDCs after 2 hours of stimulation at a MOI of 0.1, an MOI of 1 or 10 of RB50ΔclpV did expressed more MHC-II surface molecules than cells infected with RB50, suggesting that the T6SS can block early MHC-II surface expression (Figure 4A). Surprisingly, both RB50 and RB50ΔclpV induced a dose-dependent increase in CD1d surface expression compared to naïve cells, but no T6SS-mediated difference in CD1d expression was observed at any MOI (Figure 4B). At six hours, MHC-II surface expression was again higher in cells infected with RB50ΔclpV, compared to wild type, at MOIs of 1 and 10. Surprisingly, RB50 was decreased MHC-II surface expression when compared to naïve cells or the T6SS mutant, suggesting that the T6SS may decrease MHC-II expression

78 (Figure 4C). At higher MOIs after 6 hours post-inoculation, increased CD1d surface expression

was observed in RB50ΔclpV infected cells compared to those infected with RB50, indicating that

the T6SS is also capable of dampening murine BMDCs CD1d surface expression (Figure 4D).

Together, these data suggest that the T6SS contributes to both MHC-II and CD1d modulation in

vitro.

CD1d expression is dampened by the T6SS in vivo.

To determine whether the T6SS modulates MHC-II and CD1d expression in vivo,

C57BL/6 mice were inoculated with 5×105 CFU of RB50 or RB50ΔclpV, and surface expression

Figure 3-5 CD1d surface expression on immune cells in the cervical lymph nodes or lungs. A-F. The mean fluorescence intensity (MFI) of CD1d on dendritic cells (DC) (CD11c+), macrophages (MФ) (F4/80+), or B cells (B220+) were identified by flow cytometry in mice infected with either RB50 (dark bars), RB50ΔclpV (light bars), or PBS (medium bars). CD1d MFI was analyzed in the CLN or lungs 2 days (A-B), 7 days (C-D), or 14 days (E-F) post-inoculation. Statistical significance was determine by comparing each time point independently via a Student’s t-test, and * denotes p value greater than 0.05. O. Y. Rolin completed these experiments.

79 on alveolar macrophages, dendritic cells, and B cells was monitored 2, 7, and 14 days post- inoculation in the cervical lymph nodes, mediastinal lymph nodes, and lungs. Surprisingly, no

T6SS-mediated changes in MHC-II expression were observed throughout this time course in any cell type (data not shown). Furthermore, no T6SS-mediated differences were observed on dendritic cells, alveolar macrophages, or B cells in the cervical lymph nodes or lungs at 2 days post-inoculation (Figure 5A-B). However, CD1d expression was significantly increased in

RB50ΔclpV infected dendritic cells and B cells present in the cervical lymph node, but not the lungs, by 7 days post-inoculation (Figure 5C-D). Strikingly, a trend of increasing CD1d expression on DCs, macrophages, and B cells in both the cervical lymph node and lungs in

RB50ΔclpV infected mice was observed 14 days post-inoculation, and statistical differences were observed in cervical lymph node DCs and alveolar macrophages and lung macrophages

(Figure 5E-F). No differences in CD1d surface expression between RB50 and RB50ΔclpV infected mice were observed in the mediastinal lymph node (data not shown), suggesting this phenomena only occurs at the site of infection and in the cervical lymph nodes. Together, these data indicate that the T6SS is capable of dampening late CD1d expression on dendritic cells, macrophages, and B cells in vivo, which may have deleterious effects on antibody production downstream.

NK T cell recruitment is dampened by T6SS.

CD1d-/- mice are deficient in NK T cell function and lipid antigen presentation, which is

required to enhance B cell stimulation and antibody production [28,30,43–45]. To test the

hypothesis that the T6SS modulation of CD1d expression affects NK T cell recruitment or

proliferation, we monitored NK T cells in the cervical lymph nodes, mediastinal lymph nodes, and lungs at 7 and 14 days post-inoculation of C57BL/6 mice infected with either RB50 or

80 RB50ΔclpV. There was no difference in the number of NK T cells in the cervical lymph node or

mediastinal lymph node at either time point (data not shown). However, significantly higher NK

T cell numbers were observed in lungs of RB50ΔclpV infected mice on days 7 and day 14 post-

inoculation (Figure 6A). These data suggest that the T6SS is capable of modulating NK T cell

recruitment or expansion at the site of infection, which correlates with the ability of the T6SS to

dampen CD1d surface expression, suggesting a potential mechanism behind how the T6SS

mediates antibody production.

CD1d is required for T6SS-mediated antibody production modulation.

NK T cell activation through CD1d results in downstream B cell stimulation and

antibody production [43,45,46]. Because we observed that the T6SS manipulates both CD1d

surface expression and NK T cell proliferation

at the site of infection, we hypothesized that

these observations may contribute to T6SS-

mediated dampening of antibody production.

To test this, we monitored the antibody

response in CD1d deficient mice (CD1d-/-)

during either RB50 or RB50ΔclpV infection.

Figure 3-6 NK T cell numbers in the lungs and antibody Although the overall serum IgG response to production in NK T cell deficient mice. A-B. A.) Flow cytometry was used to monitor NK T cell RB50 infection was lower in CD1d-/- mice numbers via a CD1d tetramer in mice infected with either RB50 (dark bars) or RB50ΔclpV (light bars) on 7 and 14 days post-inoculation. B.) Total serum Ig titers were than in wild type C57BL/6 mice, as expected, monitored 28 days post-inoculation in CD1d-/- mice -/- infected with either RB50 (dark bars) or RB50ΔclpV RB50ΔclpV infected CD1d mice had serum (light bars). * denotes p value greater than 0.05. N. T. Jacobs performed the titer ELISAs. titers equal to RB50 infected mice on day 28

post-inoculation (Figure 6B). The higher antibody titers induced by RB50ΔclpV in wild type

81 mice (Figure 1B and 1C) were not observed in CD1d-/- mice, indicating that CD1d is required for

the T6SS-mediate manipulation of antibody production.

T6SS mediates dendritic cell cytotoxicity.

We previously established that the T6SS is required to kill murine macrophages, but dendritic cell T6SS-mediated death was never measured [42]. Because we observed that CD1d

is required for the T6SS-mediated manipulation of antibody production, we hypothesized that B.

bronchiseptica might use the T6SS to kill dendritic cells and macrophages, leading to a decrease

in the number of activated APCs throughout an

infection. To determine whether the T6SS also

enabled B. bronchiseptica to kill dendritic cells,

we cultured BMDCs from naïve C57BL/6 mice

and exposed them to either RB50 or RB50ΔclpV

Figure 3-7 In vitro BMDC cytotoxicity caused by at an MOI of 1 for 2 hours. LDH release was either RB50 or RB50ΔclpV. Lactate dehydrogenase (LDH) from BMDCs was then assayed as a proxy for dendritic cell death. measured from cell supernatants as a proxy for cell death. Cells were incubated with RB50 (dark bars) or RB50ΔclpV (light bars) for two hours before cell Neither strain produced significant cytotoxicity supernatants were obtained and LDH was monitored. Statistical significance was determine by comparing at an MOI of 0.1 (Figure 7). However, while each time point independently via a Student’s t-test, and * denotes p value greater than 0.05. RB50 was capable of inducing BMDC

cytotoxicity at an MOI of 1, RB50ΔclpV induced lower levels of cytotoxicity at an MOI of 1

(Figure 7). Even at an MOI of 10, RB50ΔclpV was only capable of inducing ~30% cytotoxicity,

still much lower than levels induced by RB50, suggesting that the T6SS contributes to B.

bronchiseptica killing of dendritic cells (Figure 7). The T6SS was also required for efficient

cytotoxicity at 4 and 6 hours post-inoculation at an MOI of 1 (data not shown). These data

82 suggest that the T6SS mediates phagocytic killing, and therefore may explain why RB50ΔclpV infected cells appear to have increased CD1d expression in vivo and in vitro.

83

Discussion

In this study, a newly discovered B. bronchiseptica virulence factor was determined to

manipulate antigen presentation, which can consequently affect antibody production. The T6SS

was again shown to be required for persistence within the lower respiratory tract in vivo, but this

study determined that dampened antibody production correlated with this phenomenon. To

investigate the mechanisms behind T6SS-mediated modulation of antibody production, the ability of the T6SS to modulate antigen presentation machinery was examined, elucidating that the T6SS can manipulate MHC-II surface expression in vitro and CD1d surface expression both in vitro and in vivo. Additionally, the T6SS was found to decrease NK T cell numbers at the site of infection, potentially by reducing CD1d surface expression on APCs. Lastly, CD1d was shown to be required for the T6SS-dependent differences in serum antibody production.

Together, these data suggest that T6SS diminishes antibody production by reducing CD1d surface expression and NK T cell accumulation at the site of infection.

Our initial data indicating that T cell independent and dependent antibody production were equally affected suggested that a T6SS-mediated mechanism affecting antibody production may have been upstream of T cell stimulation (Figure 2A-D), indicating that we should investigate effects on antigen presentation cells. In the end, this work uncovered a T cell- dependent mechanism that requires NK T cells to manipulate antibody production and may not explain how T cell independent antibody production is affected by the T6SS. We previously showed that the B. bronchiseptica T6SS is required for IL-1β, IL-6, and IL-17 production in vitro, as well as inhibition of IFN-γ production in vivo (cite), suggesting that a different innate immune stimulation may account for changes in T cell independent antibody production.

84 Additional work investigating how T6SS-dependent cytokine modulation affects antibody production is necessary to understand additional mechanisms behind how the T6SS manipulates adaptive immunity.

Although viral models suggest that CD1d modulation during infection can occur [47,48], this is the first account of CD1d surface expression manipulation by a bacterial pathogen and is also the first account of a T6SS modulating any antigen presentation machinery. The data presented here suggest a mechanistic model explaining how the T6SS can reduce antibody production by manipulating CD1d. Upon infection, B. bronchiseptica stimulates cytokine release, causes pathology, and recruits immune cells to the site of infection (cite). Recruited phagocytic cells, including macrophages and dendritic cells, can either be killed by B. bronchiseptica via various toxins or secretion systems, such as CyaA, T3SS, or T6SS, or engulf and kill the bacteria, become activated, and begin to present bacterial antigens. During this stage of infection, MHC-II and CD1d surface expression can be manipulated by the T6SS, although the molecular mechanisms underlying this observation remain unclear. Reduced transcription of

CD1d encoding genes, protein ubiquitination and degradation, or direct modulation by protein binding during surface maturation are all possible means for achieving this end. Ultimately, reduction of CD1d expression may interfere with communications between APCs and NK T cells, decreasing their ability to produce cytokines that stimulate B cells and antibody production. The T6SS also decreased B cell CD1d surface expression. This observation suggests an additional mechanism by which NK T cell-mediated activation of B cells could be dampened, because B cell CD1d expression has been shown to be required for NK T cell stimulation. Modulating CD1d surface expression on APCs and B cells provides two

85 mechanisms in which the T6SS may manipulate antibody production by interfering with NK T

cell-mediated stimulation.

To understand how the B. bronchiseptica T6SS manipulates MHC-II and CD1d, we

compared the results presented here with what is known in other bacterial pathogenesis models.

In this study, we observed that the T6SS was capable of modulating MHC-II expression in vitro

within 2 hours post-inoculation. B. pertussis ACT has been shown to stimulate MHC-II

expression on naïve dendritic cells after 24 hours of exposure, and similarly the B.

bronchiseptica T3SS was shown to induce MHC-II expression after 18 hours [49,50]. This study

monitors MHC-II expression very early during exposure (2 and 6 hours), and this variation in

experimental design could potentially explain why we observe dampening of MHC-II expression

in vitro by RB50. Although understanding how the T6SS can modulate MHC-II is interesting, these T6SS-mediate effects were not observed during infection in this study, so they were not pursued as a mechanism by which the T6SS manipulates antibody production in vivo. Secondly, we observed that the T6SS manipulated CD1d expression in vitro within six hours, so we expected that manipulation of CD1d surface expression on APCs would have been observed as early as day 2 post-inoculation in vivo, similar to what has been observed with MHC-II manipulation by Salmonella [24]. However, manipulation of CD1d surface expression in vivo was not observed until 7 days post-inoculation and was even more prominent 14 days post- inoculation. These data suggest that the T6SS-dependent effects on CD1d occur subsequent to initial APC lymph node migration and may continue throughout the infection. It remains unclear why manipulation of CD1d expression early during infection is not observed, although we hypothesize that this phenomena becomes more prominent in APCs that have survived contact with B. bronchiseptica. Similar to Pseudomonas, Vibrio, and Aeromonas, the Bordetella T6SS

86 appears to kill host macrophages and dendritic cells, suggesting that RB50ΔclpV infected cells may have more time to become increasingly activated [51]. Furthermore, in Burkholderia,

Salmonella, Francisella and Edwardsiella species, the T6SS is required for intracellular growth

and survival, suggesting that the Bordetella T6SS may also contribute to intracellular survival,

changing the host cell activation status only after bacteria have proliferated and survived for a

lengthy period intracellularly [51].

Beyond the T6SS, B. bronchiseptica possesses other virulence factors that contribute to

host cell death, namely ACT and the T3SS [15,35]. Harvill et al. showed that ACT is required

for macrophage cytotoxicity, and Yuk et al. have shown that the Bordetella T3SS is capable of

attenuating macrophage activation and causing cell death [12,13,17]. Although each of these

virulence factors also have different effects on cytokine production, cell migration, and

persistence, they share the ability to cause cell death and, in the case of the T3SS and T6SS,

dampen antibody production in vivo [17]. This suggests that virulence factors contributing to

cell death may also dampen CD1d surface expression and late cell activation in vivo, by simply

decreasing APC survival. Other pathogens may share this strategy, as numerous secretion

systems and toxins contribute to APC cell death in various model systems. One example may be

Francisella tularensis, which encodes several mechanisms by which to cause cell death, but also

is capable of decreasing MHC-II surface expression through a ubiquitin-dependent mechanism

[22,52].

Here, we have investigated how the T6SS can contribute to the manipulation of antigen

presentation machinery during infection, and we show the first example of how bacterial

pathogens can manipulate the lipid presenting molecule, CD1d. We show that CD1d modulation

in vivo is not observed early during infection, but rather after day 7 post-inoculation. Lastly, we

87 present show strong evidence supporting that CD1d cell surface expression can affect antibody

production in vivo. Further investigation to understand the molecular mechanisms behind how secreted T6SS effectors cause cell death and therefore modulate CD1d surface expression during infection may elucidate novel therapeutic mechanisms.

88

Author Contributions:

Authors: Laura S. Weyrich1,2*, Olivier Y. Rolin1,3*, Nathan T. Jacobs1, and Eric T. Harvill1

1Department of Veterinary and Biomedical Sciences, The Pennsylvania State University

115 Henning Building, University Park, PA 16802

2Graduate Program in Biochemistry, Microbiology, and Molecular Biology

3 Graduate Program in Immunology and Infectious Disease

* Authors contributed equally.

Conceived and designed experiments: LSW, OYR, ETH

Preformed experiments: LSW, OYR, NTJ

Analyzed data: LSW, OYR, ETH

Wrote manuscript: LSW, ETH

89

References

1. Mattoo S, Cherry JD (2005) Molecular Pathogenesis, Epidemiology, and Clinical Manifestations of Respiratory Infections Due to Bordetella pertussis and Other Bordetella Subspecies. Clin. Microbiol. Rev. 18: 326-382. doi:10.1128/CMR.18.2.326-382.2005

2. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, et al. (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35: 32-40. doi:10.1038/ng1227 ng1227 [pii]

3. Harvill ET, Cotter PA, Miller JF (1999) Pregenomic comparative analysis between bordetella bronchiseptica RB50 and Bordetella pertussis tohama I in murine models of respiratory tract infection. Infect Immun 67: 6109-18.

4. Cotter PA, Miller JF (1997) A mutation in the Bordetella bronchiseptica bvgS gene results in reduced virulence and increased resistance to starvation, and identifies a new class of Bvg-regulated antigens. Mol Microbiol 24: 671-85.

5. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of antibodies in immunity to Bordetella infections. Infect Immun 71: 1719-24.

6. Kirimanjeswara GS, Mann PB, Pilione M, Kennett MJ, Harvill ET (2005) The complex mechanism of antibody-mediated clearance of Bordetella from the lungs requires TLR4. J Immunol 175: 7504-11.

7. Kirimanjeswara GS, Agosto LM, Kennett MJ, Bjornstad ON, Harvill ET (2005) Pertussis toxin inhibits neutrophil recruitment to delay antibody-mediated clearance of Bordetella pertussis. J Clin Invest 115: 3594-601.

8. Mattoo S, Cherry JD (2005) Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev 18: 326-82.

9. Shrivastava R, Miller JF (2009) Virulence factor secretion and translocation by Bordetella species. Current Opinion in Microbiology 12: 88-93.

10. Horiguchi Y, Matsuda H, Koyama H, Nakai T, Kume K (1992) Bordetella bronchiseptica dermonecrotizing toxin suppresses in vivo antibody responses in mice. FEMS Microbiol. Lett. 69: 229-234.

11. Siciliano NA, Skinner JA, Yuk MH (2006) Bordetella bronchiseptica modulates macrophage phenotype leading to the inhibition of CD4+ T cell proliferation and the initiation of a Th17 immune response. J. Immunol. 177: 7131-7138.

90 12. Reissinger A, Skinner JA, Yuk MH (2005) Downregulation of mitogen-activated protein kinases by the Bordetella bronchiseptica Type III secretion system leads to attenuated nonclassical macrophage activation. Infect Immun 73: 308-16.

13. Skinner JA, Pilione MR, Shen H, Harvill ET, Yuk MH (2005) Bordetella type III secretion modulates dendritic cell migration resulting in immunosuppression and bacterial persistence. J Immunol 175: 4647-52.

14. Pilione MR, Harvill ET (2006) The Bordetella bronchiseptica type III secretion system inhibits gamma interferon production that is required for efficient antibody-mediated bacterial clearance. Infect Immun 74: 1043-9.

15. Buboltz AM, Nicholson TL, Weyrich LS, Harvill ET (2009) Role of the type III secretion system in a hypervirulent lineage of Bordetella bronchiseptica. Infect. Immun 77: 3969- 3977. doi:10.1128/IAI.01362-08

16. Mann P, Goebel E, Barbarich J, Pilione M, Kennett M, et al. (2007) Use of a Genetically Defined Double Mutant Strain of Bordetella bronchiseptica Lacking Adenylate Cyclase and Type III Secretion as a Live Vaccine. Infect Immun 75: 3665-72.

17. Yuk MH, Harvill ET, Cotter PA, Miller JF (2000) Modulation of host immune responses, induction of apoptosis and inhibition of NF-kappaB activation by the Bordetella type III secretion system. Mol Microbiol 35: 991-1004.

18. Kirveskari, He, Leirisalo‐Repo, Mäki‐Ikola, Wuorela, et al. (1999) Enterobacterial infection modulates major histocompatibility complex class I expression on mononuclear cells. Immunology 97: 420-428. doi:10.1046/j.1365-2567.1999.00803.x

19. Wuorela M, Jalkanen S, Kirveskari J, Laitio P, Granfors K (1997) Yersinia enterocolitica serotype O:3 alters the expression of serologic HLA-B27 epitopes on human monocytes. Infection and Immunity 65: 2060 -2066.

20. Zhong G, Fan P, Ji H, Dong F, Huang Y (2001) Identification of a Chlamydial Protease– Like Activity Factor Responsible for the Degradation of Host Transcription Factors. J Exp Med 193: 935-942.

21. Lapaque N, Hutchinson JL, Jones DC, Méresse S, Holden DW, et al. (2009) Salmonella regulates polyubiquitination and surface expression of MHC class II antigens. Proc. Natl. Acad. Sci. U.S.A. 106: 14052-14057. doi:10.1073/pnas.0906735106

22. Wilson JE, Katkere B, Drake JR (2009) Francisella tularensis induces ubiquitin-dependent major histocompatibility complex class II degradation in activated macrophages. Infect. Immun. 77: 4953-4965. doi:10.1128/IAI.00844-09

23. Harding CV, Ramachandra L, Wick MJ (2003) Interaction of bacteria with antigen presenting cells: influences on antigen presentation and antibacterial immunity. Current Opinion in Immunology 15: 112-119. doi:10.1016/S0952-7915(02)00008-0

91 24. Halici S, Zenk SF, Jantsch J, Hensel M (2008) Functional analysis of the Salmonella pathogenicity island 2-mediated inhibition of antigen presentation in dendritic cells. Infect. Immun. 76: 4924-4933. doi:10.1128/IAI.00531-08

25. Sugita M, Moody DB, Jackman RM, Grant EP, Rosat JP, et al. (1998) CD1--a new paradigm for antigen presentation and T cell activation. Clin. Immunol. Immunopathol. 87: 8-14.

26. Beckman EM, Porcelli SA, Morita CT, Behar SM, Furlong ST, et al. (1994) Recognition of a lipid antigen by CD1-restricted alpha beta+ T cells. Nature 372: 691-694. doi:10.1038/372691a0

27. Joyce S (2001) CD1d and natural T cells: how their properties jump-start the immune system. Cell. Mol. Life Sci. 58: 442-469.

28. Hansen DS, Siomos M, de Koning‐Ward T, Buckingham L, Crabb BS, et al. (2003) CD1d‐restricted NKT cells contribute to malarial splenomegaly and enhance parasite‐specific antibody responses. European Journal of Immunology 33: 2588-2598. doi:10.1002/eji.200323666

29. Kobrynski LJ, Sousa AO, Nahmias AJ, Lee FK (2005) Cutting Edge: Antibody Production to Pneumococcal Polysaccharides Requires CD1 Molecules and CD8+ T Cells. The Journal of Immunology 174: 1787 -1790.

30. Lang GA, Devera TS, Lang ML (2008) Requirement for CD1d expression by B cells to stimulate NKT cell–enhanced antibody production. Blood 111: 2158 -2162. doi:10.1182/blood-2007-10-117309

31. Buboltz AMNTL (2008) Enhanced TTSS function contributes to the hypervirulence of a Bordetella bronchiseptica lineage. Journal of Bacteriology Submitted.

32. Cotter PA, Miller JF (1994) BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. Infect Immun 62: 3381- 90.

33. Stainer DW, Scholte MJ (1970) A simple chemically defined medium for the production of phase I Bordetellla pertussis. J. Gen. Microbiol. 63: 211-220.

34. Exley MA, Bigley NJ, Cheng O, Shaulov A, Tahir SMA, et al. (2003) Innate immune response to encephalomyocarditis virus infection mediated by CD1d. Immunology 110: 519-526. doi:10.1111/j.1365-2567.2003.01779.x

35. Harvill ET, Cotter PA, Yuk MH, Miller JF (1999) Probing the function of Bordetella bronchiseptica adenylate cyclase toxin by manipulating host immunity. Infect Immun 67: 1493-500.

92 36. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of Antibodies in Immunity to Bordetella Infections. Infect. Immun. 71: 1719-1724. doi:10.1128/iai.71.4.1719-1724.2003

37. Zhang X, Goel T, Goodfield LL, Muse SJ, Harvill ET (2011) Decreased Leukocyte Accumulation and Delayed Bordetella pertussis Clearance in IL-6−/− Mice. The Journal of Immunology 186: 4895 -4904. doi:10.4049/jimmunol.1000594

38. Zhang X, Hester SE, Kennett MJ, Karanikas AT, Bendor L, et al. (2011) Interleukin-1 receptor signaling is required to overcome the effects of pertussis toxin and for efficient infection- or vaccination-induced immunity against Bordetella pertussis. Infect. Immun. 79: 527-541. doi:10.1128/IAI.00590-10

39. Lutz MB, Kukutsch N, Ogilvie AL., Rößner S, Koch F, et al. (1999) An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. Journal of Immunological Methods 223: 77-92. doi:10.1016/S0022- 1759(98)00204-X

40. Mattoo S, Yuk MH, Huang LL, Miller JF (2004) Regulation of type III secretion in Bordetella. Mol Microbiol 52: 1201-14.

41. Yuk MH, Harvill ET, Miller JF (1998) The BvgAS virulence control system regulates type III secretion in Bordetella bronchiseptica. Mol Microbiol 28: 945-59.

42. Weyrich, L.S., Rolin, O. Y., Muse, S. J., Park, J., Spidale, N., et al. (2012) A Type VI Secretion System Encoding Locus is Required for Bordetella bronchiseptica Immunomodulation and Persistence In Vivo. PLoS One Submitted.

43. Nishimura T, Kitamura H, Iwakabe K, Yahata T, Ohta A, et al. (2000) The interface between innate and acquired immunity: glycolipid antigen presentation by CD1d- expressing dendritic cells to NKT cells induces the differentiation of antigen-specific cytotoxic T lymphocytes. International Immunology 12: 987 -994. doi:10.1093/intimm/12.7.987

44. Kumar H, Belperron A, Barthold SW, Bockenstedt LK (2000) Cutting Edge: CD1d Deficiency Impairs Murine Host Defense Against the Spirochete, Borrelia burgdorferi. The Journal of Immunology 165: 4797 -4801.

45. Joyce S (2001) CD1d and natural T cells: how their properties jump-start the immune system. Cell. Mol. Life Sci. 58: 442-469.

46. Barral P, Eckl-Dorna J, Harwood NE, De Santo C, Salio M, et al. (2008) B cell receptor- mediated uptake of CD1d-restricted antigen augments antibody responses by recruiting invariant NKT cell help in vivo. Proceedings of the National Academy of Sciences 105: 8345 -8350. doi:10.1073/pnas.0802968105

93 47. Sanchez DJ, Gumperz JE, Ganem D (2005) Regulation of CD1d expression and function by a herpesvirus infection. Journal of Clinical Investigation 115: 1369-1378. doi:10.1172/JCI24041

48. Durante-Mangoni E, Wang R, Shaulov A, He Q, Nasser I, et al. (2004) Hepatic CD1d expression in hepatitis C virus infection and recognition by resident proinflammatory CD1d-reactive T cells. J. Immunol. 173: 2159-2166.

49. Ross PJ, Lavelle EC, Mills KHG, Boyd AP (2004) Adenylate Cyclase Toxin from Bordetella pertussis Synergizes with Lipopolysaccharide To Promote Innate Interleukin-10 Production and Enhances the Induction of Th2 and Regulatory T Cells. Infection and Immunity 72: 1568 -1579. doi:10.1128/IAI.72.3.1568-1579.2004

50. Skinner JA, Reissinger A, Shen H, Yuk MH (2004) Bordetella type III secretion and adenylate cyclase toxin synergize to drive dendritic cells into a semimature state. J Immunol 173: 1934-40.

51. Cascales E (2008) The type VI secretion toolkit. EMBO Rep 9: 735-41. doi:embor2008131 [pii] 10.1038/embor.2008.131

52. Wickstrum JR, Bokhari SM, Fischer JL, Pinson DM, Yeh H-W, et al. (2009) Francisella tularensis Induces Extensive Caspase-3 Activation and Apoptotic Cell Death in the Tissues of Infected Mice. Infection and Immunity 77: 4827 -4836. doi:10.1128/IAI.00246-09

94

Chapter 4 : Bordetella bronchiseptica Requires Cytotoxic T-cells to Displace Upper

Respiratory Tract Microflora during Infection

95

Abstract

Although respiratory pathogens are a leading cause of death worldwide, we know very little about how infectious bacterial pathogens overcome nasal cavity flora to cause disease.

Even less is known about how respiratory pathogens may exploit the immune system in order to out-compete established microorganisms. Using a zoonotic pathogen with a broad host range,

Bordetella bronchiseptica, we observed displacement of both culturable and sequencing-based identified organisms in the upper respiratory tract. In vivo species-to-species competition via both the Type VI Secretion System, Type III Secretion Systems, and the master regulatory system, BvgAS, explained much of this phenomenon; however, B. bronchiseptica was additionally found to require IL-17 and cytotoxic T-cells to displace resident flora. We propose that B. bronchiseptica-stimulated degranulation of cytotoxic T-cells results in displacement of all culturable organisms for up to 70 days, elucidating a novel mechanism in which bacterial pathogens can overcome established bacterial microflora. These results demonstrate the importance of T6SS mediated anti-microbial activity in vivo and identify a novel interaction between T-cells and common respiratory pathogens.

96 Introduction

In order for bacterial pathogens to effectively colonize a host, they must first compete with resident microflora. For respiratory tract pathogens, nasal cavity organisms, such as

Staphylococcus, Streptococcus, Niesseria, Enterobacteriaceae, Proteus, Corynebacteria, and

Mycobacterium, provide an inhospitable environment that pathogens must overcome to eventually colonize the lower respiratory tract [1]. Very little is currently known about how respiratory pathogens overcome microflora to cause disease or how host microorganisms assist or contribute to infectious diseases, even though respiratory diseases are currently the one of the leading causes of death worldwide [2]. In the gut, much more is known about microbial competition and how pathogens exploit the host immune system, but differences between gut and respiratory tract microflora communities and tissue variation between these mucosal surfaces certainly ensure that respiratory pathogens may have adapted novel mechanisms in which to compete with microflora.

Currently, three main competition classifications exist by which pathogens could interact with their microflora counterparts. Direct or interference competition, where one organism is killed by another organism through warfare-like mechanisms, can be used to target individual competitors, such as the release of colicins [3–5]. Secondly, indirect or exploitative competition, where the ability to obtain nutrients or factors for growth is at an advantage for one organism over another organism, can be used; an example of this would be the use of siderophores to capture iron [6–8]. Lastly, immune-mediated or apparent competition can be used to out- compete microbial species within a host, for example by causing an immune response that favors the pathogen [7,9]. These three mechanisms can easily be applied to competition between organisms that inhabit the upper respiratory tract.

97 The recently discovered bacterial Type VI Secretion System (T6SS) in opportunistic

respiratory pathogen Pseudomonas was shown to directly secrete toxins into other, competitor

bacterial species [10]. In the upper respiratory tract, Streptococcus pneumoniae has been shown

to use interference competition by releasing hydrogen peroxide as a means of directly out-

competing Haemophilus, while Bacillus has been shown to release antibiotics that alter the

ability of Streptococcus to colonize [11,12]. Interestingly, when Haemophilus and

Streptococcus were co-inoculated into animals to investigate immune-mediated competition,

Lysenko et al. determined that increased neutrophil recruitment was required for Haemophilus to

deplete Staphylococcus from the respiratory tract [13]. Although these examples of interspecies

competition have been well characterized, very little is known about how this competition affects

the entire microflora community during infection or how the addition of multiple bacteria affects

competition.

Bordetella bronchiseptica is a very common respiratory pathogen that infects a wide

range of animals, from cats, dogs, and mice to lions, seals, and humans [14]. The two causative

agents of whooping cough, Bordetella pertussis and Bordetella parapertussis are believed to

have independently evolved from a B. bronchiseptica like progenitor, through massive gene loss,

which is believed to cause human host restriction [15]. Although much is known about the

collection of bordetellae virulence mechanisms and the immune response and pathology that

follow colonization, extremely little is known how B. bronchiseptica uses these well studied

virulence factors to colonize during a natural infection, compete with bacterial species within the

host, and ultimately become a successful commensal-like organism after the initial infection has

waned. Because of its wide host range and low infectious dose, B. bronchiseptica can be used as

98 an excellent model system in which to study interactions between a pathogen and host microflora during natural infection.

In this study, we use the respiratory pathogen B. bronchiseptica to investigate how the upper respiratory tract microbiome changes during infection during a murine infection model and then examine the mechanisms this pathogen employs to achieve those interactions. Using both culture and metagenomic sequencing tools, we observed that B. bronchiseptica is capable of displacing nasal cavity microflora in a natural host, mice. In vitro, B. bronchiseptica requires a functional B. bronchiseptica T3SS, T6SS, and a BvgAS system to outcompete both Gram- positive and Gram-negative species. Immunodeficient mice and depletion experiments were utilized to determine that T-cells are also required for competition in vivo, unveiling a novel mechanism by which a respiratory pathogen can displace microflora as a means of initial colonization.

99 Results and Discussion

B. bronchiseptica displaces culturable upper respiratory tract microflora.

The goal of this study was to investigate how respiratory pathogens interact with and

overcome upper respiratory tract microflora as means to establish infection. To examine this, we

used a low dose, low

volume inoculation

(LDLV) model,

mimicking natural

infection, to inoculate

caged C57BL/6

littermates with

Bordetella bronchiseptica

strain RB50. A group of

three un-inoculated mice

were euthanized and

Figure 4-1 B. bronchiseptica displaces culturable murine nasal cavity flora. dissected, and bacteria A.) Total culturable nasal cavity flora (lines) and B. bronchiseptica (bars) colonization was monitored before and days 3, 7, and 70 days after a low dose, low present in a healthy, volume RB50 inoculation. Error bars represent +/- one standard deviation. B.) Each individual species cultured in Figure 2A was determine by 16S rRNA encoding sequence identification and graphed according to the number of CFU normal nasal cavity were identified in each mouse. S. J. Muse produced Figure 4-1B. cultured on blood agar

(BA) media (Figure 1A) and identified by sequencing the 16S rRNA encoding genome region of

each microflora (Figure 1B). After inoculation, three remaining mice were dissected at 3, 7, and

70 days post-inoculation, and both B. bronchiseptica and host microflora were identified on

100 Bordet Gengou agar with streptomycin (BGS) and BA. Although nearly 800 CFU of nasal cavity

flora were obtained prior to infection, no culturable microflora were obtained 3 days post-

inoculation when B. bronchiseptica grew to levels above 10,000 CFU, indicating the B.

bronchiseptica is capable of displacing culturable host microflora species upon colonization

(Figure 1A). Multiple Staphylococcus species, Bacillus clausii, Klebsiella PAMU, and

Enterobacter hormachei were all displaced within 3 days post-inoculation, indicating that this

early displacement affects both Gram-positive and Gram-negative species (Figure 1B).

Interestingly, two microflora species, Klebsiella PAMU and Enterobacter hormachei, were

recovered at very low levels on 7 days post-inoculation, suggesting that a few culturable species

may minimally survive the initial B. bronchiseptica colonization (Figure 1B). Both of these

species recovered on day 7 post inoculation are Gram-negative, which may indicate different levels of competition between more phylogenetically and evolutionarily similar species.

However, no culturable microflora were observed up to 70 days post-inoculation, indicating that

B. bronchiseptica is able to maintain the displacement of some microflora species, perhaps for the life of the animal (Figure 1A). Because similar results have been observed by pathogens in the gut and skin, these results suggest that displacing host microflora may be a shared pathogenesis mechanism for invading microbes anywhere in the body [16,17]. Additionally, these results may explain why B. bronchiseptica is able to transition from a pathogen into a more commensal lifestyle and why live, attenuated B. bronchiseptica nasal vaccines for animals are effective; an already colonized B. bronchiseptica strain is probably capable of displacing any future competitor species.

101 RB50 requires known virulence factors to compete with microflora.

Recently, the Type VI Secretion Systems (T6SS) in Pseudomonas, Vibrio, and Serratia has been shown to mediate microbial competition in vitro by secreting anti-microbial peptides

[10,22,23]. B. bronchiseptica was recently shown to contain a functional T6SS that contributes to immunopathology, persistence, and colonization in the murine model [24].

Additionally, B. bronchiseptica maintains numerous virulence factors shown to be required for similar processes in vivo, including a Type III Secretion System (T3SS), adenylate cyclase toxin

(ACT), and a master regulatory system of virulence factors, BvgAS. Therefore, we hypothesized that the microflora displacement phenotype we observed could be mediated by competition via known Bordetella virulence factors. Additionally, because Lactobacillus has been shown to displace E. coli and Salmonella strains from mucus glycoproteins and Caco-2 cell surfaces through an adherence dependent mechanisms, we additionally hypothesized that adherence factors, such as filamentous hemagglutinin (FHA) or pertactin (PRN), may also play a role in the capacity of B. bronchiseptica to adhere properly to the nasal epithelium and therefore to displace flora [16]. B. bronchiseptica mutants lacking a BvgAS system (RB54), FHA (RBX9), PRN

(SP5), T3SS (WD3), and T6SS (RB50ΔclpV) were LDLV inoculated into C57BL/6 mice, and culturable host microflora both prior to and after inoculation were determined (Figure 2A).

While culturable microflora displacement in wild-type RB50 inoculated mice was observed, mutants lacking major adherence factors, FHA and PRN, were able to displace large amounts of flora, but they were not as effective at displacing microflora as wild type B. bronchiseptica, suggesting the ability to adhere may play a partial role in this phenomenon. Additionally, mutants lacking BvgAS, T3SS, or T6SS stimulated growth of the host microflora species, suggesting that these known virulence factors play a role in the ability of B. bronchiseptica to

102 displace microflora. Two major hypothesizes exist to explain why these three virulence factors may contribute to host microflora displacement. First, these factors may be required for the direct or indirect killing of other prokaryotic species. Secondly, these factors could be required for a specialized immune response that results in the clearance of host microorganisms.

To determine if the BvgAS, T3SS, and T6SS played a role in direct killing or elimination of microflora from the nasal cavity, we tested the ability of isolated microflora species to

Table 1 Cross-streaking competition. Cross streaking competition results between B. bronchiseptica wild type strain RB50 and mutants lacking BvgAS regulatory system (ΔBvgAS), a T3SS (ΔBscN), or T6SS (ΔClpV). – designates that RB50 was not able to inhibit growth of a particular host microflora species, while + denotes inhibition interactions were observed. C. Safi performed this experiment.

compete with B. bronchiseptica in vitro. Two different assays were employed to investigate this observation: cross-streaking and co-culturing. Cross streaking was performed by culturing a line of B. bronchiseptica down the center of a BA plate, incubating for one day, and then culturing a line of an indicated microflora species perpendicular to and on top of the Bordetella for an additional day (Table 1). Mutants lacking a BvgAS, T3SS, or T6SS were additionally testing in this manner. When competing microflora species against B. bronchiseptica in direct contact on

103 agar, we observed that wild type RB50 was able to out-compete S. cohnii, S. lentus, and S. sp T-

02 (Table 1). Interestingly, a functional Bordetella T6SS was required to out-compete

Staphyloccous lentus and a T3SS was required to out-compete Staphylococcus cohnii. A BvgAS

system was not required for competition on agar. Furthermore, S. xylosus, S. sp. S16, Bacillus

clausii, Enterobacter cancerogenus, Klebsiella PAMU, and Enterococcus sp. were all capable of

growing in the presence of any Bordetella strain, suggesting that other mechanisms of

competition exist in vivo or that different environmental cues may be required for competition.

We additionally used a co-culture method to examine direct competition between

Bordetella and various host microflora species, where the two organisms were grown in while

shaking in PBS for two hours, simulating a very nutrient poor, moist environment. Again, we

tested wild type RB50 and mutants lacking BvgAS, T3SS, or T6SS function. We observed that

Entercoccus was capable of growing in the presence of RB50, while Staphylococcus was almost

completely killed in this scenario (Figure 2B). Klebsiella, Rhizobium, and Kytococcus species

neither grew nor were killed during this time. However, when we observed these conditions

with a T6SS mutant, RB50ΔclpV, Klebsiella and Staphylococcus species survived better

compared to co-culture with RB50, suggesting that a functional T6SS is required to compete

with these species during co-culture (Figure 2C). Similarly, a functional T3SS was observed to

be required for competition with every species except for Kytotococcus (Figure 2D).

104 Surprisingly, RB54, a RB50 derivative lacking the master virulence regulatory system BvgAS,

Figure 4-2 T3SS, T6SS, and BvgAS are required to displace flora in vitro. A. Host microflora recovered 3 days post-inoculation, as a percentage of microflora isolated from uninfected animals in the same cage, is graphed according to each B. bronchiseptica mutant tested. Error bars represent +/- one standard deviation. In vitro competition experiments were completed, competing wild type RB50 (B), a T6SS mutant, RB50ΔclpV (C), a T3SS mutant, RB50ΔbscN (D), and a mutant lacking a function BvgAS system, RB54 (E) with five microflora strains, including Klebsiella, Enterococcus, Rhizobium, Staphylococcus lentus, and Kytococcus, by growing each bordetellae strain with a microflora for 4 hours in PBS at 37°C.

was less capable of competing with every species tested, even though RB54 is believed to be

locked in a state better suited for environmental survival, and thus competition (Figure 2E).

105 Together, these data suggest that a functional T6SS, T3SS, and BvgAS system are

required for direct competition with numerous Gram-positive and Gram-negative host microflora species. To our knowledge, this is the first observation where a T3SS or BvgAS is required for anti-microbial functions, and this suggests that alternative secretion systems beyond the T6SS may possess anti-microbial functions. It is easy to postulate that anti-microbial peptides could be secreted through a T3SS or a T6SS. Because the B. bronchiseptica T3SS has already been shown to be BvgAS regulated, this may explain why BvgAS would be required [25].

Nevertheless, why BvgAS would be required for Kytotococcus competition, when the T3SS and

T6SS are not, remains undetermined and highlights how very little we know about the ability of bordetellae to compete with the microbiome. Additionally, when we only compare the ability of wild type RB50 to compete with flora in vitro, there are multiple instances where the host microflora is capable of growing or surviving in the presence of RB50, which is contradictory to what we observed in vivo. This suggests that another in vivo mechanism may play a role in the ability of B. bronchiseptica to displace microflora, beyond direct competition.

T-cells are required for microflora displacement and low dose colonization.

To investigate if B. bronchiseptica additionally utilizes immune-mediate competition as another means to displace respiratory microflora, we LDLV inoculated mice deficient in complement (C3-/-), B-cells (µMT), T and B-cells (Rag1-/-), T-cells (Tcrβδ-/-), Interferon-gamma

(IFN-γ), or interleukin-17 (IL-17-/-) with RB50 and identified the subsequent nasal cavity flora

three days later (Figure 3A). The levels of nasal cavity microflora increased 327% and 77% in

Rag1-/- and Tcrβδ-/- mice, respectively, three days post-inoculation with RB50, when compared with uninfected caged littermates. Additionally, the microflora in mice lacking IFN-γ increased

8% and IL-17-/- increased 18%, while mice lacking complement or B-cells alone were not

106 statistically different from wild-type mice. One potential explanation for these results is that the

total, culturable and unculturable, bacterial

composition of these mouse strains is vastly

different, because immune factors required for

maintaining host flora have been disrupted.

The interference or exploitative competition

between B. bronchiseptica and resident flora

could be vastly different in immunodeficient

mice than in C57BL/6 mice. Another

explanation suggests that these mice utilize

apparent competition, by exploiting T-cells,

IFN-γ, or IL-17 to displace flora. Interestingly, Figure 4-3 T-cells, IFN-γ, and IL-17 are required to displace microflora in vivo. IL-17 is largely known to stimulate anti- A.) Host microflora from mice deficient in complement (C3-/-), B-cells (µMT), B and T-cells (Rag1-/-), T-cells (Tcrβδ-/-), interferon-γ (IFN-γ-/-), and interfleukin-17 microbial peptide release from epithelial cells, receptor (IL-17-/-), recovered 3 days post-inoculation with RB50, as a percentage of microflora isolated from and B. bronchiseptica is known to be resistant uninfected animals in the same cage. Error bars represent +/- one standard deviation. *denotes p value to many of these peptides. These observations less than 0.5, and ** denotes a p value less than 0.01. B.) RB50 colonization in wild type mice (C57), mice deficient in B-cells and T-cells (Rag1-/-), and B-cell could potentially suggest a mechanism that deficient mice (mMT) was monitored after mice were given 5 CFU RB50 in 5 µL on days 3, 7, 14, and 28 explains how B. bronchiseptica may stimulate post-inoculation. O. Y. Rolin produced Figure 4-3B. an IL-17 release and production of anti-

microbial peptides, while itself remains unaffected [26–31]. However, in comparison to IL-17-/-

mice, T-cells deficient mice were much less capable of displacing flora, suggesting that T-cells,

and perhaps their activation, play additional roles beyond IL-17 production that are required for

B. bronchiseptica-mediated flora displacement.

107 Host microflora may act as decoys to phagocytic cells, provide nutrients or nutritious by-

products, or provide protection against additional invading species. However, displacing host

microflora via an immune-mediated mechanism could be advantageous for numerous reasons.

Displacing microflora could provide more nutrients for the bordetellae, open physical space

within the respiratory tract, change the immune response elicited by the pathogen, or even allow

the pathogen to colonize. Because we are using a LDLV infection model, we hypothesized that

the ability to displace host microflora may correlate with the ability of B. bronchiseptica to

initially colonize a host. To determine if T-cells were also required for initial colonization, we

-/- -/- inoculated wild type, Rag1 , Tcrβδ , or µMT mice with 10 CFU, a dose slightly above the ID50

that we found to colonize all wild type mice. We observed that all of the wild type and µMT

mice lacking B-cells were colonized at this dose; however, none of the mice lacking T-cells,

Rag1-/- and Tcrβδ-/-, were colonized, indicating that T-cells may be required for natural colonization (Figure 3B). Because these are the same mice in which the ability to displace flora is abrogated, this suggests that the ability to displace flora may serve as a mechanism for natural colonization. Very little is known about natural Bordetella infection and transmission, and although highly speculative, host microflora may contribute to a pathogens ability to infect particular hosts or even may contribute to the ability of respiratory pathogens to differentially cause disease in susceptible individuals.

Cytotoxic T-cells are required to displace culturable microflora.

We observed that T-cells were required for the displacement of respiratory tract microflora by B. bronchiseptica; however, we wanted to delve deeper into the mechanism behind this phenomenon. In the gut, T-cells have been shown to play a vital role in communicating with antigen sampling cells within Peyer’s Patches and regulating the immune responses toward flora

108 [32]. To determine which T-cell subtypes may be involved in this phenomena, we again LDLV

inoculated C57BL/6 mice RB50 and monitored T-cell populations in the nasal cavity, nasal

associated lymphoid tissue (NALT), and cervical * lymph nodes (CLN) at 12, 24, and 60 hours post-

inoculation (Figure 4A). Although the numbers

of CD4+ T-cells only tended to decrease, we did

observe a significant difference in the number of

CD4+CD8+ double positive T-cells as early as 12

hours post-inoculation. Most CD4+CD8+ are

tissue resident memory cells, most prevalent in

epithelial tissues, that retain functions of both T-

helper and T-effector cells types and respond to

antigen recall [33,34]. They were recently shown

Figure 4-4 CD4+CD8+ and CD8+ T-cells are to have anti-viral functions, capable of homing to recruited to the nasal cavity and required for microflora displacement. the site of infection [35,36]. We can speculate A.) Percentages of nasal cavity CD4+, CD4+CD8+ double positive, and CD8+ T-cells from the total CD3+ T-cell population are graphed in naïve that these T-cells may also have anti-microbial C57BL/6 mice and 12, 24, and 60 hours post- inoculation with wild type RB50. B.) α-CD4, α- functions, as observed with CD8+ effector T- CD8 or both antibodies were intraperitioneally injected into C57BL/6 mice 3 days prior to cells, which could then negatively affect nasal inoculation with RB50. Levels of host microflora were then monitored 3 days post-inoculation and graphed as a percentage of microflora recovered from cavity flora [37]. Furthermore, this study un-infected caged littermates. * denotes p value less than 0.5, and error bars represent +/- one standard provides insight into the similarities and deviation. differences between interactions of gut and

respiratory tract microflora and bacterial pathogens.

109 To determine if these double positive cells are indeed required for the displacement of

host microflora, we depleted either CD4+, CD8+ or both CD4+ and CD8+ T-cells from wild type C57BL/6 mice using monoclonal antibodies and determined the microflora levels three days post-inoculation with RB50. Interestingly, a 16,816% increase in nasal cavity microflora was observed in mice depleted of both CD4 and CD8, and a 705% increase in flora was observed in mice when CD8 was solely depleted (Figure 4B). Mice administered α-CD4 or an isotype control decreased microflora by 74% and 24%, but were not capable of completely clearing their nasal cavity flora. These data suggest that cytotoxic T-cell populations, either CD4+CD8+ double positive or CD8+ single positive T-cells, are required by B. bronchiseptica to decrease nasal cavity flora. Although it is plausible that CD4+ and CD8+ play an additional synergistic or combined role, CD4+ T-cells were not seen to accumulate in the nasal cavity during infection,

indicating that cytoxic cells are

most likely the major contributor.

Cytotoxic T-cells have been well

established to release perforin,

granulysin, and granzyme A and

Figure 4-5 Host microflora, but not B. bronchiseptica, are susceptible to B when activated during granzyme B in vitro. A.) Four microflora strains and B. bronchiseptica were exposed to three infection [36–40]. Both varying concentrations of granzyme B for 2 hours at 37°C in HBSS. Bacteria that survived over the two hours in comparison to bacteria incubated in HBSS alone were graphed as percent survival. granulysin and granzyme B have been shown to have antimicrobial functions, which could explain their role in displacing microflora [41,42]. However, granulysin has not been identified in mice, suggesting that granzyme B may be the mechanism in which CD8+ cells cause microbial death [43]. To test the

110 ability of granzyme B to kill host microflora species and whether or not B. bronchiseptica is

resistant to granzyme B, we exposed four host microflora species, S. saprophyticus, S. lentus, S. xylous, and Klebsiella PAMU, and B. bronchiseptica to varying concentrations of granzyme B

(Figure 5). While these microflora species were generally unaffected by granzyme B at 1.66×10-

4 and 1.66×10-2 µgs, all of the host microflora were negatively affected by 1.66 µg; only 59% of

S. lentus were recoverable after a 2 hour exposure at this dose, while 88% of S. saprophyticus

was the greatest percentage of any species recovered. Strikingly, B. bronchiseptica was capable

of growing while exposed to the higher granzyme B concentration tested, suggesting that it is resistant to granzyme-B mediated killing.

From these data, we can predict a model to explain how B. bronchiseptica successfully

displaces much of the host microbial community during infection. Upon inoculation, B.

bronchiseptica immediately engages with nasal cavity flora by utilizing both T3SS and T6SS, of

which potentially BvgAS regulated factors have bacteriostatic effects on host microflora. These

virulence factors may additionally stimulate CD8+ single positive and CD4+CD8+ double

positive memory T-cells. Activation of these cells results in granzyme B production, of which

the survival of both Gram-positive and Gram-negative species is negatively affected. B.

bronchiseptica is protected against granzyme B release, potentially by a yet undiscovered outer

membrane protein. Granzyme B could simply be released into the nasal cavity environment or in

conjuctions with perforin to displace flora inhabiting dendritic cells [44]. Another possibility lay

within the ability of CD8+ T-cells to stimulate epithelial cells to release anti-microbial peptides,

by producing IL-17 [45].

This research provides a novel mechanism in which a respiratory pathogen can interact

and ultimately overcome resident microflora in the nasal cavity. This paper proves the

111 essentiality for antimicrobial functions of the T6SS in vivo and additionally suggests that the

T3SS may have similar functions. Furthermore, it elucidates a novel function of CD8+ T-cells during infection, which can have repercussions on any pathogen infecting mucosal surfaces.

Lastly, we show novel functionality for CD4+CD8+ double positive cells, which will need to be studied in further detail to investigate other potential mechanisms they have in exposed epithelial tissue. Future research to determine the mechanism behind how B. bronchiseptica stimulates cytotoxic T-cells to displace respiratory tract microflora will open doors for numerous novel therapeutic agents, ideally to better manipulate and strengthen our own respiratory tract host microflora to prevent infection.

112

Materials and Methods

Bacterial Strains.

The wild type B. bronchiseptica strain RB50 and various RB50 derived mutants used in this study have been described elsewhere: wild type RB50 [46], Bvg- phased locked mutant

(RB54) [46], filamentous hemmagluttinin mutant (RBX9) [47], pertactin mutant (SP5) [48],

Type III Secretion System mutant (WD3) [49], and a Type VI Secretion System mutant

(RB50ΔclpV) (Chapter 2). Host microflora species isolated in this study were initially cultured on Luria-Bertani agar (LB) or Blood Agar (BA) (BD Biosciences) supplemented with 5% sheep blood (Hema Resources) for 48 hours at 37°C and were maintained on BA thereafter.

Bordetellae were maintained on Bordet-Gengou agar (Difco) supplemented with 10% sheep blood (Hema Resources) with 20 µg/ml streptomycin (Sigma) (BGS). Liquid bordetellae

cultures were grown in exponential phase while shaking in Stainer-Scholte (SS) broth at 37°C, while liquid host microflora cultures were grown in LB broth (BD Biosciences) while shaking at

37°C.

Mouse Experiments.

Wild type C57BL/6, B cell deficient: B6.129S2-igh-6tm1cgn/J (µMT), combined B and T

cell deficient: B6.129S7-Rag1tm1Mom/J (Rag1-/-), T cell deficient: B6.129P2-Tcrbtm1Mom

Tcrdtm1Mom/J (TcrB, and Interferon-γ deficient B6.129S7-Ifngtm1Ts/J mice were obtained from

Jackson Laboratories in Bar Harbor, ME. Mice deficient in complement (C3-/-) were a kind gift

from Dr. Rick Wetzel at Baylor University. Mice were bred and housed at a specific pathogen-

free facility at The Pennsylvania State University in University Park, PA. All animal

experiments were carried out according to institutional guidelines and were performed as

previously described [47,50,51] with these exceptions. Mice used for these experiments were all

113 caged littermates, co-housed for at least 5 weeks before experimentation to ameliorate their host

microflora. A low dose, low volume (LDLV) inoculation method was used for all of these

experiments where 10 or 100 CFU of Bordetella in 10 uL were gently pipetted onto the external

nares of a sedated mouse. For in vivo bacterial quantification and host microflora identification,

mice were then euthanized on the designated day post-inoculation, and the nasal cavity was

excised, homogenized in PBS, serially diluted, and plated onto BGS, LB, and BA agar. After 48 hrs incubation at 37°C, the numbers of Bordetella were determined by counting colonies on BGS,

while all other bacteria were counted and isolated from either LB or BA.

CD4+ and CD8+ Depletion

Monoclonal antibodies were a kind gift from Andrew Gunderson at The Pennsylvania

State University. Briefly, depleting monoclonal antibodies were purified from hybridoma

supernatants grown and collected in CELLine 1000 culture flasks (Argos Technologies) over a

Protein G Sepharaose column (GE Healthcare) and dialyzed in PBS overnight (Spectrum

Laboratories). Dilute antibody solutions were concentrated using a centrifugal filter unit

(Millipore). Specific leukocytes were depleted with 100 µg of GK1.5 (α-CD4) or YTS169.4 (α-

CD8β) antibodies suspended in sterile PBS were administered intraperitoneally three days prior to inoculation.

Host Microflora Species Identification

All culturable nasal cavity host microflora species from each mouse were cultured on agar, described morphologically, isolated, grown in liquid culture, and preserved in 20% glycerol

(Sigma) at -80°C. One mL of liquid culture was obtained, and DNA extraction (Qiagen) was performed on each individual species. The 1,200 bp 16S rRNA V2 region was amplified from

each species and sent for sequencing at The Pennsylvania State University Genomics Core

114 Facility [52]. Both forward and reverse sequences were aligned via MEGA4 [53], and the consensus sequence was submitted to NCBI BLAST for identification. The top known hit was used for identification.

454 Sequencing.

Nasal cavity samples were obtained in a sterilized hood utilizing UV sterilized, DNA-zap

(Ambion) treated dissecting tools and placed 1 mL of in DNA and RNA free PBS (Invitrogen).

The samples were homogenized via vortexing, and 200 µL was removed using DNA and RNA free filter tips in a sterilized hood. Samples were then frozen at -80°C until DNA extraction was performed. Total DNA was extracted from each sample using the BiOstic FFPE Tissue DNA

Isolation Kit (MoBio) as per the manufacturer’s instructions, but increasing the 55°C incubation length in step 3 to 12 hours. DNA quality was assessed via Nanodrop and Qubit, and samples above 2 ng/µL were used for sequencing. At The Pennsylvania State University Core Genomics

Facility, samples were barcoded, and targeted region sequencing using GS FLX Titanium series amplicons 28F

(CGTATCGCCTCCCTCGCGCCATCAGATATCGCGAGAGAGTTTGATCMTGGCTCAG) and 518R

(CCTATCCCCTGTGTGCCTTGGCAGTCTCAGACGAGTGCGTWTTACCGCGGCTGCTG

G) primers were used to amplify the 16S rRNA V2 regions present within the sample. The Core

Genomics Facility also completed 454 sequencing for each of the eight samples, obtaining over

10,000 reads per sample.

Metagenomic Analysis

Artificial barcodes were trimmed from the 454 raw sequence reads, and sequences were screened based on the quality file from 454 sequencer. We only considered the unique sequences

115 to run standalone BLASTN against the rRNA database, including both Ribosomal Database

Project (RDP) and greengenes databases [54]. MEtaGenome ANalyzer (MEGAN) was then run

to place BLAST output on National Center for Biotechnology Information (NCBI) taxonomic

tree and to compare each sample [55].

Bacterial Competition Experiments.

Staphylococcus saprophyticus, Kytococcus, Klebsiella, Rhizobium, Enterococcus

hormachea, and RB50 cultures were grown overnight to an optical density of ~0.8. Using the

optical densities to estimate culturable CFUs, all cultures were then all diluted to 106 CFU in

sterile PBS. To complete cross streaking, one loopful of RB50 was taken an streaked in a line on

BA, followed by streaking a loopfull of each host microflora species perpendicular to RB50.

Bacterial streaks were then grown for 48 hours at 37ºC to determine if the growth of either had

been inhibited. To complete co-culture, each culture was mixed with an equal part of RB50,

centrifuged into pellicle form, and incubated at 37ºC for 6 hours with shaking. Controls for all

six bacteria were also made by adding a 1:1 ratio of 106 bacteria with PBS. Each of the mixtures and controls were plated at before and after 6 hours incubation on BA and BGS. The plates were

incubated for 48 hours at 37ºC, and resulting colonies were counted thereafter.

Flow cytometry.

Prior to dissection 10-20 Ml of PBS was perfused through the left ventricle of mice while

venous runoff was collected from the orbit. Nasal bones were dissected and placed in 1ml of

DMEM containing 5% FBS and 1mg/ml collagenase D. Samples were incubated for 45 minutes

at 37oC and subsequently disaggregated into a single cell suspension by mechanical disruption

over a 70µm mesh screen. Mediastinal and cervical lymph nodes were harvested in DMEM

containing 5% FBS and disaggregated directly without further treatment. 2x106 cells per well

116 were then added to 96 well plates. Samples were resuspended in FC blocking buffer (200:1 anti-

CD16/32 BDbiosciences in PBS + 2% FBS) and incubated on ice for 20 minutes. Following washing, cell surface markers were labeled with the following antibodies in PBS + 2% FBS: anti-CD45 APC-cy7 BD 400:1, anti-CD8α PerCP (200:1), anti-CD3 Horizon V450 (200:1), anti-

CD4 Horizon V500, anti-CD69 FITC (100:1), anti-CD127(IL-7R) PE (400:1), anti CD44 APC

(800:1) (BD bioscience).

117

Author Contributions:

Authors: Laura S. Weyrich1,2, Olivier Y. Rolin1,3, Sarah J. Muse1,2, Jihye Park1,4, Chetan Y.

Safi1, Sarah E. Young1, and Eric T. Harvill1

1Department of Veterinary and Biomedical Sciences, The Pennsylvania State University

W-213 Millennium Science Complex, University Park, PA 16802

2Graduate Program in Biochemistry, Microbiology, and Molecular Biology

3 Graduate Program in Immunology and Infectious Disease

4 Graduate Program in Bioinformatics and Genomics

Conceived and designed experiments: LSW, OYR, ETH

Preformed experiments: LSW, OYR, SJM, JP, CYS, SEY

Analyzed data: LSW, ETH

Wrote manuscript: LSW, ETH

118

References

1. Efremova SA, Kurina SA, Grig MG, Sazpnova IM (1977) [Microflora of the nasal cavity and its sensitivity to antibiotics in patients with chronic rhinitis treated by cryotherapy]. Antibiotiki 22: 324-327.

2. WHO | The world health report 2004 - changing history (2004). Available: http://www.who.int/whr/2004/en/. Accessed 12 Jan 2012.

3. Normal Flora - Medical Microbiology - NCBI Bookshelf (n.d.). Available: http://www.ncbi.nlm.nih.gov/books/NBK7617/. Accessed 8 Jan 2012.

4. Zambrano MM, Siegele DA, Almirón M, Tormo A, Kolter R (1993) Microbial competition: Escherichia coli mutants that take over stationary phase cultures. Science 259: 1757-1760.

5. Van Melderen L, Saavedra De Bast M (2009) Bacterial Toxin–Antitoxin Systems: More Than Selfish Entities? PLoS Genet 5: e1000437. doi:10.1371/journal.pgen.1000437

6. Smith VH, Holt RD (1996) Resource competition and within-host disease dynamics. Trends in Ecology & Evolution 11: 386-389. doi:10.1016/0169-5347(96)20067-9

7. Read AF, Taylor LH (2001) The Ecology of Genetically Diverse Infections. Science 292: 1099 -1102. doi:10.1126/science.1059410

8. Griffin AS, West SA, Buckling A (2004) Cooperation and competition in pathogenic bacteria. Nature 430: 1024-1027. doi:10.1038/nature02744

9. Stecher B, Robbiani R, Walker AW, Westendorf AM, Barthel M, et al. (2007) Salmonella enterica Serovar Typhimurium Exploits Inflammation to Compete with the Intestinal Microbiota. PLoS Biol 5: e244. doi:10.1371/journal.pbio.0050244

10. Hood RD, Singh P, Hsu F, Güvener T, Carl MA, et al. (2010) A Type VI Secretion System of Pseudomonas aeruginosa Targets a Toxin to Bacteria. Cell Host Microbe 7: 25-37. doi:10.1016/j.chom.2009.12.007

11. Pericone CD, Overweg K, Hermans PWM, Weiser JN (2000) Inhibitory and Bactericidal Effects of Hydrogen Peroxide Production by Streptococcus pneumoniae on Other Inhabitants of the Upper Respiratory Tract. Infection and Immunity 68: 3990 -3997. doi:10.1128/IAI.68.7.3990-3997.2000

12. Gonzalez DJ, Haste NM, Hollands A, Fleming TC, Hamby M, et al. (2011) Microbial competition between Bacillus subtilis and Staphylococcus aureus monitored by imaging mass spectrometry. Microbiology 157: 2485 -2492. doi:10.1099/mic.0.048736-0

119 13. Lysenko ES, Ratner AJ, Nelson AL, Weiser JN (2005) The Role of Innate Immune Responses in the Outcome of Interspecies Competition for Colonization of Mucosal Surfaces. PLoS Pathog 1: e1. doi:10.1371/journal.ppat.0010001

14. Mattoo S, Cherry JD (2005) Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev 18: 326-82.

15. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, et al. (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35: 32-40. doi:10.1038/ng1227 ng1227 [pii]

16. Lee Y-K, Puong K-Y, Ouwehand AC, Salminen S (2003) Displacement of bacterial pathogens from mucus and Caco-2 cell surface by lactobacilli. Journal of Medical Microbiology 52: 925 -930. doi:10.1099/jmm.0.05009-0

17. Park B, Nizet V, Liu GY (2008) Role of Staphylococcus aureus Catalase in Niche Competition against Streptococcus pneumoniae. Journal of Bacteriology 190: 2275 -2278. doi:10.1128/JB.00006-08

18. Costello EK, Lauber CL, Hamady M, Fierer N, Gordon JI, et al. (2009) Bacterial Community Variation in Human Body Habitats Across Space and Time. Science 326: 1694 -1697. doi:10.1126/science.1177486

19. Fierer N, Hamady M, Lauber CL, Knight R (2008) The influence of sex, handedness, and washing on the diversity of hand surface bacteria. Proceedings of the National Academy of Sciences 105: 17994-17999. doi:10.1073/pnas.0807920105

20. Gosalbes MJ, Durbán A, Pignatelli M, Abellan JJ, Jiménez-Hernández N, et al. (n.d.) Metatranscriptomic Approach to Analyze the Functional Human Gut Microbiota. PLoS One 6. doi:10.1371/journal.pone.0017447

21. Contreras M, Costello EK, Hidalgo G, Magris M, Knight R, et al. (2010) The bacterial microbiota in the oral mucosa of rural Amerindians. Microbiology 156: 3282-3287. doi:10.1099/mic.0.043174-0

22. MacIntyre DL, Miyata ST, Kitaoka M, Pukatzki S (2010) The Vibrio cholerae type VI secretion system displays antimicrobial properties. Proceedings of the National Academy of Sciences 107: 19520 -19524. doi:10.1073/pnas.1012931107

23. Murdoch SL, Trunk K, English G, Fritsch MJ, Pourkarimi E, et al. (2011) The Opportunistic Pathogen Serratia marcescens Utilizes Type VI Secretion To Target Bacterial Competitors. J. Bacteriol. 193: 6057-6069. doi:10.1128/JB.05671-11

24. Weyrich, L.S., Rolin, O. Y., Muse, S. J., Park, J., Spidale, N., et al. (2012) A Type VI Secretion System Encoding Locus is Required for Bordetella bronchiseptica Immunomodulation and Persistence In Vivo. PLoS One Submitted.

120 25. Mattoo S, Yuk MH, Huang LL, Miller JF (2004) Regulation of type III secretion in Bordetella. Mol Microbiol 52: 1201-14.

26. Fernandez RC, Weiss AA (1996) Susceptibilities of Bordetella pertussis strains to antimicrobial peptides. Antimicrobial Agents and Chemotherapy 40: 1041 -1043.

27. Banemann A, Deppisch H, Gross R (1998) The Lipopolysaccharide of Bordetella bronchiseptica Acts as a Protective Shield against Antimicrobial Peptides. Infection and Immunity 66: 5607 -5612.

28. Kabelitz D, Schröder J-M (2005) Mechanisms of epithelial defense. Karger Publishers. p.

29. Liang SC, Tan X-Y, Luxenberg DP, Karim R, Dunussi-Joannopoulos K, et al. (2006) Interleukin (IL)-22 and IL-17 are coexpressed by Th17 cells and cooperatively enhance expression of antimicrobial peptides. The Journal of Experimental Medicine 203: 2271 - 2279. doi:10.1084/jem.20061308

30. Nizet V (2006) Antimicrobial peptide resistance mechanisms of human bacterial pathogens. Curr Issues Mol Biol 8: 11-26.

31. Selsted ME, Szklarek D, Lehrer RI (1984) Purification and antibacterial activity of antimicrobial peptides of rabbit granulocytes. Infection and Immunity 45: 150 -154.

32. Cerf-Bensussan N, Gaboriau-Routhiau V (2010) The immune system and the gut microbiota: friends or foes? Nat Rev Immunol 10: 735-744. doi:10.1038/nri2850

33. Federico A Z (1999) Extrathymic CD4/CD8 double positive T cells. Veterinary Immunology and Immunopathology 72: 55-66. doi:10.1016/S0165-2427(99)00118-X

34. Zuckermann FA, Husmann RJ (1996) Functional and phenotypic analysis of porcine peripheral blood CD4/CD8 double-positive T cells. Immunology 87: 500-512.

35. Nascimbeni M, Pol S, Saunier B (2011) Distinct CD4+CD8+ Double-Positive T Cells in the Blood and Liver of Patients during Chronic Hepatitis B and C. PLoS ONE 6: e20145. doi:10.1371/journal.pone.0020145

36. Nascimbeni M, Shin E-C, Chiriboga L, Kleiner DE, Rehermann B (2004) Peripheral CD4+CD8+ T cells are differentiated effector memory cells with antiviral functions. Blood 104: 478 -486. doi:10.1182/blood-2003-12-4395

37. Harty JT, Tvinnereim AR, White DW (2000) CD8+ T Cell Effector Mechanisms in Resistance to Infection. Annual Review of Immunology 18: 275-308. doi:10.1146/annurev.immunol.18.1.275

38. Lauw FN, Simpson AJ, Hack CE, Prins JM, Wolbink AM, et al. (2000) Soluble granzymes are released during human endotoxemia and in patients with severe infection due to gram- negative bacteria. J. Infect. Dis. 182: 206-213. doi:10.1086/315642

121 39. Thiery J, Walch M, Jensen DK, Martinvalet D, Lieberman J (n.d.) Isolation of Cytotoxic T Cell and NK Granules and Purification of Their Effector Proteins.. Available: http://onlinelibrary.wiley.com/doi/10.1002/0471143030.cb0337s47/abstract. Accessed 13 Jan 2012.

40. Winkler U, Fraser SA, Hudig D (1997) Perforin-Enhancing Protein, a Low Molecular Weight Protein of Cytotoxic Lymphocyte Granules, Enhances Perforin Lysis. Biochemical and Biophysical Research Communications 236: 34-39. doi:10.1006/bbrc.1997.6899

41. Stenger S, Hanson DA, Teitelbaum R, Dewan P, Niazi KR, et al. (1998) An antimicrobial activity of cytolytic T cells mediated by granulysin. Science 282: 121-125.

42. Shafer W, Pohl J, Onunka V, Bangalore N, Travis J (1991) Human lysosomal cathepsin G and granzyme B share a functionally conserved broad spectrum antibacterial peptide. J. Biol. Chem. 266: 112-116.

43. Krensky AM, Clayberger C (2009) Biology and clinical relevance of granulysin. Tissue Antigens 73: 193-198. doi:10.1111/j.1399-0039.2008.01218.x

44. Macpherson AJ, Uhr T (2004) Induction of Protective IgA by Intestinal Dendritic Cells Carrying Commensal Bacteria. Science 303: 1662 -1665. doi:10.1126/science.1091334

45. Happel KI, Zheng M, Young E, Quinton LJ, Lockhart E, et al. (2003) Cutting Edge: Roles of Toll-Like Receptor 4 and IL-23 in IL-17 Expression in Response to Klebsiella pneumoniae Infection. The Journal of Immunology 170: 4432 -4436.

46. BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. (1994). Available: http://medjournal.hmc.psu.edu:2060/pmc/articles/PMC302969/?tool=pubmed. Accessed 24 Jul 2011.

47. Cotter PA, Yuk MH, Mattoo S, Akerley BJ, Boschwitz J, et al. (1998) Filamentous Hemagglutinin of Bordetella bronchiseptica Is Required for Efficient Establishment of Tracheal Colonization. Infect. Immun. 66: 5921-5929.

48. Edwards JA, Groathouse NA, Boitano S (2005) Bordetella bronchiseptica Adherence to Cilia Is Mediated by Multiple Adhesin Factors and Blocked by Surfactant Protein A. Infect Immun 73: 3618-3626. doi:10.1128/IAI.73.6.3618-3626.2005

49. Yuk MH, Harvill ET, Miller JF (1998) The BvgAS virulence control system regulates type III secretion in Bordetella bronchiseptica. Mol Microbiol 28: 945-59.

50. Harvill ET, Cotter PA, Miller JF (1999) Pregenomic comparative analysis between bordetella bronchiseptica RB50 and Bordetella pertussis tohama I in murine models of respiratory tract infection. Infect Immun 67: 6109-18.

122 51. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of antibodies in immunity to Bordetella infections. Infect Immun 71: 1719-24.

52. Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173: 697-703.

53. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596-9. doi:msm092 [pii] 10.1093/molbev/msm092

54. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J. Mol. Biol 215: 403-410. doi:10.1006/jmbi.1990.9999

55. Huson DH, Auch AF, Qi J, Schuster SC (2007) MEGAN analysis of metagenomic data. Genome Res 17: 377-86. doi:gr.5969107 [pii] 10.1101/gr.5969107

123

Chapter 5 : Murine Upper Respiratory Tract Microflora Inhibit Bordetella pertussis

Colonization

124

Abstract

The human restricted pathogen, Bordetella pertussis, is highly successful in humans but requires high doses to colonize the murine nasal cavity. A closely related pathogen, Bordetella bronchiseptica, has a broad host range, including humans, and very efficiently colonizes the murine nasal cavity at low doses. We hypothesized that interactions with nasal cavity microflora may mediate these differences, potentially contributing to infectiousness and host restriction.

Using a low volume inoculation, we determined that B. bronchiseptica able to colonize mice at low doses and was able to displace culturable flora. In stark contrast, B. pertussis was unable to colonize or displace murine nasal cavity organisms. The ID50 of B. pertussis was decreased 99% by administering a broad spectrum antibiotic prior to inoculation, indicating that murine microflora inhibit B. pertussis colonization at low doses. Staphylococcus, Klebsiella,

Kytococcus, and Rhizobium species were isolated from the nasal cavities of healthy mice and were then grown in the presence of B. pertussis to assay competition. Staphylococcus,

Kytococcus, and Rhizobium species were all capable of inhibiting B. pertussis growth, while

Klebsiella was capable of killing B. pertussis in vitro. Following antibiotic treatment, single doses of Staphylococcus and Klebsiella species blocked B. pertussis colonization in vivo, suggesting these organisms may be used as upper respiratory tract prophylactics to protect against disease. Understanding how resident microflora block pathogen colonization has important applications both in the improved use of animal models, as well as disease treatments and preventions.

125

Introduction

Bordetella pertussis is the major causative agent of whooping cough, which continues to

persist in both highly vaccinated and unvaccinated populations worldwide. The World Health

Organization estimates that 20-40 million whooping cough cases occur annually worldwide,

resulting in over 400,000 deaths each year [1,2]. Acellular and whole cell B. pertussis vaccines

have been highly effective in decreasing the number of whooping cough cases, especially in developed countries [3]. However, in recent years, annual B. pertussis incidence has increased even in highly vaccinated populations, with over 25,000 cases reported in 2005 in the United

States alone [4,5]. Waning immunity in adolescence and immune evasion through pathogen evolution are all proposed to contribute to this recent whooping cough resurgence [6].

Furthermore, B. pertussis is even more endemic in under developed countries where the case- fatality rates are much higher [1,2]. Therefore, novel and inexpensive solutions to control and prevent whooping cough are necessary to effectively combat B. pertussis in the 21st century.

B. pertussis is one of the classical bordetellae, a group of three highly conserved bacteria

that commonly cause respiratory infections, additionally including Bordetella parapertussis and

Bordetella bronchiseptica. Unlike B. bronchiseptica, B. pertussis and B. parapertussis appear to

be host restricted; B. bronchiseptica has a very diverse host range, including dogs, seals, lions,

and humans [7]. Additionally, B. bronchiseptica can cause a wide range of disease severities,

from asymptomatic colonization to lethal pneumonia, unlike the two human restricted pathogens

[8]. In a murine infection model, all three classical bordetellae colonize the lower respiratory

tract when administered at high dose inoculations of 5 ×105 CFU onto the external nares [9–11].

However, the ID50s of these three organisms vary wildly, from 5 CFU (B. bronchiseptica strain

126 RB50) to a yet undetermined dose (B. pertussis) [8]. Additionally, B. bronchiseptica remains in the nasal cavity of mice for the life of the animals, assuming a commensal-like behavior in the upper respiratory tract after the lower respiratory tract infection has waned [9,12]. This is in stark contrast to the two human pathogens, where B. pertussis is cleared from the murine upper respiratory tract within 28 days post-inoculation, while B. parapertussis is eliminated within 49 days post-inoculation [9,13,14].

In this study, we aimed to understand why B. pertussis lacks the ability to colonize the murine nasal cavity similar to B. bronchiseptica. We hypothesized that the differences between these pathogens during murine infection could be explained by three potential mechanisms: differences in adherence to murine cells, varied immune responses between mice and humans in vivo, or differential competition with murine microflora. Interestingly, all three classical

Bordetella species have been shown to effectively adhere to murine epithelial cells in vitro, and although these pathogens each stimulate a unique immune response, each pathogen can colonize a myriad of immunodeficient mice [8]. These studies suggest that adherence or variations in the immune response are probably not the major governing factors behind differences in ID50 and nasal cavity colonization. Although these observations rule out differences in adherence and varied immune responses, these data do suggest that there are unknown mechanisms contributing to this observation; for example, interactions with host microflora may contribute to host restriction and ability to effectively establish colonization in the upper respiratory tract.

Although mice are not the primary host of B. pertussis, the murine infection model has been highly effective in understanding B. pertussis infection kinetics, immune response, and vaccination efficacy, even if this is achieved by administering large doses (~5×105 CFU) [15].

However, another goal of this study is to investigate novel methods in which to create a more

127 natural B. pertussis infection model with the ability to use lower inoculation doses and volumes.

Interestingly, numerous other pathogens cannot be studied in mice, because murine infection

cannot be established even when large doses are utilized. Therefore, researchers must use

immunodeficent mice, predispose mice to chemicals, or even genetically engineer mice to create

a model system. For example, Psuedomonas aeurginosa requires a burn mouse model to study

the pathogenesis of wound sepsis [16], neonatal mice to study pneumonia [17], and neutropenic

mice to understand how virulence factors contribute to disease [18]. By understanding

interactions with host flora and how these microorganisms contribute to an infection may allow

researchers to create novel animal models.

We aimed to investigate the impact of murine microflora on the ability of B. pertussis to

colonize the upper respiratory tract. We observed that the low volume ID50 of B. pertussis can

be lowered ~100 fold if antibiotics are administered to the external nares prior to inoculation,

suggesting that murine nasal cavity flora are capable of blocking low dose colonization.

Staphylococcus, Kytococcus, Klebsiella, Rhizobium, and Enterococcus species were identified in

the upper respiratory tract of healthy mice and were then competed against B. pertussis in vitro.

Host microflora prevented B. pertussis growth in vitro and initial colonization in vivo, suggesting that murine host microflora may be used as prophylactics against B. pertussis infection. Lastly, we determined that a closely related, zoonotic pathogen, B. bronchiseptica, is capable of displacing murine microflora, while B. pertussis was observed to be capable of out-competing human flora in vitro, suggesting pathogens may evolve to compete with that pathogens have developed mechanisms to effectively compete with microflora common to their natural host.

Together, these data provide the initial framework to investigate how respiratory pathogens overcome host microflora to initiate infection.

128

Material and Methods

Bacterial Strains

B. pertussis strain 536 and B. bronchiseptica strain RB50 are described elsewhere

[19,20]. Bordetellae were maintained on Bordet-Gengou agar (Difco) with 10% sheep blood

(Hema Resources) and 20 µg/ml streptomycin (Sigma). Host microflora isolates were obtained

by culture on Luria-Bertani agar (LB) or Blood Agar (BA) (BD Biosciences) supplemented with

5% sheep blood (Hema Resources). B. bronchiseptica and host microflora species were grown

on agar for 48 hours at 37°C, while B. pertussis was grown for 5 days at 37°C. To obtain liquid

cultures, bordetellae were grown overnight to exponential phase in Stainer-Scholte (SS) broth

while shaking at 37°C, whereas host microflora isolates were similarly grown in LB broth (BD

Biosciences).

Mouse Experiments

C57BL/6 wild type mice were obtained from Jackson Laboratories (Bar Harbor, ME) and were bred and housed in a pathogen free facility at The Pennsylvania State University in

University Park, PA. All animal experiments were completed according to institutional guidelines. Animal experiments were done as previously described with the following modifications [20–22]. Mice were inoculated by gently pipetting 100 CFU of bordetellae or host microflora suspended in 10 µL sterile PBS onto the external nares of anesthetized mice to mimic a more natural infection. Three days post-inoculation, mice were euthanized using CO2, and the

nasal cavity was excised, homogenized, and plated on BG, BA ,and LB to identify quantities of bordetellae (BG) and host microflora (BA and LB).

Host Species Identification

129 All culturable nasal cavity host microflora cultured on agar were described

morphologically, isolated in pure culture, and preserved in 20% glycerol (Sigma) at -80°C.

DNA extraction using QIAmp DNA Mini Kit (Qiagen) was performed on each individual

species, and the 16S rRNA V2 region was amplified using 16SF

(AGAGTTTGATCATGGCTCAG) and 16SR (AAGGAGGTGATCCAACC) primers in the

following conditions: 94°C for 5 min; 95°C for 30 s, 56°C for 1 min, 72°C for 1.5 min, 30

times; and 72°C for 8 min [23]. Following amplification, the ~1,500 bp PCR products was

isolated and purified using MinElute Gel Extraction Kit (Qiagen) as per manufacturer’s

instructions. Sequencing of each 16S rRNA encoding region was completed at The

Pennsylvania State University Genomics Core Facility. Both forward and reverse sequences

were aligned via MEGA4 [24], and the consensus sequence was submitted to NCBI BLAST for

identification [25]. The top known hit was used for identification, as long as it was above 70%

homologous.

Metagenomic Sequencing

Three nasal cavities from uninfected mice were obtained from our breeding facility.

Total DNA isolation was performed in a UV sterilized hood with DNA-zap (Ambion) treated instruments and DNA and RNA free filtered pipette tips. BiOstatic FFPE Tissue DNA Isolation

Kit (Mo Bio Laboratories) was used to isolate total DNA, with the following modifications: the samples were left at 55° C overnight, instead of one hour (step 3) and step 4 was omitted. 454 amplicon sequencing, library preparations, and sequencing was completed at The Pennsylvania

State University Core Genomic facility. Bar-coded samples were sequenced together on a 454 quad to achieve at least 20,000 reads per sample. Samples below 100 bp were omitted from sequence analysis. Sequences were initially analyzed by Mothur to assess quality and identify

130 species based on a BLAST method, identifying species based on sequence similarities within the

GreenGenes database (http://greengenes.lbl.gov) [26]. Sequences were then analyzed via MEta

Genome ANalyzer (MEGAN), producing phylogenetic classification and relationships to known

NCBI taxonomic trees [27].

Bacterial Competition Experiments

S. saprophyticus, S. lentus, S. xylous, Kytococcus, Klebsiella, Rhizobium and B. pertussis cultures were grown to exponential phase. Using the optical densities to estimate culturable

CFUs, all cultures were then all diluted to 1×107 CFU in SS liquid media. To compete each

microflora with B. pertussis, each host species was mixed with an equal part of RB50 and

incubated at 37ºC for 40 hours with shaking. B. pertussis was also grown alone in SS, starting

with 1×107 CFU as a control. Each mixture and control was performed in triplicate and was

plated at 0, 2, 4, 8, 24, 36 hours BA and BG with 60 ng/mL streptomycin. The plates were

incubated for 48 hours (microflora) or 5 days (B. pertussis) at 37ºC, and resulting colonies were counted thereafter.

131

Results

B. pertussis requires 10,000 CFU to colonize a murine nasal cavity.

To our knowledge, the B. pertussis infectious dose (ID50), in which half the animals become colonized using large inoculation doses of 50 µL, has not been determined.

Furthermore, the ID50s of neither B. bronchiseptica nor B. pertussis has been determined using a low dose, low volume inoculation that mimics the initial events of natural colonization. To

answer this question, C57BL/6 wild type mice

were inoculated with 5, 50, 100, 500, 1×103,

5×103, 1×104, 1×105, or 1×106 CFU in 5 µL of

either B. pertussis strain 536 or B.

bronchiseptica strain RB50, and levels of

colonization were determined three days post-

inoculation (Figure 1). 5 CFU of B.

Figure 5-1 104 CFU B. pertussis are required to bronchiseptica in 5 µL was sufficient to colonize colonize murine nasal cavity. 10µL of B. bronchiseptica (□) or B. pertussis (◊) was the nasal cavity of half of the mice inoculated, inoculated into groups of 3 wild type C57BL/6 mice. Nasal cavity colonization was assessed three days later, and the percentage of colonized animals 3 days while at least 10,000 CFU of B. pertussis in 5 µL post-inoculation was graphed. was required to observe any remaining B. pertussis three days post-inoculation. This result suggests that B. bronchiseptica, a natural murine pathogen, and B. pertussis, a human restricted pathogen, have different capabilities to colonize the upper respiratory tract and therefore suggest these pathogens utilize different mechanisms to establish infection in the nasal cavity.

132 Murine nasal cavity microflora inhibit low dose colonization.

Because we observed that B. pertussis was inefficient at colonizing the murine nasal

cavity compared to B. bronchiseptica, we hypothesized that this phenomena may be affected by

the presence of host microflora present in the upper respiratory tract. To investigate this,

C57BL/6 mice were intranasally treated with a

broad spectrum antibiotic, enrofloxacin, a

bacteriocidal compound shown to affect both

Gram-positive and Gram-negative species, prior

to inoculation with B. pertussis. Both antibiotic

treated and control mice were inoculated with Figure 5-2 Baytril treatment reduced B. pertussis ID50. 100, 500, 1,000, 5, 000, or 10,000 CFU B. Wild type C57BL/6 mice were either treated three times with Baytril (light grey) or left untreated (dark grey). All mice were then inoculated with various pertussis 536, and bacterial load was determined 3 doses of B. pertussis strain 536, and levels of nasal cavity colonization were assessed 3 days post- days post-inoculation. B. pertussis was able to inoculation. Error bars represent +/- one standard deviation. L. S. Weyrich, O. Y. Rolin, and C. Safi all contributed to this experiment. colonize mice treated with enrofloxacin when

inoculated with 100 CFU, but 10,000 CFU was

required to colonize mice with their microflora still intact (Figure 2). This finding indicates that

the microflora present in the upper respiratory tract of mice can inhibit low dose inoculation and

may additionally indicate that broad spectrum antibiotics may support B. pertussis infections.

To identify the species responsible for this in vivo inhibition, microflora present in uninfected littermates of mice utilized in the previous experiments were cultured on blood agar and identified by sequencing the 16S rRNA encoding genome region. In these untreated and uninfected mice, 8 different culturable organisms were identified, including Staphylococcus,

Kytococcus, Klebsiella, and Rhizobium species, including the day in which they were identified,

133 and the number of CFU present in the nasal cavity when originally isolated (Table 2). These data suggests that common

Table 2 Host microflora isolated from nasal cavities of healthy mice.

respiratory organisms isolated from mice are capable of inhibiting low dose colonization by B. pertussis in the murine nasal cavity.

Murine host microflora directly outcompete B. pertussis in vitro.

To investigate how murine nasal cavity flora could exclude B. pertussis in vivo, we utilized an in vitro system to cultured host microflora and B. pertussis organisms together. This assay examines fitness of an organism in the presence of another organism, under both nutrient rich and poor conditions. Exponential phase cultures of both B. pertussis and bacterial species identified in Table 2,

Staphylococcus saprophyticus, Figure 5-3 Murine nasal cavity flora inhibits B. pertussis growth. B. pertussis was grown alone or in the presence of either S. saprophyticus, S. lentus, S. xylous, or Klebsiella in SS media over Staphylococcus cohnii, Staphylococcus 32 hours, and surviving CFU were determined by plating on either LB (Staphylococcus) or BG with 60ng/mL streptomycin xylous, or Klebsiella PAMU, were (B. pertussis). H. A. Feaga completed this experiment.

134 inoculated into SS media in equal concentrations of 1×107 CFU/mL with shaking at 37°C for 36

hours. The growth of each host microflora species was unaffected by the presence of B.

pertussis compared to growth in single culture (data not shown). When B. pertussis was grown

alone, it grew to ~5×109 CFU. In comparison, B. pertussis only grew to ~1×108 CFU when it

was co-cultured with host organisms (Figure 3), suggesting murine nasal cavity flora are capable

of out-competing B. pertussis in vitro. Microflora could potentially directly inhibit B. pertussis

growth by releasing antmicrobial compounds or indirectly out-compete this pathogen by simply obtaining nutrients more efficiently.

Klebsiella and Staphylococcus species can prevent low dose B. pertussis colonization.

A recent survey investigating the use of oral probiotic organisms for respiratory disease prevention found that Lactobacillus species could shorten disease duration but not prevent disease [28]. Intra-nasal administration of Lactobacillus has been shown to protect against

Figure 5-4 Inoculation with known microflora inhibits B. pertussis colonization. C57BL/6 mice were treated with enrofloxacin and then inoculated with ~5,000 CFU of S. saprophyticus, S. lentus, S. xylous, or Klebsiella. A.) Colonization of these organisms was monitored at 0 and 3 days post- inoculation and at day 6 post-inoculation, which is 3 days after inoculation of ~5,000 CFU of B. pertussis. B.) B. pertussis colonization on day 6 post-inoculation of microflora and three days after B. pertussis inoculation was determined for each of the groups initially inoculated with individual host microflora species.

135 subsequent Streptococcus infection and lower influenza infections [29,30]. The finding

presented here suggests that murine microflora may be successful in preventing B. pertussis

infection. To test this hypothesis, wild type C57BL/6 mice were intranasally treated with

enrofloxacin and subsequently inoculated with ~1,000 CFU of either Klebsiella PAMU, S.

saprophyticus, S. lentus, or S. xylous. Three days after host microflora inoculation, microflora

levels were determined, and mice were then inoculated with 5,000 CFU of B. pertussis, a dose

only shown to colonize antibiotic treated mice (Figure 2). Levels of Klebsiella PAMU, S.

saprophyticus, S. lentus, or S. xylous were maintained through six days post-inoculation (Figure

4A), although Kytococcus was not able to persist past day three post-inoculation and was

therefore not included in this experiment (data not shown). However, even though B. pertussis

was inoculated at a dose that would have resulted in ~5,000 CFU in antibiotic treated animals, B.

pertussis was barely recovered from mice additionally inoculated with Klebsiella, S. saprophyticus, S. lentus, and S. xylous, at an average of ~10 CFU per mouse (Figure 4B). These

data suggest that murine microflora prevent B. pertussis colonization and indicate that

respiratory tract probiotic organisms may be used to prevent respiratory disease.

Bordetella bronchiseptica can displace murine microflora and colonize at low doses.

We hypothesize that pathogens may have evolved mechanisms to best compete

organisms within their own niche, in this case, the upper respiratory tract. Although murine

microflora block B. pertussis colonization, a contributing factor to this observation may be that

mice are not the natural host of B. pertussis. To investigate this observation, we decided to

examine how a very closely related murine pathogen, B. bronchiseptica, competes with murine

nasal cavity flora. These two strains, B. pertussis strain 536 and B. bronchiseptica strain RB50,

share 65% of their genome, and B. bronchiseptica is believed to the evolutionary progenitor –

136 like strain of B. bronchiseptica. Unlike B. pertussis, B. bronchiseptica has a very low LD50 and a very wide host range, suggesting that B. bronchiseptica may possesses mechanisms to compete with murine microflora. To test this, we again took 9 co-housed C57BL/6 littermates, dissected

3 healthy uninfected mice, and inoculated the remaining mice with 100 CFU of B. bronchiseptica, the same B. pertussis dose found to colonize antibiotic treated mice. Not only was B. bronchiseptica able to colonize the upper respiratory tract, as expected, it was also capable of displacing the culturable microflora observed in these animals (Figure 5A).

Additionally, we also examined the ability of B. parapertussis to interact with murine nasal

Figure 5-5 Zoonotic pathogen B. bronchiseptica can displace murine flora. Wild type C57BL/6 mice were inoculated with 100 CFU of B. bronchiseptica strain RB50 in 10 µL. Both host microflora (lines) and bordetellae (bars) colonization was assessed prior to inoculation and 3 days post-inoculation. Error bars represent +/- one standard deviation. cavity flora in vivo. B. pertussis is found to infect only humans or sheep, depending on the lineage and is therefore also believed to be host restricted. 100 CFU of B. parapertussis was inoculated into caged littermates, and levels of host microflora and bordetellae were monitored three days later. Interestingly, although B. parapertussis was able to colonize mice, unlike B. pertussis, it was not capable of displacing flora (Figure 5B). These experiments suggest that pathogens may have evolved the ability to compete with microflora present in their host

137 environment, and we can speculate that pathogens with broad host ranges may have the ability to displace various microflora types from various mammalian species.

138

Discussion

We set out to determine why human-adapted B. pertussis was unable to colonize the murine nasal cavity similarly to the zoonotic and closely related pathogen B. bronchiseptica. We uncovered that host microflora contribute to the ability of B. pertussis to colonize mice, and identified the microbiota responsible for this phenomena. Several murine flora isolates were capable of inhibiting B. pertussis growth in vitro and were additionally able to prevent B. pertussis colonization in vivo, indicating that these strains could be utilized in disease prevention.

However, a broad host range and known murine pathogen, B. bronchiseptica, was capable of

displacing all culturable murine microflora, suggesting that interactions between these pathogens

and microflora may contribute to host-adaptation and that mechanisms governing this competition may also contribute to a pathogens ability to infect a host.

The numerous species isolates from the murine nasal cavity in this study are diverse, but many of them mimic what is observed in human nasal cavities. Staphylococcus, Kytotococcus, and Klebsiella species are commonly observed in the upper respiratory tract of humans

[31,31,32]. This research additionally suggests that these interactions may be true in humans. It is plausible that humans receiving broad spectrum antibiotics may be more susceptible to B. pertussis infections. Additionally, it may provide another explanation, beyond insufficient immunity, as to why B. pertussis is successful in newborn infants that have yet to be fully colonized by environmental microflora [31]. Although the majority of newborn flora is obtained during the birthing process from the birth canal, this may contain particular species capable of inhibiting B. pertussis infection, such as Staphylococcus or Klebsiella. Novel strategies involving introduction of upper respiratory tract probiotics very early on in life, or

139 even after the administration of antibiotics, may be a new preventative measure against B. pertussis infection. However, determining a subset of microflora to include as respiratory probiotics would be difficult. The intranasal administration of Lactobacillus has also been shown to decrease the severity of influenza infections, but patients who took an oral probiotic cocktail of Lactobacillus, Bifidobacterium, and Streptococccus reduced Staphylococcus aureus,

Streptococcus pneumoniae, ß-hemolytic streptococci, and Haemophilus influenzae in the upper respiratory tract, most plausibly through an immune response induced mechanism [30,33].

Although particular bacterial species may be beneficial against a subset of pathogens, that species may additionally inhibit closely related species essential in protecting against B. pertussis infection, such as S. saprophyticus, S. lentus, S. xylous, or Klebsiella species. Both immune mediated mechanisms and direct competition, as our study demonstrates, must be taken into consideration when creating effective respiratory tract probiotics for disease prevention.

Interestingly, we isolated a Rhizboium species from the upper respiratory tract of mice, on more than one occasion. Rhizobium species are normally thought to be plant pathogens or symbiotes, but not murine nasal cavity flora, so we hypothesize that it may have been obtained from the food in our breed facility, which is partially plant based [34]. Observations such as this may mimic the diversity and fluidity currently observed in the human microbiome [35]. How uncultured, or currently unknown, respiratory organisms contribute to these interactions is yet to be determined. Furthermore, understanding interactions on a bacterial community scale may result in very different observations.

In this study, we observed that several murine host microflora species were capable of out-competing or even killing B. pertussis in vitro. The mechanisms behind this interaction remain unclear, but we can speculate that B. pertussis may lack factors necessary for microbial

140 competition. For example, the Type VI Secretion System (T6SS) in Pseudomonas, Vibrio, and

Serratia has been shown to be required for microbial competition [36–38]. Recently, our laboratory identified the T6SS loci in bordetellae, and additionally observed that this locus is missing in B. pertussis [39]. Interestingly, a functional T6SS locus in B. bronchiseptica was identified, and here, we observed that this zoonotic pathogen is capable of competing with nasal cavity flora in vivo. Although we can speculate that the T6SS plays a role during in vivo microbial competition with bordetellae, almost certainly additional bacterial factors and mechanisms contribute to this phenomenon. In this study, we observed that B. pertussis was capable of competing with human microflora, suggesting that it possesses some factors required for microbial competition, but perhaps not the same ones B. bronchiseptica uses in mice.

Experiments examining the competitions between B. pertussis mutants lacking vital in vivo virulence factors may identify bordetellae factors essential for microbial competition.

We observed that human adapted B. pertussis is not capable of competing with murine flora in vitro or in vivo, while a natural murine pathogen and the evolutionary-like progenitor strain of B. pertussis, B. bronchiseptica, is able to out-compete murine microflora through a yet unknown mechanism. Together, these observations suggest that pathogens may evolve or acquire mechanisms to compete with the microflora only present in their niche environment.

These results may also suggest that B. pertussis is only capable of out-competing lower respiratory tract microflora and infects its host by first establishing the trachea. Interestingly, B. pertussis is believed to have evolved from B. bronchiseptica by genome loss, in which human host adaptation occurred. It is possible that the genes required to compete with murine flora were also loss during this transition into human adaptation. This also suggests that unique mechanisms may be required to displace different bacterial communities or even the same

141 species isolated from different hosts. Whole genome sequencing of individual microflora strains or metagenomic sequencing comparisons between host species may elucidate major difference in pathogen host specificity and how these animal species use microbiota to protect themselves against infection. Our data additionally suggests that new animal models could be made by the administration of antibiotics into a non-human animal.

This study examines why B. pertussis is incapable of colonizing mice at low doses, demonstrating that murine nasal cavity flora are capable of out-competing this human pathogen and blocking colonization. We additionally show that these species are capable of blocking B. pertussis growth, suggesting numerous uses for them in novel anti-microbial research or as a prophylactic treatment against B. pertussis. Lastly, our work suggests that pathogens have evolved mechanisms to compete with microflora within their natural host environment, and that host microflora may contribute to pathogen host-specificity.

142

Author Contributions:

Authors: Laura S. Weyrich1,2, Heather A. Feaga1,2, Jihye Park1,4, Chetan Y. Safi1, Sarah J.

Muse1,2, Sarah E. Young1, Olivier Y. Rolin1,3, and Eric T. Harvill1

1Department of Veterinary and Biomedical Sciences, The Pennsylvania State University

W-213 Millennium Science Complex, University Park, PA 16802

2Graduate Program in Biochemistry, Microbiology, and Molecular Biology

3 Graduate Program in Immunology and Infectious Disease

4 Graduate Program in Bioinformatics and Genomics

Conceived and designed experiments: LSW, HAF, OYR, ETH

Preformed experiments: LSW, HAF, OYR, SJM, JP, CYS, SEY

Analyzed data: LSW, HAF, ETH

Wrote manuscript: LSW, ETH

143

References

1. WHO | The world health report 2004 - changing history (2004). Available: http://www.who.int/whr/2004/en/. Accessed 12 Jan 2012.

2. Tan T, Trindade E, Skowronski D (2005) Epidemiology of Pertussis. The Pediatric Infectious Disease Journal 24. Available: http://journals.lww.com/pidj/Fulltext/2005/05001/Epidemiology_of_Pertussis.3.aspx.

3. Cherry JD (1996) Historical Review Of Pertussis And The Classical Vaccine. Journal of Infectious Diseases 174: S259-S263. doi:10.1093/infdis/174.Supplement_3.S259

4. Center for Disease Control and Prevention (2008) Pertussis.. Available: http://www.cdc.gov/vaccines/pubs/pinkbook/downloads/pert.pdf. Accessed 12 Jan 2012.

5. Pam D (2002) Whooping cough makes global comeback. The Lancet Infectious Diseases 2: 322. doi:10.1016/S1473-3099(02)00308-0

6. Mielcarek N, Debrie A-S, Raze D, Bertout J, Rouanet C, et al. (2006) Live Attenuated B. pertussis as a Single-Dose Nasal Vaccine against Whooping Cough. PLoS Pathog 2: e65. doi:10.1371/journal.ppat.0020065

7. Goodnow RA (1980) Biology of Bordetella bronchiseptica. Microbiol Rev 44: 722-38.

8. Mattoo S, Cherry JD (2005) Molecular Pathogenesis, Epidemiology, and Clinical Manifestations of Respiratory Infections Due to Bordetella pertussis and Other Bordetella Subspecies. Clin. Microbiol. Rev. 18: 326-382. doi:10.1128/CMR.18.2.326-382.2005

9. Wolfe DN, Kirimanjeswara GS, Harvill ET (2005) Clearance of Bordetella parapertussis from the lower respiratory tract requires humoral and cellular immunity. Infect Immun 73: 6508-13.

10. Mann PB, Wolfe D, Latz E, Golenbock D, Preston A, et al. (2005) Comparative toll-like receptor 4-mediated innate host defense to Bordetella infection. Infect Immun 73: 8144-52.

11. Andreasen C, Powell DA, Carbonetti NH (2009) Pertussis Toxin Stimulates IL-17 Production in Response to Bordetella pertussis Infection in Mice. PLoS ONE 4: e7079. doi:10.1371/journal.pone.0007079

12. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of antibodies in immunity to Bordetella infections. Infect Immun 71: 1719-24.

13. Heininger U, Cotter PA, Fescemyer HW, Martinez de Tejada G, Yuk MH, et al. (2002) Comparative phenotypic analysis of the Bordetella parapertussis isolate chosen for genomic sequencing. Infect Immun 70: 3777-84.

144 14. Harvill ET, Cotter PA, Miller JF (1999) Pregenomic comparative analysis between bordetella bronchiseptica RB50 and Bordetella pertussis tohama I in murine models of respiratory tract infection. Infect Immun 67: 6109-18.

15. Mills KH (2001) Immunity to Bordetella pertussis. Microbes Infect. 3: 655-677.

16. Stieritz DD, Holder IA (1975) Experimental Studies of the Pathogenesis of Infections Due to Pseudomonas aeruginosa: Description of a Burned Mouse Model. Journal of Infectious Diseases 131: 688 -691. doi:10.1093/infdis/131.6.688

17. Tang HB, DiMango E, Bryan R, Gambello M, Iglewski BH, et al. (1996) Contribution of specific Pseudomonas aeruginosa virulence factors to pathogenesis of pneumonia in a neonatal mouse model of infection. Infection and Immunity 64: 37 -43.

18. Motley ST, Lory S (1999) Functional Characterization of a Serine/Threonine Protein Kinase of Pseudomonas aeruginosa. Infection and Immunity 67: 5386 -5394.

19. Stibitz S, Yang MS (1991) Subcellular localization and immunological detection of proteins encoded by the vir locus of Bordetella pertussis. J. Bacteriol. 173: 4288-4296.

20. BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. (1994). Available: http://medjournal.hmc.psu.edu:2060/pmc/articles/PMC302969/?tool=pubmed. Accessed 24 Jul 2011.

21. Pishko EJ, Kirimanjeswara GS, Pilione MR, Gopinathan L, Kennett MJ, et al. (2004) Antibody-mediated bacterial clearance from the lower respiratory tract of mice requires complement component C3. Eur J Immunol 34: 184-93.

22. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of Antibodies in Immunity to Bordetella Infections. Infect. Immun. 71: 1719-1724. doi:10.1128/iai.71.4.1719-1724.2003

23. Weisburg WG, Barns SM, Pelletier DA, Lane DJ (1991) 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173: 697-703.

24. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596-9. doi:msm092 [pii] 10.1093/molbev/msm092

25. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J. Mol. Biol 215: 403-410. doi:10.1006/jmbi.1990.9999

26. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, et al. (2009) Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities. Appl. Environ. Microbiol. 75: 7537- 7541. doi:

10.1128/AEM.01541-09

145 27. Huson DH, Auch AF, Qi J, Schuster SC (2007) MEGAN analysis of metagenomic data. Genome Res 17: 377-86. doi:gr.5969107 [pii] 10.1101/gr.5969107

28. Vouloumanou EK, Makris GC, Karageorgopoulos DE, Falagas ME (2009) Probiotics for the prevention of respiratory tract infections: a systematic review. International Journal of Antimicrobial Agents 34: 197.e1-197.e10. doi:10.1016/j.ijantimicag.2008.11.005

29. Cangemi de Gutierrez R, Santos V, Nader‐Macías ME (2001) Protective effect of intranasally inoculated Lactobacillus fermentum against Streptococcus pneumoniae challenge on the mouse respiratory tract. FEMS Immunology & Medical Microbiology 31: 187-195. doi:10.1111/j.1574-695X.2001.tb00519.x

30. Hori T, Kiyoshima J, Shida K, Yasui H (2001) Effect of Intranasal Administration ofLactobacillus casei Shirota on Influenza Virus Infection of Upper Respiratory Tract in Mice. Clinical and Diagnostic Laboratory Immunology 8: 593 -597. doi:10.1128/CDLI.8.3.593-597.2001

31. Normal Flora - Medical Microbiology - NCBI Bookshelf (n.d.). Available: http://www.ncbi.nlm.nih.gov/books/NBK7617/. Accessed 8 Jan 2012.

32. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, et al. (2007) The Human Microbiome Project. Nature 449: 804-810. doi:10.1038/nature06244

33. Glück U, Gebbers J-O (2003) Ingested probiotics reduce nasal colonization with pathogenic bacteria (Staphylococcus aureus, Streptococcus pneumoniae, and ß-hemolytic streptococci). The American Journal of Clinical Nutrition 77: 517 -520.

34. Van Berkum P, Beyene D, Eardly BD (1996) Phylogenetic Relationships among Rhizobium Species Nodulating the Common Bean (Phaseolus vulgaris L.). International Journal of Systematic Bacteriology 46: 240 -244. doi:10.1099/00207713-46-1-240

35. Caporaso JG (2011) Moving pictures of the human microbiome. Genome Biology 12. doi:10.1186/gb-2011-12-5-r50

36. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, et al. (2011) Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475: 343-347. doi:10.1038/nature10244

37. MacIntyre DL, Miyata ST, Kitaoka M, Pukatzki S (2010) The Vibrio cholerae type VI secretion system displays antimicrobial properties. Proceedings of the National Academy of Sciences 107: 19520 -19524. doi:10.1073/pnas.1012931107

38. Murdoch SL, Trunk K, English G, Fritsch MJ, Pourkarimi E, et al. (2011) The Opportunistic Pathogen Serratia marcescens Utilizes Type VI Secretion To Target Bacterial Competitors. J. Bacteriol. 193: 6057-6069. doi:10.1128/JB.05671-11

146 39. Weyrich, L.S., Rolin, O. Y., Muse, S. J., Park, J., Spidale, N., et al. (2012) A Type VI Secretion System Encoding Locus is Required for Bordetella bronchiseptica Immunomodulation and Persistence In Vivo. PLoS One Submitted.

147

Chapter 6 : Bordetella bronchiseptica Small RNA Molecules are Predicted to Target

Major Virulence Factors and Contribute to Strain Specificity

148

Abstract

Small, non-coding RNA (sRNA) molecules have been shown to contribute to the virulence of numerous respiratory pathogens, such as Pseudomonas, Streptococcus, and

Mycobacterium; however, how these sRNA molecules might contribute to the virulence of the broad host ranged respiratory pathogen, Bordetella bronchiseptica, remains undetermined. Here, bioinformatic prediction and SOLiD sequencing were used to identify 181 novel sRNAs in B. bronchiseptica. Several of these sRNAs were shown to be regulated by σE, similarly to

Escherichia coli, suggesting they may play a role in cell envelope stress responses. We predicted putative targets for each of these sRNAs, identifying the transcripts encoding the Type

II, III, and VI Secretion Systems, adenylate cyclase toxin, fimbriae, filamentous hemagglutinin,

O-Antigen, and BvgS as putative targets. Two sRNAs, bbsrna28 and bbsrna70 are believed to target a novel, putative T6SS effector, tssJ. Lastly, 138 of these sRNAs were identified in the genomes of closely related human pathogens, Bordetella pertussis and Bordetella parapertussis, suggesting they are highly conserved. We additionally determined that the 6 sRNA molecules unique to B. bronchiseptica may contribute to strain specificity and provide explanations to unique observations in the bordetellae field.

149

Introduction

Currently, the Bordetella genus includes nine species of fastidious, Gram-negative

coccobacilli that can cause a wide range of respiratory diseases. Through massive gene loss, B.

pertussis and B. parapertussis, the causative agents of whooping cough, are believed to have

diverged from a B. bronchiseptica like-progenitor, creating the three classical bordetellae [1,2].

B. bronchiseptica is known to infect diverse mammalian hosts, including immunocompromised humans, and its colonization leads to various disease severities, from asymptomatic carriage to fatal pneumonia [3–5]. Since laboratory animals are natural B. bronchiseptica hosts, studies analyzing B. bronchiseptica host-pathogen interactions have provided insight into the pathogenesis of the two human pathogens [3,4,6].

In the host, certain factors, such as pertussis toxin (Ptx), adenylate cyclase toxin (ACT), filamentous hemagglutinin (FHA), pertactin (PRN), fimbriae (Fim2 and Fim3), dermonecrotic toxin (DNT), tracheal cytotoxin (TCT), Type III Secretion System (T3SS), and Type VI

Secretion System (T6SS), must be up-regulated in order for Bordetella to cause disease [7–10].

Generally, this global up-regulation of virulence genes occurs through a two-component, phospho-relay system, BvgAS, which involves a sensor kinase and a transcriptional regulator

[11]. The BvgAS system is turned on during in vitro growth or infection and activates transcription of virulence associated genes; down-regulation of virulence repressed genes is also achieved by the addition of a third protein, BvgR [12–14]. BvgA and BvgR down-regulate environmentally associated genes, such as motility, metabolism, and transporter loci during infection, although the requirements of these genes during infection are much less known [11].

Because this global regulatory network controls so many of the widely studied Bordetella

150 virulence factors, little research has explored regulatory mechanisms secondary to BvgAS. We

hypothesize that B. bronchiseptica possess virulence factor regulatory mechanisms that are

secondary or alternative to the global BvgAS system, such as untranslated, small RNA (sRNA)

molecules.

In multiple other Gram-negative pathogens, a sRNA regulatory system helps to “fine

tune” the regulation exhibited by the global regulatory systems, providing post transcriptional

regulation [15,16]. These untranslated, regulatory RNAs modulate messenger RNAs (mRNAs) by binding a complementary target sequence or protein to alter the protein function. Cis or trans sRNA molecules positively or negatively regulate their targets by three main mechanisms: altering mRNA stability, targeting mRNAs for degradation, or increasing translation [17–19].

Many sRNAs additionally use a chaperone protein, Hfq, to find their target. To date, the majority of work identifying prokaryotic sRNAs, determining sRNA regulatory mechanisms, and deducing target mRNAs has been done in Escherichia coli and Salmonella. For example, the E. coli sRNA, micF, inhibits production of an outer membrane protein, OmpF, which has been implicated in adherence in Enteropathogenic E. coli [20]. Since then, sRNAs involved in iron homeostasis, metabolic pathways, and pathogenesis have been characterized in various species, including Escherichia, Salmonella, Shigella, Yersinia, Vibrio, Mycobacterium, Streptococcus, and Staphyloccocus [17,21–29]. Several mutants of the sRNA chaperone protein, Hfq, have resulted in decreased virulence, suggesting that sRNAs are critical for virulence factor regulation

[30–33]. Most recently, 23 candidate sRNA molecules in B. pertussis were identified in silico and verified by Northern blotting [34]. However, no work has been done to investigate how sRNAs directly contribute to Bordetella pathogenesis and virulence factor gene regulation.

151 Furthermore, total RNA sequencing of Bordetella sRNAs has not been completed, suggesting

there are many novel, unidentified sRNAs.

Here, we seek to identify sRNA molecules present in B. bronchiseptica, the evolutionary

progenitor-like species of B. pertussis and B. parapertussis, and better understand the role sRNAs play in Bordetella pathogenesis. sRNA identification protocol using high-throughput

technologies (SIPHT) prediction software and SOLiD RNA sequencing were employed to

identify 181 sRNA molecules in B. bronchiseptica strain RB50. As expected, sRNAs were

evenly distributed throughout the genome and were >20 to <900 bp in size. TargetPredict

software was then used to predict mRNA targets for each sRNA, identifying several sRNAs

predicted to target genes encoding ACT, FHA, Fim, T3SS, T6SS, and the flagellar apparatus.

Quantitative RT-PCR confirmed that two putative sRNAs are transcribed in coordination with a

putative T6SS effector protein. Furthermore, we confirmed the presence of 138 non-coding

sRNAs in B. pertussis and B. parapertussis, suggesting these small regulatory molecules are well conserved. Together, this data suggests that B. bronchiseptica possesses a vast array of sRNA molecules that have the potential to contribute to B. bronchiseptica pathogenesis.

152

Materials and Methods

Bacterial Strains

Wild type B. bronchiseptica strain RB50 and a Bvg- phase locked derivative, strain

RB54, were obtained from Jeff Miller and are previously described [35]. All strains were maintained on Bordet-Gengou (BG) agar (Difco) containing 10% sheep's blood (Hema

Resources) and 20 μg/ml streptomycin (Sigma) [36]. Liquid cultures were grown in Stainer

Scholte media overnight at 37°C with shaking until cultures reached logarithmic phase (OD

~0.7), as measured by optical density at 600 nm [37].

RNA Isolation and Library Preparation

Bacteria for RNA isolation were collected from a liquid culture of B. bronchiseptica strain RB50 in logarithmic phase (OD ~0.7) and placed immediately into RNA Later

(Invitrogen). Total RNA was extracted using the mirVana miRNA Isolation Kit (Ambion) following the total RNA protocol that was provided. RNA was stored at -80°C until sequencing and library preparation at The Genomic Core Facility at The Pennsylvania State University in

University Park, PA, USA. Briefly, size separation was performed on a flashPAGE Fractionator

(Ambion), and RNA fragments smaller than 45 bp were collected and used for library preparation according to the manufacturer’s instructions (Applied Biosystems).

SOLiD Sequencing and Analysis

All sequencing was completed at the Genomic Core Facility at The Pennsylvania State

University in University Park, PA, USA. The SOLiD 3 sequencing platform were used to obtain

3,103,032 reads for strain RB50. 27% of the sequencing reads from SOLiD were mapped onto the reference genome of B. bronchiseptica RB50 with SHort Read Mapping Package (SHRiMP)

153 using the default setting [38]. The start and end position of coding sequences as well as the

strand information (positive or negative) in reference RB50 genome was downloaded from

National Center for Biotechnology Information (NCBI) database. Reads mapping to known

genes, tRNAs, and rRNAs in the annotated direction were excluded. Reads mapping to the

opposite orientation of the annotated gene, tRNA, or rRNA were maintained. The remaining

reads were concatenated when at least one bp overlapped on the same strand. The concatenated

sequences that were mapped at least 50 times on the reference genome were labeled “putative

sRNAs.”

sRNA identification protocol using high-throughput technologies (SIPHT) Prediction.

SIPHT prediction on B. bronchiseptica strain RB50 was done as previously described

[39]. To determine which sRNAs are present in both the SIPHT-predicted and our sequencing analysis, the genomic position of each sequenced and predicted sRNA were compared, and overlap between these positions on the same strand was used as criteria to determine counted as common sRNA candidates between two analyses.

Genome Context and Strain Relatedness Analysis

sRNA candidate sequences were then checked with the reference genome, and reference genome sequences were used for the further analysis. These candidate sequences were first reverse complemented, and the position of each sRNA was determined by sequence similarity, using standalone BLASTN, allowing 10 mismatches [40]. Distances between the sRNA

candidate and the nearest genes were calculated as the difference between the ends of the

putative sRNA and the ORF closest to each other on the same strand. To determine which

sRNAs sequenced from B. bronchiseptica are also found in the other classical bordetellae

154 genomes, each sRNA sequence was used to search B. parapertussis (taxid:519) and B. pertussis

(taxid:520) genomes by BLASTN (NCBI).

Northern Blotting

Bacterial RNA was isolated by using the TRIzol Max Bacterial RNA Isolation Kit

(Invitrogen) following the recommended protocol. The RNA quality and concentration were determined by 1.2% agarose-formaldehyde gel electrophoresis and readings of absorbance at 260 and 280nm, using a NanoDrop 2000 Spectrophotometer (Thermo Scientific). Northern blotting was performed by denaturing total RNA (10μg per lane) and electrophoresis on a 6% polyacrylamide denaturing gel (National Diagnostics) containing 7.5M urea. RNA was transferred to Hybond-N+ nylon membranes (GE Healthcare) and UV cross-linked. Membranes were prehybridized in 5X SSC/0.1% SDS followed by a 4 hour hybridization of [γ-32P]ATP radiolabeled oligonucleotide probes at 42-45°C. Washes of 15 minutes each with 2xSSC/0.1%

SDS, 1xSSC/0.1% SDS, and 0.1xSSC were performed before detection of signal with a Typhoon imaging system (GE Healthcare).

TargetRNA Analysis

To predict messenger RNA (mRNA) targets for each sRNA, each candidate sRNA was analyzed using TargetRNA, as previously described [41]. Briefly, potential targets for each sRNA candidate in the B. bronchiseptica genome were determined by the ability for basepair binding after RNA folding conformations. If available, the top twenty hits were all considered valid.

155

Results

SIPHT predicts over 450 B. bronchiseptica specific small RNA molecules.

Many small RNAs present in pathogens have first been identified by their prediction,

then later verification via hybridization. SIPHT, developed by Livny et al., was used to predict

possible sRNA molecules in B. bronchiseptica [25]. This approach identified 450 small RNA

molecules in B. bronchiseptica strain RB50 (Appendix A). These predicted sRNAs span the

genome and do not appear to cluster in any specific area. This prediction suggests that sRNA

molecules may be present and play some role in the ability of B. bronchiseptica to successfully

regulate its genes.

SOLiD sequencing identifies 181 B. bronchiseptica sRNAs.

Because a large number of sRNAs were predicted by SIPHT, we utilized SOLiD3

sequencing to identify small RNAs expressed during exponential growth phase, when bordetellae

are believed to be in Bvg+ phase, expressing the majority of their known virulence factors. Over

3 million total reads were obtained, and 27% of the reads mapped directly to the B. bronchiseptica RB50 genome, using a Short Read Mapping Package (SHRiMP) [38]. Any

sequence that mapped to a known messenger, transfer, or ribosomal RNA was dismissed

utilizing an in-house pipeline designed by Y. Zhang. Additionally, small RNAs encoded within

a messenger RNA but transcribed in an opposite direction of the mRNA were obtained using

another in-house pipeline. Overlapping sequences were concatenated to determine start and end

sites. This analysis identified 158 non-coding sRNAs and 23 sRNAs encoded within messenger

RNAs, resulting in a total of 181 putative B. bronchiseptica sRNA molecules (Table 3). First, we compared our putative sRNA molecules to those identified via SIPHT using mySQL, and

156 Table 3 Small RNA molecules identified in B. bronchiseptica strain RB50. Small RNAs identified by SOLiD sequencing were named in order of their coverage. The genome start and end position for each sRNA, the strand position, and overall predicted length are listed. The read coverage for the concatenated sequence and per base pair of each sRNA were calculated. Coverage was also compared to expression calculated for the 16S ribosomal RNA. J. Park analyzed this data.

Name Start End Strand Length Coverage Coverage Expression per base Compared to 16S RNA Expression bbsrna1 4484877 4484974 - 98 4041 41.234694 12.860659 bbsrna1 3620838 3620935 - 98 4041 41.234694 12.860659 bbsrna1 1974453 1974550 + 98 4041 41.234694 12.860659 bbsrna2 1972480 1972918 + 439 2975 6.7767654 2.1136005 bbsrna2 3622470 3622908 - 439 2974 6.7744875 2.11289 bbsrna2 4486509 4486947 - 439 2973 6.7722096 2.1121796 bbsrna3 4493425 4493727 - 303 837 2.7623762 0.8615555 bbsrna4 5147238 5147452 - 215 644 2.9953488 0.9342172 bbsrna5 928735 928936 + 202 557 2.7574257 0.8600115 bbsrna6 1534240 1534372 + 133 556 4.1804511 1.3038379 bbsrna7 2206723 2206863 - 141 501 3.5531915 1.1082023 bbsrna8 3617287 3617457 - 171 451 2.6374269 0.8225852 bbsrna8 1977931 1978101 + 171 451 2.6374269 0.8225852 bbsrna8 4481326 4481496 - 171 447 2.6140351 0.8152895 bbsrna9 4484503 4484641 - 139 440 3.1654676 0.9872754 bbsrna9 3620464 3620602 - 139 440 3.1654676 0.9872754 bbsrna9 1974786 1974924 + 139 440 3.1654676 0.9872754 bbsrna10 4073804 4073886 - 83 414 4.9879518 1.5556887 bbsrna11 4401795 4401862 - 68 411 6.0441176 1.8850955 bbsrna12 4633004 4633137 + 134 326 2.4328358 0.7587754 bbsrna13 3601640 3601836 + 197 309 1.5685279 0.489207 bbsrna14 3719812 3601836 + 135 294 2.1777778 0.6792255 bbsrna15 1456010 3601836 - 89 283 3.1797753 0.9917378 bbsrna16 2769741 3601836 + 59 277 4.6949153 1.4642937 bbsrna17 1761259 3601836 + 48 273 5.6875 1.7738703 bbsrna18 4397155 3601836 - 57 247 4.3333333 1.3515202 bbsrna19 5233190 3601836 + 274 205 0.7481752 0.2333478 bbsrna20 4234331 3601836 - 158 199 1.2594937 0.3928226 bbsrna21 4636206 3601836 - 63 195 3.0952381 0.9653716 bbsrna22 2767242 3601836 + 176 184 1.0454545 0.3260661 bbsrna23 1054449 3601836 + 140 182 1.3 0.4054561 bbsrna24 170646 3601836 - 56 179 3.1964286 0.9969318 bbsrna25 4148588 3601836 - 45 178 3.9555556 1.2336954 bbsrna26 4804958 3601836 + 109 177 1.6238532 0.5064624 bbsrna27 3604725 3601836 + 182 168 0.9230769 0.2878978

157 bbsrna28 2736123 3601836 + 467 168 0.359743 0.1122 bbsrna29 4834896 3601836 + 62 160 2.5806452 0.8048756 bbsrna30 2767205 3601836 - 168 155 0.922619 0.287755 bbsrna31 4827227 3601836 + 36 151 4.1944444 1.3082023 bbsrna32 217610 3601836 - 93 145 1.5591398 0.486279 bbsrna33 1978220 3601836 + 247 144 0.582996 0.1818302 bbsrna34 4032089 3601836 - 36 142 3.9444444 1.2302299 bbsrna35 4117374 3601836 - 78 141 1.8076923 0.5637999 bbsrna36 535582 3601836 - 95 139 1.4631579 0.4563433 bbsrna37 3131675 3601836 - 84 139 1.6547619 0.5161025 bbsrna38 2558978 3601836 - 72 139 1.9305556 0.6021196 bbsrna39 2403491 3601836 - 71 139 1.9577465 0.6106001 bbsrna40 1427632 3601836 - 54 138 2.5555556 0.7970504 bbsrna41 3379965 3601836 - 108 137 1.2685185 0.3956373 bbsrna42 3819649 3601836 - 50 136 2.72 0.8483388 bbsrna43 26212 3601836 - 167 134 0.8023952 0.2502585 bbsrna44 530604 3601836 - 55 132 2.4 0.7485343 bbsrna45 4235832 3601836 - 93 130 1.3978495 0.4359743 bbsrna46 2117387 3601836 + 65 128 1.9692308 0.614182 bbsrna47 2384958 3601836 + 57 127 2.2280702 0.6949112 bbsrna48 395761 3601836 + 60 124 2.0666667 0.6445712 bbsrna49 3169500 3601836 + 78 124 1.5897436 0.495824 bbsrna50 1359516 3601836 + 81 124 1.5308642 0.4774601 bbsrna51 5247657 3601836 + 91 123 1.3516484 0.4215646 bbsrna52 4223871 3601836 - 38 123 3.2368421 1.0095364 bbsrna53 3913443 3601836 + 62 118 1.9032258 0.5935957 bbsrna54 4178156 3601836 - 113 117 1.0353982 0.3229296 bbsrna55 1041354 3601836 - 86 117 1.3604651 0.4243145 bbsrna56 3249958 3601836 - 165 116 0.7030303 0.2192676 bbsrna57 2100138 3601836 - 58 116 2 0.6237786 bbsrna58 3314666 3601836 - 75 115 1.5333333 0.4782302 bbsrna59 1761266 3601836 - 50 115 2.3 0.7173453 bbsrna60 4223117 3601836 - 64 114 1.78125 0.5555528 bbsrna61 3885556 3601836 - 120 112 0.9333333 0.2910967 bbsrna62 5254046 3601836 - 37 111 3 0.9356678 bbsrna63 2721230 3601836 - 111 111 1 0.3118893 bbsrna64 1504333 3601836 + 65 111 1.7076923 0.5326109 bbsrna65 2703745 3601836 + 135 110 0.8148148 0.254132 bbsrna66 4537607 3601836 - 86 109 1.2674419 0.3953015 bbsrna67 2379049 3601836 + 119 106 0.8907563 0.2778173 bbsrna68 2268339 3601836 + 93 105 1.1290323 0.3521331 bbsrna69 100532 3601836 + 201 105 0.5223881 0.1629272 bbsrna70 2327272 3601836 + 39 103 2.6410256 0.8237076

158 bbsrna71 1496239 3601836 - 220 102 0.4636364 0.1446032 bbsrna72 2298090 3601836 + 142 98 0.6901408 0.2152475 bbsrna73 4454106 3601836 - 76 96 1.2631579 0.3939654 bbsrna74 4695015 3601836 + 78 94 1.2051282 0.3758666 bbsrna75 63114 3601836 + 62 85 1.3709677 0.4275901 bbsrna76 3324932 3601836 - 44 84 1.9090909 0.595425 bbsrna77 2484722 3601836 + 36 84 2.3333333 0.7277417 bbsrna78 4365504 3601836 - 41 83 2.0243902 0.6313856 bbsrna79 3851536 3601836 + 54 83 1.537037 0.4793854 bbsrna80 2110463 3601836 - 75 83 1.1066667 0.3451575 bbsrna81 886765 3601836 - 45 79 1.7555556 0.547539 bbsrna82 3687177 3601836 - 71 79 1.1126761 0.3470317 bbsrna83 2842702 3601836 - 204 78 0.3823529 0.1192518 bbsrna84 4938829 3601836 - 46 76 1.6521739 0.5152953 bbsrna85 3979063 3601836 - 86 76 0.8837209 0.2756231 bbsrna86 176964 3601836 + 79 76 0.9620253 0.3000454 bbsrna87 5151360 3601836 + 148 74 0.5 0.1559446 bbsrna88 4661821 3601836 + 195 74 0.3794872 0.118358 bbsrna89 4429341 3601836 - 50 74 1.48 0.4615961 bbsrna90 321225 3601836 + 71 73 1.028169 0.3206749 bbsrna91 2694629 3601836 + 127 73 0.5748031 0.1792749 bbsrna92 4666337 3601836 + 123 72 0.5853659 0.1825693 bbsrna93 2111745 3601836 - 125 72 0.576 0.1796482 bbsrna94 535509 3601836 + 172 70 0.4069767 0.1269317 bbsrna95 901479 3601836 + 37 69 1.8648649 0.5816314 bbsrna96 62968 3601836 - 91 69 0.7582418 0.2364875 bbsrna97 3983228 3601836 - 41 69 1.6829268 0.5248868 bbsrna98 3978590 3601836 - 144 69 0.4791667 0.1494469 bbsrna99 3679824 3601836 + 34 69 2.0294118 0.6329518 bbsrna100 6317 3601836 + 110 68 0.6181818 0.1928043 bbsrna101 530826 3601836 - 41 68 1.6585366 0.5172798 bbsrna102 5272553 3601836 + 40 68 1.7 0.5302118 bbsrna103 4479641 3601836 - 93 68 0.7311828 0.2280481 bbsrna104 2704810 3601836 + 111 67 0.6036036 0.1882575 bbsrna105 4647582 3601836 - 48 66 1.375 0.4288478 bbsrna106 4694954 3601836 - 52 65 1.25 0.3898616 bbsrna107 4122916 3601836 - 181 64 0.3535912 0.1102813 bbsrna108 1103656 3601836 - 46 64 1.3913043 0.4339329 bbsrna109 3518771 3601836 - 71 62 0.8732394 0.272354 bbsrna110 4912409 3601836 - 127 61 0.480315 0.1498051 bbsrna111 42201 3601836 + 217 61 0.281106 0.0876739 bbsrna112 3164003 3601836 - 96 61 0.6354167 0.1981796 bbsrna113 300128 3601836 - 73 61 0.8356164 0.2606198

159 bbsrna114 2736205 3601836 - 128 61 0.4765625 0.1486347 bbsrna115 11542 3601836 + 82 61 0.7439024 0.2320152 bbsrna116 2632555 3601836 + 42 60 1.4285714 0.4455561 bbsrna117 4672037 3601836 - 37 59 1.5945946 0.497337 bbsrna118 2702324 3601836 - 101 59 0.5841584 0.1821927 bbsrna119 3478166 3601836 - 113 58 0.5132743 0.1600848 bbsrna120 2171928 3601836 + 170 58 0.3411765 0.1064093 bbsrna121 886737 3601836 + 74 57 0.7702703 0.240239 bbsrna122 5337023 3601836 + 130 57 0.4384615 0.1367515 bbsrna123 4005152 3601836 + 69 57 0.826087 0.2576477 bbsrna124 3824032 3601836 + 72 57 0.7916667 0.2469123 bbsrna125 2229521 3601836 + 121 57 0.4710744 0.1469231 bbsrna126 2101054 3601836 - 45 57 1.2666667 0.3950598 bbsrna127 4816284 3601836 + 46 56 1.2173913 0.3796913 bbsrna128 2874729 3601836 - 35 56 1.6 0.4990229 bbsrna129 2378229 3601836 + 37 56 1.5135135 0.4720486 bbsrna130 4114378 3601836 - 56 55 0.9821429 0.3063198 bbsrna131 152712 3601836 - 55 55 1 0.3118893 bbsrna132 97393 3601836 - 74 54 0.7297297 0.2275949 bbsrna133 49743 3601836 + 74 54 0.7297297 0.2275949 bbsrna134 4910616 3601836 - 116 54 0.4655172 0.1451898 bbsrna135 2964013 3601836 + 43 54 1.255814 0.3916749 bbsrna136 189040 3601836 - 86 54 0.627907 0.1958375 bbsrna137 1175152 3601836 - 72 54 0.75 0.233917 bbsrna138 4494700 3601836 + 234 53 0.2264957 0.0706416 bbsrna139 3953167 3601836 + 60 53 0.8833333 0.2755022 bbsrna140 3373878 3601836 - 52 53 1.0192308 0.3178872 bbsrna141 3336834 3601836 + 36 53 1.4722222 0.4591703 bbsrna142 1139209 3601836 - 63 53 0.8412698 0.262383 bbsrna143 929038 3601836 + 48 52 1.0833333 0.3378801 bbsrna144 3616764 3601836 - 251 52 0.2071713 0.0646145 bbsrna145 34160 3601836 + 61 52 0.852459 0.2658728 bbsrna146 1782322 3601836 + 37 52 1.4054054 0.4383309 bbsrna147 42078 3601836 + 123 51 0.4146341 0.1293199 bbsrna148 3540524 3601836 + 97 51 0.5257732 0.163983 bbsrna149 4888933 3601836 + 31 50 1.6129032 0.5030472 bbsrna150 4480719 3601836 - 241 50 0.2074689 0.0647073 bbsrna151 4268230 3601836 + 107 50 0.4672897 0.1457427 bbsrna152 4191356 3601836 + 116 50 0.4310345 0.134435 bbsrna153 411311 3601836 - 58 50 0.862069 0.2688701 bbsrna154 3989538 3601836 + 72 50 0.6944444 0.2165898 bbsrna155 3679712 3601836 - 37 50 1.3513514 0.421472 bbsrna156 3184599 3601836 - 101 50 0.4950495 0.1544006

160 bbsrna157 2055598 3601836 - 61 50 0.8196721 0.255647 bbsrna158 203877 3601836 + 69 50 0.7246377 0.2260067

observed that only 22 small RNAs were common between SIPHT prediction and SOLiD sequencing (Table 4). This could be due to false positives in SIPHT prediction or could suggest

Table 4 Small RNAs conserved amongst SIPHT prediction and SOLiD sequencing. Small RNAs predicted by SIPHT (red) are listed next to sRNAs identified by RNA sequencing (black). Start and end positions within the genome, strand, and length are also listed for both the predicted and sequences sRNAs.

sRNA sRNA sRNAstart sRNAend strand length >BbSIPHT448 6309 6425 + 116 bbsrna100 6317 6426 + 110 >BbSIPHT304 321225 321339 + 114 bbsrna90 321225 321295 + 71 >BbSIPHT168 1504253 1504432 + 179 bbsrna64 1504333 1504397 + 65 >BbSIPHT184 1782347 1782385 + 38 bbsrna146 1782322 1782358 + 37 >BbSIPHT231 2206705 2206916 - 211 bbsrna6 2206723 2206863 - 141 >BbSIPHT242 2298145 2298293 + 148 bbsrna 72 2298090 2298231 + 142 >BbSIPHT90 2558978 2559118 - 140 bbsrna38 2558978 2559049 - 72 >BbSIPHT303 3184577 3184701 - 124 bbsrna156 3184599 3184699 - 101 >BbSIPHT325 3478165 3478234 - 69 bbsrna119 3478166 3478278 - 113 >BbSIPHT40 4114080 4114479 - 399 bbsrna130 4114378 4114433 - 56 >BbSIPHT381 4122955 4123062 - 107 bbsrna107 4122916 4123096 - 181 >BbSIPHT389 4223118 4223204 - 86 bbsrna60 4223117 4223180 - 64 >BbSIPHT397 4397117 4397197 - 80 bbsrna18 4397155 4397211 - 57 >BbSIPHT36 4401767 4401808 - 41 bbsrna11 4401795 4401862 - 68 >BbSIPHT398 4429325 4429395 - 70 bbsrna89 4429341 4429390 - 50 >BbSIPHT25 4494724 4494854 + 130 bbsrna138 4494700 4494933 + 234

161 >BbSIPHT130 4666193 4666348 + 155 bbsrna92 4666337 4666459 + 123 >BbSIPHT428 5147268 5147438 - 170 bbsrna4 5147238 5147452 - 215 >BbSIPHT390 4234461 4234537 + 76 bbsrna20 4234331 4234488 - 158 >BbSIPHT118 4268210 4268353 - 143 bbsrna151 4268230 4268336 + 107 that many bordetellae small RNAs are only transcribed under very specific conditions and thus

were not transcribed when bordetellae was grown in vitro to exponential phase.

These 181 sRNA sequences were then mapped onto a linear RB50 genome, allowing us

to determine if small RNAs are clustered within the genome (Figure 1). Although the highest

Figure 6-1 Small RNAs are distributed evenly throughout the B. bronchiseptica genome. Each non-coding (blue) or coding (red) sRNA molecule was mapped to a linear B. bronchiseptica strain RB50 genome. Bar height for each sRNA represents read frequency, and sRNAs identified multiple times within the genome were shown at all possible locations. Y. Zhang constructed this figure.

density of sRNAs is located in the first tenth of the genome, B. bronchiseptica small RNAs are

dispersed throughout the genome and vary wildly in their levels of transcription. Additionally,

three sRNA molecules were mapped to several different locations within the genome. To better

determine where these sRNA molecules are transcribed in relation to open reading frames

(ORF), we analyzed the distance of each sRNAs from the nearest known ORF. The majority of

the putative sRNA molecules were present 50 bp away from the nearest ORF (Figure 2). Small

RNAs near the 5’ end of an ORF appeared to be the most common, suggesting that the 5’

162

Figure 6-2 Small RNAs are commonly transcribed near the 5’ end of an ORF. The orientation of each sRNA molecule was determined via standalone BLAST, and grouped according to its orientation to the nearest open reading frame. Outside of the ORF, sRNAs were grouped into 50 bp bins and the total number in each bin was graphed above the x-axis if sRNA transcription occurred from the 5’ DNA strand, or below the x-axis if the sRNA was transcribed from the negative strand, relative to the ORF. Small RNAs mapping to within the ORF were also grouped into 100 bp bins and graphed according to the their location within the ORF. Y. Zhang constructed this figure.

untranslated region (UTR) of messenger transcripts is the most abundant location to encode

sRNAs. These 5’ UTR sRNAs could be non-functional and are simply a by-product of mRNA

transcription, or they could be novel mechanisms in which B. bronchiseptica regulates other

genes after a particular gene is transcribed and processed.

Small RNA transcription regulation.

Because several of these sRNAs could be an overlap of mRNA transcription, we wanted

to determine if these sRNAs had specific regulators or terminators. First, Rho-independent

terminators for the majority of the 158 non-coding putative sRNA molecules were identified

using TransTermHP (Appendix B) [43]. Although Rho-independent terminators were identified

for each sRNA molecule, rho-dependent terminators may overlap. Next, we used another

bioinformatic approach to identify some of the possible transcription regulators. In E. coli, RpoE

or σE regulate and is regulated by sRNAs; σE is also turned on during cell envelope stress [44].

In B. bronchiseptica, σE was recently shown to regulate hfq, which encodes the sRNA chaperone,

163 and contribute to B. bronchiseptica lethality (S.E. Barchinger, Submitted to BMC Microbiology).

Using an in-house prediction program, 5 putative sRNAs had a consensus sigE binding sequence

directly upstream, suggesting that these small RNAs may be σE regulated (Table 5). These

Table 5 Small RNA molecules with predicted σE binding sites upstream. Small RNAs with predicted σE binding sites upstream, designated by underlining. The predicted messenger RNA targets for each sRNA are also listed. S. E. Barchinger constructed this figure. sRNA name Sequence

bbsRNA047 tgaattttcggctattgttcccatATCAGACGTAACCGACGTGGAGACCTGGGTC

TGCCCTTGCCCCGTCCCCCATTACGC

Pr dicted Target Product target

BB1808 probable GntR-family transcriptional regulator

BB0780 putative type II secretion system protein

BB2756 permease component of ABC transporter protein

bbsRNA125 agaaattcgcgcgaacctataatcaaaAGCGAGGGGCTTTTCGATTTGCCTGTCC

ACCCACGCACCAAGAATGCCGCCAGGGCTCGCCTGGTCGAAGGCG

GCCACGCGTGTCTTGTGTTTGTTCAATCCTAGGGGATCATCACTAG

GA

Predicted target Target Product

BB1355 hypothetical protein

BB4346 putative AraC-family transcriptional regulator

BB0820 putative lipoprotein

BB3363 hypothetical protein

BB0157 3-deoxy-D-manno-octulosonic-acid transferase

BB1910 exodeoxyribonuclease VII small subunit

BB0595 hypothetical protein

BB4543 sensor kinase protein

164 bbsRNA058 ggaacttgtgcgtggcagcatagtccaattacCAGAAGCGGGATTAGCACCACCCG

CCGACGTCGCGCGTTGCAGCCGCACGCGTTTTGCAAGTAACGGAC

TGGGAG

Predicted target Target Product

BB2860 hypothetical protein

BB0889 putative two-component system sensor protein

BB0882 putative glycosyltransferase

BB3087 hypothetical protein

BB1081 LysR-family transcriptional regulator

BB2172 isoleucyl-tRNA synthetase

BB2803 putative heme export protein

BbSIPHT144 ggaaattgcttcgctgtcggccaactaacaggcCGCTGGCACATAGAACTTTGGGCC

GCATCCGGACTTGTCCGGATGCTGCTGTCAAAGACCGCTGGAGCG

CTAGCCGGTTGTTTGGCGAATGCTTAATCCCGAAGCTGGTTCAGTT

CGGATCCAGCGTAGACGGCGTCCCCAGGAAGATTTCTTTACCCGA

AGAACTCAACCAGGCGCCGCATTCGGTTGACGCGCAAGGCTAGCG

CCCCAAGCGTCTTTTGA

Predicted target Target Product

BB4638 putative polysaccharide deacetylase

Predicted target Target Product

BB2666 putative transmembrane regulator

BB3710 putative transport permease

BB4407 phosphomethylpyrimidine kinase

BB3277 ABC transporter, permease protein

BB2235 phage-related hypothetical protein

165

BB2831 iron utilization protein

BB3485 phage-related hypothetical protein

BB4270 putative acetyltransferase

BB4135 hypothetical protein

BbSIPHT399 agaactcaaacatcgccacctcgctcagaacagGCTGAAGTCCGCGCGCCGACGCA

CGGCATCCGCCCGCGCGGCGCGACTCGGAAAACCAGCTTAGCAGG

CTGCGGGATCGCGCAGGCAGGCTGGCGCGGCGGTTTCCCGATTGT

CGGGAAACCGCGTCGCGGCAAGCGGATATCGAAGGCGGGAAGAT

GGCCGCGCGCAACAACCAGGCGCACCGCATGAAAAAGGCCAGCA

AGAAAAAAGCCGCTTCCATGAGCGGCTTTTTTCT

Target Product Predicted target

BB1081 LysR-family transcriptional regulator

BB4038 putative transcriptional regulator

BB1440 recombination protein RecR

phosphopantothenoylcysteine BB2170 synthase/decarboxylase

BB0883 putative glycosyltransferase

BB1530 N-carbamoyl-L-amino acid amidohydrolase

BB0648 probable regulator

BB1377 hypothetical protein

BB3081 putative membrane protein

BB3370 putative transmembrane transport protein

BB2766 putative short chain dehydrogenase

BB4949 biotin carboxylase

BB1855 putative integral membrane protein

166

BB1142 putative amidase

BB1556 ABC transporter, ATP-binding protein

BB2260 hypothetical protein

BB2810 cytochrome C-type biogenesis protein

BB2702 carboxymethylenebutenolidase

BB3687 aconitate hydratase

BB4810 probable transcriptional regulator

BB2421 hypothetical protein

BB3063 hypothetical protein

BB3300 RNA polymerase sigma-70 factor

sRNA molecules may additionally play a role in environmental, outer membrane stress responses, or contribute to the role sigE plays in bordetellae virulence. Together, these data suggest that our putative sRNA molecules are indeed under the control of a specific transcriptional response.

Figure 6-3 Northern blotting of 4 identified sRNAs. Northern blotting was performed on bbsrna7, 11, 23,and 133 by probing denatured RNA obtained from log-phase cultures with [γ-32P]ATP radiolabeled oligonucleotide probes designed against the first 20 bp of each sRNA molecule. The expected size from SOLiD sequencing is additionally listed. S. E. Barchinger and D. E. Place completed this experiment and analyzed the data.

167 To additionally determine that these putative sRNAs are not background “noise” and

indeed have directed regulation, we completed Northern blotting on four random sRNAs that

were transcribed at different levels. Bacteria were grown under the same conditions, but a

different RNA extraction method was used to ensure the extraction process had not biased our

results. Bbsrna7, 11, 23, and 133 were all verified by Northern blotting (Figure 3). As

predicted, small RNA molecules were similar to their predicted sizes and appeared to be

expressed at levels comparable to their copy numbers, further confirming the quality of our

SOLiD sequencing analysis to identify sRNA molecules.

Small RNA target prediction yields various virulence factors.

Currently, the general purpose of sRNA molecules is to provide post-transcriptional regulation of various genes, as a means of “fine-tuning” a transcriptional response. Small RNAs can act through several mechanisms; however, the most common is by binding the target in a reverse complementary manner. TargetRNA is designed to model the structure of the small

RNA, determine which portion of the sRNA could bind to any given tertiary structure, and then scan the host genome to determine if any complementary RNA sequences exist in mRNA regulatory regions [41]. Using TargetRNA, we predicted putative target messenger RNA sequences for each of the putative sRNAs (Table 6). Predicted targets included regulatory factors, like LysR transcription regulators and sigma 70, and metabolic enzymes, to phage

Table 6 Virulence factor targets of sequenced sRNAs. Targets predicted for each sequenced sRNA, if they were predicted to bind known virulence factors, are listed. The genome start and end position, strand, length and coverage are additionally listed. Predicted target information includes the index number for each sRNA, gene designation if know, NCBI gene number, and TargetRNA score (- 100 highest probability score).

Name Start End Strand Length Coverage Predicted Target bbsnra13 3601640 3601836 + 197 309 1 BB0782 BB0782 -71 bbsrna21 4636206 4636268 - 63 195 7 bscR BB1632 -59

168 bbsrna23 1054449 1054588 + 140 182 6 fimD BB2989 -65 9 bscQ BB1631 -63 bbsrna28 2736123 2736589 + 467 168 26 BB0804 BB0804 -75 bbsrna41 3379965 3380072 - 108 137 1 ptlB BB4896 -74 bbsrna45 4235832 4235924 - 93 130 45 BB0782 BB0782 -60 bbsrna47 2384958 2385014 + 57 127 2 BB0780 BB0780 -64 bbsrna70 2327272 2327310 + 39 103 3 bcrD BB1612 -68 4 bscL BB1627 -63 7 BB0804 BB0804 -61 bbsrna78 4365504 4365544 - 41 83 2 bscL BB1627 -61 bbsrna87 5151360 5151507 + 148 74 2 fimX BB3426 -68 bbsrna103 4479641 4479733 - 93 68 21 fhaL BB1936 -60 bbsrna114 2736205 2736332 - 128 61 1 cyaB BB0325 -77 27 bscQ BB1631 -63 bbsrna133 49743 49816 + 74 54 20 bvgS BB2995 -61 bbsrna144 3616764 3617014 - 251 52 8 BB0782 BB0782 -75 9 BB4900 BB4900 -74 bbsrna148 3540524 3540620 + 97 51 2 bscQ BB1631 -87 bbsrna150 4480719 4480959 - 241 50 3 cyaD BB0326 -80 6 BB0782 BB0782 -78 bbsrna153 411311 411368 - 58 50 2 wbmO BB0130 -77 bbsrna158 203877 203945 + 69 50 1 cyaB BB0325 -81

related proteins, and to hypothetical proteins. Several various groups of targeted proteins were

identified, such as flagellar genes (fliT, fliO, flgG, flhA, fliA, fliR, fliI, and fliQ) and the Type III

Secretion System (bscR, bscQ, bscL, and bcrD), suggesting that sRNAs may play a post- transcriptional role in regulating large loci that require assembly of numerous proteins. This may also suggest that these sRNAs are used to down regulate genes within a locus at times when the remaining genes are required. The Type III Secretion System was not the only putative sRNA target that was a known bordetellae virulence factor. Other known virulence factor targets included a putative Type II Secretion System (BB0782), fimbrial genes (fimD and fimX), a pertussis toxin transport gene (ptlB), filamentous hemagglutinin (fhaL), adenylate cyclase toxin

(cyaB and cyaD), the master regulatory system BvgAS (bvgS), the locus encoding O-antigen

169 (wbmO), and the Type VI Secretion System (BB0804). This predictive analysis suggests that sRNA molecules probably play a large role in virulence factor regulation and may contribute to observed phenotypic differences between classical bordetellae strains.

Small RNAs bbsrna28 and bbsrna70 are up-regulated in response to serum and may

regulate a putative Type VI Secretion System Effector.

During our predictive analysis, we noticed that bbsrna70 was predicted to

regulate not only bcrD and bcrL, two highly conserved T3SS proteins, but also another gene

encoded within a putative T6SS locus, BB0804. Our lab earlier defined BB0804 as tssJ, a gene

Figure 6-4 Bbsrna 23 and 70 bind a similar region upstream of tssJ. A.) A schematic was constructed demonstrating where bbsrna23 and bbsrna70 are predicted to bind upstream of tssJ. The putative binding sequences are also illustrated and compared to the known DNA sequence in this region.

residing within the 26 gene T6SS locus and containing sequence encoding known lysozyme

domains [45]. In an independent study, we also observed that this gene was highly up-regulated

when bacteria were exposed to serum, while the remainder of the T6SS locus was not highly

expressed (S. E. Hester and L. S. Weyrich, unpublished data). Furthermore, the increased

expression of this protein in response to serum exposure also correlated with increased

cytotoxicity in vivo, suggesting it indeed has lysozyme functions (S. E. Hester and L. S.

Weyrich, unpublished data). Because of these findings, we hypothesized that tssJ encodes a

170 putative T6SS effector protein and has the independent regulation compared to the remaining

T6SS locus genes. First, we mapped where these putative sRNA molecules would bind to

regulate tssJ (Figure 4A). The predicted sRNA binding site was between -30 to -2 bp upstream

of the open reading frame, suggesting that it may interfere or modulate with tssJ translation and

ribosomal binding. Although highly speculative, bbsrna28 and bbsrna70 may negatively

regulate tssJ during in vitro conditions but allow translation of this transcript when bacteria are

responding to an in vivo signal, such as serum.

Small RNA molecule conservation amongst the classical bordetellae.

Because the classical bordetellae genomes share high similarity, we hypothesized that

many of these sRNA molecules would be

conserved amongst the classical

bordetellae. Hot et al. recently identified 138 13 sRNAs in B. pertussis, using both 2 18 bioinformatic prediction and verification 6 via Northern blotting [34]. Surprisingly,

Figure 6-5 B. bronchiseptica sRNAs are conserved amongst when sRNA sequence similarity was the classical bordetellae. A Venn diagram was constructed to illustrate sRNAs compared between the B. pertussis sRNA conserved amongst B. bronchiseptica strain RB50, B. pertussis strain Tohama I, and B. parapertussis strain 12822. Because SOLiD sequencing of B. pertussis and B. molecules identified by Hot et al. and the parapertussis has not been completed, these numbers and comparisons are left blank. B. bronchiseptica ones identified here using

SOLiD sequencing, we found no overlap between the sRNAs identified in these studies. This

may be due to different approaches or because of genomic differences between B. pertussis and

B. bronchiseptica. To determine if any of the sRNAs identified in our study were conserved, we

171 used BLAST to compare our putative non-coding B. bronchiseptica sRNA sequences with the genomes of B. pertussis strain Tohama I and B. parapertussis strain 12822 (Figure 5). Of the

158 non-coding B. bronchiseptica sRNAs, 138 of these were also conserved in B. pertussis and

B. parapertussis. Two B. bronchiseptica sRNA molecules were not identified in B. parapertussis, while 19 were not identified in B. pertussis. Together, these data indicate that sRNA were conserved during the divergence of these species and may play critical roles in survival or pathogenesis; however, these results additionally suggest that there is still much to learn about how these sRNA molecules contribute to gene regulation within the classical bordetellae.

The BLAST comparison of sRNAs amongst the classical bordetellae identified 6 sRNAs that were unique to only B. bronchiseptica (Table 5). These sRNAs, bbsrna87, 116, 143, 145,

150, and 154, are transcribed throughout the genome and are predicted to regulate a wide range of targets, from hypothetical proteins to cyaD, a component of the adenylate cyclase toxin encoding locus. Interestingly, bbsrna87 is predicted to regulate fimX, a fimbriae encoding gene that has been observed to have little or no transcription, as detected by classical methodologies.

Furthermore, remaining B. bronchiseptica sRNA specific targets include mostly metabolic genes, such as dehydrogenases, hydratases, acetyltransferase, ABC transporter permease, and pseudouridine synthase, suggesting that these sRNA molecules may play a role in B. bronchiseptica lineage specificity, perhaps by regulating metabolism.

172

Discussion

This is the first description of sRNA molecules in the zoonotic pathogen, B. bronchiseptica. These data suggest that sRNAs are more numerous and diverse in Bordetella than ever imagined, and that these sRNAs may play a major role in fine-tuning virulence factor regulation in the bordetellae, especially in large, multi-protein machineries. We now know that bordetellae sRNA molecules are not only BvgAS regulated, but also regulated by sigE, suggesting they play an essential role in stress responses. Additionally, two sRNAs were identified to putatively regulate a T6SS effector protein, which was recently linked to increased eukaryotic cell death [45]. Lastly, we determined by sequence similarity that many of these sRNAs are conserved amongst the classical bordetellae, suggesting that they play a vital role in the ability of Bordetella to cause disease. Interestingly, the few unique B. bronchiseptica sRNA molecules may explain some of the species variation and unique traits specific to this zoonotic pathogen.

In this study, we chose to analyze the sRNA molecules transcribed during exponential growth in vitro, which allowed 181 novel sRNA molecules to be identified. Although there are many advantages to this approach, many small RNAs are also often only expressed in stationary phase and therefore would not be observed during our experimental conditions [46]. However, we are confident that SOLiD sequencing may be able to identify numerous stationary phase sRNAs in future sequencing attempts. Interestingly, 14 of the 21 sRNAs identified by Hot et al. were transcribed after 30 hours of growth, of which none were identified in this study [34].

However, when bordetellae are grown to stationary phase, a large proportion of individual cells can become Bvg- mutants, which can drastically change the transcriptome and may also have

173 significantly impacted the differential sRNA identification [47,48]. Additionally, SIPHT

identified over 400 predicted sRNA molecules, suggesting that our sequencing approach under

these conditions has probably only identified a small subset of bordetellae sRNAs. Nevertheless,

this study provides a much broader identification of bordetellae sRNA molecules and begins to

lend insight into their physiological functions, although continued sRNA identification under

multiple, additional conditions will be required to understand how sRNAs fully contribute to

bordetellae evolution, virulence, and survival.

By using a bioinformatic approach, TargetRNA, we were able to predict putative mRNA

targets for every identified putative sRNA. Many of the sRNAs had multiple targets, suggesting

that sRNAs may play a role in a very complex web of regulation; sRNAs may target multiple

mRNAs as part of a much larger response. For example, 3 sRNAs identified here had σE binding sites upstream and were predicted to regulate a GntR and an AraC transcriptional regulator and outer membrane associated proteins, suggesting that these sRNAs play a role in an outer membrane stress response cascade. We additionally identified several sRNAs that targets multiple flagellar genes and genes within the T3SS locus, suggesting that sRNAs may help fine tune regulation of large loci. By microarray analysis, the flagellar locus has been shown to be down regulated during exponential phase in vitro, because this locus is BvgAS regulated

[13,13,12,49]. These flagellar targeted sRNAs may increase the down-regulation by targeting flagellar mRNA for degredation. We also identified 6 different sRNAs predicted to target bscR, bscQ, bcrD, and bscL, although none of the sRNAs targeted btrS, the BvgAS regulated sigma factor found to control transcription of the T3SS locus, suggesting that these sRNA molecules are able to post-transcriptionally modulate the T3SS locus in vitro [50,51]. Furthermore, two

genes encoding CyaB and CyaD were also identified as putative sRNA targets, suggesting that

174 sRNAs may additionally regulate secretion of ACT, an essential toxin conserved amongst the

classical bordetellae [52–55,2]. These large complexes that require multiple proteins to be

produced and their delicate assembly orchestrated for proper function would definitely benefit

from fine tuning their regulation, allowing the bacteria to very precisely turn on and off transcripts as needed.

Understanding how the sRNAs regulate virulence genes and entire systems, including the flagellum system and the T3SS, could lead to new strategies for treating and managing disease; for example, ectopic expression of flagellar genes can limit the ability of B. bronchiseptica to colonize, suggesting that abrogating these sRNAs may also affect the ability of B. bronchiseptica to cause disease [56]. Interestingly, bbsrna133 is predicted to target the transcript encoding

BvgS, the sensory protein of the master regulatory system in bordetellae [57]. To our knowledge, this is the first report of alternative regulation of bvgS, and we hypothesize that this sRNA may limit BvgS production, perhaps as an immune evasion mechanism to limit inflammation or pathology. Understanding the mechanism behind how this sRNA regulates BvgS production could reveal additional antimicrobial strategies against bordetellae infection

Several sRNAs identified here are not only regulating known virulence factors, but may also explaining puzzling observations within the bordetellae field. 6 sRNA molecules were uniquely identified in B. bronchiseptica and appear to have been lost in the evolution, and perhaps species adaption, of the two human pathogens, B. pertussis and B. parapertussis. One of these sRNAs targets fimX, a fimbriae encoding gene present in only B. pertussis and B. bronchiseptica. Transcription of this gene has been extremely hard to detect, causing researchers to suggest that the fimX transcript may not even be expressed at all [58]. Transcripts of fimX may not have been detectable because they are being targeted by a sRNA for degradation.

175 Another interesting sRNA, bbsrna153, is predicted to target wbmO, a gene within the O-antigen locus, although it has remained of unknown function [59,60]. The function of WbmO may have been difficult to study because its encoding transcript may have been constantly silenced, not producing a functional protein even when the transcripts are detectable. The potential function of WbmO may only be determined if bbsrna153 can be disrupted.

Together, these findings provide significant evidence that sRNAs are important for virulence factor regulation within the classical bordetellae. Disruption of each sRNA and determining its contributions to virulence will need to occur before we will fully be able to understand the contributions of these sRNAs during infection. This approach to identifying bordetellae sRNAs has elucidated potential mechanisms behind virulence factor regulation and species variation, and given insight into confounding observations in the field.

176

Author Contributions:

Authors: Laura S. Weyrich1,3, Sarah E. Barchinger2,3, Jihye Park1,4, Ying Zhang1, David E.

Place1,5, Sarah E. Ades2, and Eric T. Harvill1

1 Department of Veterinary and Biomedical Sciences, The Pennsylvania State University,

115 Henning Bldg, University Park, PA 16802

2 Biochemistry and Molecular Biology Department, The Pennsylvania State University,

101 Althouse Laboratory, University Park, PA 16802

3 Biochemistry, Microbiology, and Molecular Biology Graduate Program

4 Bioinformatics and Genomics Graduate Program

5 Immunology and Infectious Disease Graduate Program

Conceived and designed experiments: LSW, SEB, ETH

Preformed experiments: LSW, SEB, JP, YZ, DEP

Analyzed data: LSW, SEB, JP, SEA, ETH

Wrote manuscript: LSW, ETH

177

References

1. Diavatopoulos DA, Cummings CA, Schouls LM, Brinig MM, Relman DA, et al. (2005) Bordetella pertussis, the causative agent of whooping cough, evolved from a distinct, human-associated lineage of B. bronchiseptica. PLoS Pathog 1: e45.

2. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, et al. (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35: 32-40. doi:10.1038/ng1227 ng1227 [pii]

3. Goodnow RA (1980) Biology of Bordetella bronchiseptica. Microbiol Rev 44: 722-38.

4. Mattoo S, Cherry JD (2005) Molecular pathogenesis, epidemiology, and clinical manifestations of respiratory infections due to Bordetella pertussis and other Bordetella subspecies. Clin Microbiol Rev 18: 326-82.

5. Mattoo S, Foreman-Wykert AK, Cotter PA, Miller JF (2001) Mechanisms of Bordetella pathogenesis. Front Biosci 6: E168-86.

6. Giles CJ, Leman AD, L. SW, L. MW, D. DS, et al. (1992) Bordetellosis. In: Diseases of Swine. Ames: Iowa State University Press. pp. 436-445.

7. Mattoo S, Cherry JD (2005) Molecular Pathogenesis, Epidemiology, and Clinical Manifestations of Respiratory Infections Due to Bordetella pertussis and Other Bordetella Subspecies. Clin. Microbiol. Rev. 18: 326-382. doi:10.1128/CMR.18.2.326-382.2005

8. Cotter PA, Miller JF (1997) A mutation in the Bordetella bronchiseptica bvgS gene results in reduced virulence and increased resistance to starvation, and identifies a new class of Bvg-regulated antigens. Mol Microbiol 24: 671-85.

9. Merkel TJ, Stibitz S, Keith JM, Leef M, Shahin R (1998) Contribution of regulation by the bvg locus to respiratory infection of mice by Bordetella pertussis. Infect Immun 66: 4367-73.

10. BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. (1994). Available: http://medjournal.hmc.psu.edu:2060/pmc/articles/PMC302969/?tool=pubmed. Accessed 24 Jul 2011.

11. Cotter PA, Jones AM (2003) Phosphorelay control of virulence gene expression in Bordetella. Trends Microbiol 11: 367-73.

12. Merkel TJ, Barros C, Stibitz S (1998) Characterization of the bvgR locus of Bordetella pertussis. J Bacteriol 180: 1682-90.

13. Merkel TJ, Boucher PE, Stibitz S, Grippe VK (2003) Analysis of bvgR expression in Bordetella pertussis. J Bacteriol 185: 6902-12.

178 14. Williams CL, Boucher PE, Stibitz S, Cotter PA (2005) BvgA functions as both an activator and a repressor to control Bvg phase expression of bipA in Bordetella pertussis. Mol Microbiol 56: 175- 88.

15. Gottesman S (2005) Micros for microbes: non-coding regulatory RNAs in bacteria. Trends Genet. 21: 399-404. doi:10.1016/j.tig.2005.05.008

16. Majdalani N, Vanderpool CK, Gottesman S (2005) Bacterial small RNA regulators. Crit. Rev. Biochem. Mol. Biol. 40: 93-113. doi:10.1080/10409230590918702

17. Gottesman S (2004) THE SMALL RNA REGULATORS OF ESCHERICHIA COLI : Roles and Mechanisms*. Annual Review of Microbiology 58: 303-328. doi:10.1146/annurev.micro.58.030603.123841

18. Vogel J (2009) A rough guide to the non-coding RNA world of Salmonella. Molecular Microbiology 71: 1-11. doi:10.1111/j.1365-2958.2008.06505.x

19. Massé E, Escorcia FE, Gottesman S (2003) Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes & Development 17: 2374 -2383. doi:10.1101/gad.1127103

20. Deborah Chen H, Frankel G (2005) Enteropathogenic Escherichia coli: unravelling pathogenesis. FEMS Microbiology Reviews 29: 83-98. doi:10.1016/j.femsre.2004.07.002

21. Alejandro Toledo-Arana, Francis Repoila, Pascale Cossart (2007) Small noncoding RNAs controlling pathogenesis. Curr Opin Micro 10: 182-188.

22. Arnvig KB, Young DB (2009) Identification of small RNAs in Mycobacterium tuberculosis. Molecular Microbiology 73: 397-408. doi:10.1111/j.1365-2958.2009.06777.x

23. Chabelskaya S, Gaillot O, Felden B (2010) A Staphylococcus aureus Small RNA Is Required for Bacterial Virulence and Regulates the Expression of an Immune-Evasion Molecule. PLoS Pathog 6: e1000927. doi:10.1371/journal.ppat.1000927

24. Chen Y, Indurthi DC, Jones SW, Papoutsakis ET (January) Small RNAs in the Genus Clostridium. mBio 2. Available: http://mbio.asm.org/content/2/1/e00340-10.abstract. Accessed 7 Mar 2011.

25. Livny J (2006) Identification of 17 Pseudomonas aeruginosa sRNAs and prediction of sRNA- encoding genes in 10 diverse pathogens using the bioinformatic tool sRNAPredict2. Nucleic Acids Research 34: 3484-3493. doi:10.1093/nar/gkl453

26. Padalon-Brauch G, Hershberg R, Elgrably-Weiss M (n.d.) Small RNAs encoded within genetic islands of Salmonella typhimurium show host-induced expression and role in virulence.. Available: http://nar.oxfordjournals.org/content/36/6/1913.short. Accessed 12 Sep 2011.

27. Perez N, Treviño J, Liu Z, Ho SCM, Babitzke P, et al. (2009) A Genome-Wide Analysis of Small Regulatory RNAs in the Human Pathogen Group A Streptococcus. PLoS ONE 4: e7668. doi:10.1371/journal.pone.0007668

179 28. Song T, Mika F, Lindmark B, Liu Z, Schild S, et al. (2008) A new Vibrio cholerae sRNA modulates colonization and affects release of outer membrane vesicles. Mol Microbiol 70: 100-111. doi:10.1111/j.1365-2958.2008.06392.x

29. Sonnleitner E, Gonzalez N, Sorger-Domenigg T, Heeb S, Richter AS, et al. (2011) The small RNA PhrS stimulates synthesis of the Pseudomonas aeruginosa quinolone signal. Molecular Microbiology 80: 868-885. doi:10.1111/j.1365-2958.2011.07620.x

30. Christiansen JK, Larsen MH, Ingmer H, Sogaard-Andersen L, Kallipolitis BH (2004) The RNA- Binding Protein Hfq of Listeria monocytogenes: Role in Stress Tolerance and Virulence. Journal of Bacteriology 186: 3355-3362. doi:10.1128/JB.186.11.3355-3362.2004

31. Schiano CA, Bellows LE, Lathem WW (2010) The Small RNA Chaperone Hfq Is Required for the Virulence of Yersinia pseudotuberculosis. Infection and Immunity 78: 2034-2044. doi:10.1128/IAI.01046-09

32. Shakhnovich EA, Davis BM, Waldor MK (2009) Hfq negatively regulates type III secretion in EHEC and several other pathogens. Molecular Microbiology 74: 347-363. doi:10.1111/j.1365- 2958.2009.06856.x

33. Sittka A, Pfeiffer V, Tedin K, Vogel J (2007) The RNA chaperone Hfq is essential for the virulence of Salmonella typhimurium. Mol Microbiol 63: 193-217. doi:10.1111/j.1365-2958.2006.05489.x

34. Hot D, Slupek S, Wulbrecht B, D’Hondt A, Hubans C, et al. (2011) Detection of small RNAs in Bordetella pertussis and identification of a novel repeated genetic element. BMC Genomics 12: 207. doi:10.1186/1471-2164-12-207

35. Cotter PA, Miller JF (1994) BvgAS-mediated signal transduction: analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. Infect Immun 62: 3381-90.

36. Bordet J, Gengou O (1906) Le microbe de la coqueluche. Ann. Inst. Pasteur 20: 48-68.

37. Stainer DW, Scholte MJ (1970) A simple chemically defined medium for the production of phase I Bordetellla pertussis. J. Gen. Microbiol. 63: 211-220.

38. Rumble SM, Lacroute P, Dalca AV, Fiume M, Sidow A, et al. (2009) SHRiMP: Accurate Mapping of Short Color-space Reads. PLoS Comput Biol 5: e1000386. doi:10.1371/journal.pcbi.1000386

39. Livny J (2006) Identification of 17 Pseudomonas aeruginosa sRNAs and prediction of sRNA- encoding genes in 10 diverse pathogens using the bioinformatic tool sRNAPredict2. Nucleic Acids Research 34: 3484-3493. doi:10.1093/nar/gkl453

40. Altschul SF, Gish W, Miller W, Myers EW, DJ L (1990) Basic local alignment search tool. J Mol Biol 215: 403-410.

41. Tjaden B (2008) TargetRNA: a tool for predicting targets of small RNA action in bacteria. Nucleic Acids Research 36: W109-W113. doi:10.1093/nar/gkn264

42. Buboltz AMNTL (2008) Enhanced TTSS function contributes to the hypervirulence of a Bordetella bronchiseptica lineage. Journal of Bacteriology Submitted.

180 43. Kingsford CL, Ayanbule K, Salzberg SL (2007) Rapid, accurate, computational discovery of Rho- independent transcription terminators illuminates their relationship to DNA uptake. Genome Biol. 8: R22. doi:10.1186/gb-2007-8-2-r22

44. Thompson KM, Rhodius VA, Gottesman S (2007) σE Regulates and Is Regulated by a Small RNA in Escherichia coli. J Bacteriol 189: 4243-4256. doi:10.1128/JB.00020-07

45. Weyrich, L.S., Rolin, O. Y., Muse, S. J., Park, J., Spidale, N., et al. (2012) A Type VI Secretion System Encoding Locus is Required for Bordetella bronchiseptica Immunomodulation and Persistence In Vivo. PLoS One Submitted.

46. Opdyke JA, Kang J-G, Storz G (2004) GadY, a small-RNA regulator of acid response genes in Escherichia coli. J. Bacteriol. 186: 6698-6705. doi:10.1128/JB.186.20.6698-6705.2004

47. Stibitz S (1994) Mutations in the bvgA gene of Bordetella pertussis that differentially affect regulation of virulence determinants. J Bacteriol 176: 5615-21.

48. Stibitz S, Yang MS (1999) Genomic plasticity in natural populations of Bordetella pertussis. J Bacteriol 181: 5512-5.

49. Akerley BJ, Miller JF (1993) Flagellin gene transcription in Bordetella bronchiseptica is regulated by the BvgAS virulence control system. J Bacteriol 175: 3468-79.

50. Yuk MH, Harvill ET, Miller JF (1998) The BvgAS virulence control system regulates type III secretion in Bordetella bronchiseptica. Mol Microbiol 28: 945-59.

51. Mattoo S, Yuk MH, Huang LL, Miller JF (2004) Regulation of type III secretion in Bordetella. Mol Microbiol 52: 1201-14.

52. Ehrmann IE, Gray MC, Gordon VM, Gray LS, Hewlett EL (1991) Hemolytic activity of adenylate cyclase toxin from Bordetella pertussis. FEBS Lett 278: 79-83.

53. Glaser P, Danchin A, Ladant D, Barzu O, Ullmann A (1988) Bordetella pertussis adenylate cyclase: the gene and the protein. Tokai J Exp Clin Med 13 Suppl: 239-52.

54. Gueirard P, Druilhe A, Pretolani M, Guiso N (1998) Role of adenylate cyclase-hemolysin in alveolar macrophage apoptosis during Bordetella pertussis infection in vivo. Infect Immun 66: 1718-25.

55. Gueirard P, Guiso N (1993) Virulence of Bordetella bronchiseptica: role of adenylate cyclase- hemolysin. Infect Immun 61: 4072-8.

56. Akerley BJ, Cotter PA, Miller JF (1995) Ectopic expression of the flagellar regulon alters development of the Bordetella-host interaction. Cell 80: 611-20. doi:0092-8674(95)90515-4 [pii]

57. Cotter PA, Miller JF (1997) A mutation in the Bordetella bronchiseptica bvgS gene results in reduced virulence and increased resistance to starvation, and identifies a new class of Bvg-regulated antigens. Mol Microbiol 24: 671-85.

181 58. Riboli B, Pedroni P, Cuzzoni A, Grandi G, de Ferra F (1991) Expression of Bordetella pertussis fimbrial (fim) genes in Bordetella bronchiseptica: fimX is expressed at a low level and vir- regulated. Microb. Pathog. 10: 393-403.

59. Vinogradov E, King JD, Pathak AK, Harvill ET, Preston A (2010) Antigenic Variation among Bordetella: Bordetella bronchiseptica strain MO149 expresses a novel o chain that is poorly immunogenic. J. Biol. Chem. 285: 26869-26877. doi:10.1074/jbc.M110.115121

60. Reeves P (1995) Role of O-antigen variation in the immune response. Trends Microbiol 3: 381-6. doi:S0966-842X(00)88983-0 [pii]

182

Chapter 7 : Respiratory Tract Microbiome Dysbiosis Contributes to Obstructive

Pulmonary Disease and Asthma

183

Abstract

Chronic Obstructive Pulmonary Disease (COPD) and asthma combined affect over 260 million people worldwide, and understanding how lower respiratory tract bacteria contribute to these diseases has become a major area of study. Identifying microbes associated with each disease could give insight into disease etiology, progress, and symptom management.

Haemophilus influenzae, Moraxella catarrhalis, Streptococcus pneumoniae, Enterobacteriaceae, and Pseudomonas have been identified in patients suffering from COPD, and asthma is believed to predispose patients to similar bacterial infections. Using both culture and a novel metagenomic approach, this study aimed to identify microbes associated with COPD, asthma, healthy humans, and patients suffering with respiratory problems without any disease onset. By culture, we detected Rothia dentocariosa, Streptococcus salivarius, Neisseria subflava, and

Streptococcus mitis species in healthy and diseased individuals, suggesting this subset is common amongst humans and does not contribute to disease. This study detected several known

COPD-associated bacterial species and also detected Veillonella as a novel contributing species.

In asthma patients, microbial diversity was much higher than in patients suffering with COPD.

Generally, healthy humans had more Epsilonproteobacteria and Bacteroidia species, suggesting that a dysbiosis of the lower respiratory tract microbiome may contribute to these diseases.

Together, this initial analysis of a much larger study identified novel pathogens associated with

COPD and asthma and identified microbial species associated with healthy lung activity.

184 Introduction

Although the lungs were once considered sterile, today, we are beginning to understand

how respiratory tract microflora can contribute to disease [1]. Diseases such as Chronic

Obstructive Pulmonary Disease (COPD) and asthma affected 35 million and 235 million people,

respectively, worldwide in 2004 according to the World Health Organization [2]. Advanced

culture techniques have allowed researchers to identify individual species that are associated

with common obstructive respiratory diseases, such as Chronic Obstructive Pulmonary Disease

(COPD) or asthma. Bacterial infections have been found to be associated with both COPD and

asthma. Haemophilus influenzae, Moraxella catarrhalis, Streptococcus pneumoniae,

Enterobacteriaceae, and Pseudomonas has been identified in COPD patients [3–9]; however, the

most prevalent organism identified in COPD patients is highly variable: S. pneumonia in China,

M. catarrhalis in Saudi Arabia, and H. influenzae in Germany [10–12]. In asthma, bacterial

infections including Chlamydia pneumoniae, Mycoplasma pneumoniae and Bordetella pertussis

were thought to trigger asthma exacerbations, but several recent studies have found these agents

in the lungs of healthy humans and no correlation between viral and bacterial colonization and

asthma severity [13,14]. Because these pathogens can be isolated from healthy humans and only encompass a small percentage of total respiratory tract organisms, there is still much to learn

about how respiratory tract microorganisms contribute to COPD and asthma disease symptoms,

prevention, and long-term progress.

While the etiologies of these obstructive respiratory diseases are fairly well understood,

how the respiratory tract microbiome and pathogen colonization contribute to disease onset,

symptoms, and progress is less understood. We currently know that COPD patients who are

treated with antibiotics have lower relapse rates and decreased symptoms, regardless of disease

185 state or progression [3,15]. However, current recommendations suggest only severely ill patients

or ones at high risk for respiratory failure and death receive antibiotics, because of negative

effects surrounding antibiotic resistance [16]. The use of antibiotics during early childhood has

been highly associated with the development of asthma, and the addition of antibiotics to drug

treatment cocktails during emergency room asthma presentations may even increase adverse

effects [17–19]. These studies suggest that antibiotics may cause or worsen asthma symptoms, contrary to effects of antibiotic treatment in COPD patients, suggesting bacteria may contribute differently to these etiologically similar diseases.

This study aims to investigate bacterial associations with COPD and asthma using a

novel metagenomic sequencing approach that will identify both culturable and unculturable

organisms. Additionally, we will compare metagenomic sequencing results from COPD and

asthma patients to those of healthy humans from the same community. Using this technique, we

will be able to identify organisms solely associated with COPD or asthma, determine the

respiratory tract microbiome composition of healthy individuals, and, ultimately, elucidate how

entire bacterial communities may contribute to obstructive respiratory disease. Here, our

metagenomics approach identified more Bacilli bacteria in patients with COPD and asthma,

while Bacteroidia species, such as Prevotella, are more common in healthy individuals.

186

Materials and Methods

Sputum Sample Collection.

Sputum samples used in this study were obtained from The Penn State Milton S. Hershey

Medical Center, Hershey, Pennsylvania, USA from May, 2010 to January 2012. 19 total patient sputum samples were obtained from healthy humans (7), COPD patients (6), or asthma patients

(6). Sputum was obtained by using the Pari LC®D Nebulizer, nebulizing the patient with 3% saline for 7 minutes. Oxygen saturation and spirometric tests were then performed, and nebulization continued if the forced expiratory volume (FEV1) had not fallen by more than 20%.

Over the next 7 minutes, sputum was collected into sterile sputum pots. Samples were transported to The Pennsylvania State University, University Park, Pennsylvania, USA on ice.

Nearly 1 mL was removed for immune cell identification, and 0.3 mL were taken for bacterial culture analysis. The remaining samples was then stored at -80 °C in the presence of an equal volumetric ratio of RNAlater (Invitrogen).

Bacterial Isolation and Growth

0.2 mL of fresh sputum from each of the 11 patients was removed from the sputum pot upon arrival to University Park, PA in a sterile, ventilated hood using certified DNA/RNA free pipette tips and diluted to 10-1 and 10-4 CFU concentrations using sterile phosphate buffered saline (PBS). Diluted samples were plated on Blood Agar (BA), Luria Bertani Agar (LB),

MacConkey Agar (Mc), Bordet Gengou Agar (BG), Nutrient Agar (NA), Tryptic Soy Agar

(TSA) and grown at 37°C for one day; a duplicate set was also allowed to grow at 37°C for one week. Total colonies were counted for each media type, dilution, and incubation time. Plates containing an uncountable lawn of bacteria at 10-4 were estimated at 1×107 CFU per sample. The

187 6 most frequently recovered bacterial species grown on BA were streaked again on BA for

isolation, grown in pure culture in Luria Bertani broth, and preserved in 20% glycerol in PBS at -

80°C.

16S rRNA Sequencing

To identify organisms that were obtained in pure culture, sequencing of the 16S rRNA encoding region were preformed on each bacteria. Only 8 patients produced 6 isolatable bacteria present at levels above 105 CFU. Genomic DNA from isolated liquid cultures was extracted

using QIAamp DNA mini kit (Qiagen), and 16S rRNA gene binding primers 16SF

(AGAGTTTGATCATGGCTCAG) and 16SR (AAGGAGGTGATCCAACC) were used to

amplify the V2 region in a ~1,500 bp product. The following conditions were used for the primer

set: 94°C for 5 min; 95°C for 30 s, 56°C for 1 min, 72°C for 1.5 min, 30 times; and 72°C for 8

min. Following amplification, the PCR products were separated on a 1% agarose gel, and bands

corresponding to a target product length of 1,500 bp were excised using a UV machine (Foto

Prep). DNA was extracted from bands using MinElute Gel Extraction Kit (Qiagen) as per

manufacturer’s instructions. Samples were sent to The Pennsylvania State University Genomics

Core Facility for DNA sequencing. Forward and reverse sequences were concatenated using

MEGA4 [20]. NCBI BLAST analysis was performed using the consensus nucleotide sequences

(www.ncbi.nlm.nih.gov/blast/). The top known BLAST result was used for identification, unless

it was under 70% homology, in which case it was labeled as unknown. If BLAST hits only

returned “uncultured species,” the bacteria were labeled as uncultured.

Metagenomic sequencing and analysis.

Six initial samples were chosen for the first round of 454 metagenomics sequencing

analysis to identify unculturable organisms. These six samples came from 2 symptomatic

188 patients with no identifiable disease (NID), 1 COPD patient, 2 asthma patients, and 1 healthy

individual. Total DNA isolation was performed in a UV sterilized hood with DNA-zap

(Ambion) treated instruments and DNA and RNA free filtered pipette tips. BiOstatic FFPE

Tissue DNA Isolation Kit (Mo Bio Laboratories) was utilized to obtain total DNA with the following modifications: the samples were left at 55°C overnight, instead of one hour (step 3) and step 4 was omitted. 454 amplicon sequencing, library preparations, and sequencing was completed at The Pennsylvania State University Core Genomic facility. Samples were bar-coded and sequenced together on a 454 quad to achieve at least 20,000 reads per sample. Samples below 100 bp were omitted from sequence analysis. Sequences were initially analyzed by

Mothur to assess quality, and species were then identified using a BLAST method that compared sequences to the greengenes database (http://greengenes.lbl.gov) [21]. Sequences were then

analyzed via MEta Genome ANalyzer (MEGAN), producing phylogenetic classification and

relationships to known NCBI taxonomic trees [22].

189

Results

Blood agar isolated most culturable species from sputum.

Previously, Shahnawaz showed that the best culturable media for lung sputum is blood

Agar (BA) and MacConkey Agar (Mc) [23]. In this study, we used several additional non-

selective media including LB, NA, and TSA in the hope of achieving maximal culturing

conditions. We also included BG to potentially isolate B. pertussis, which has been associated

with asthma. Overall, more total organisms were isolated from BA than from any other media

(Figure 1), while LB, TSA, BG and NA all yielded similar numbers. Interestingly, contrary to

what was observed by Shahnawaz et al., Mc

medium was the least efficient at isolating sputum

organisms. These data suggest that BA is the most

ideal medium for isolating large quantities of

bacterial species from lung sputum.

Figure 7-1 Blood Agar isolates the most sputum Streptococcus, Niesseria, and Rothia are microorganisms. The total number of organisms isolated from either Bordet-Gengou agar (BG), blood agar (BA), nutrient commonly cultured from human sputum, agar (NA), tryptic soy agar (TSA), luria bertani agar (LB), and MacConkey’s agar (Mc) is graphed regardless of disease state. according to media type. D. X. Mina helped with this experiment. The human lower respiratory tract

microbiome has been shown to include Niesseria and Haemophilus species, but exclude some

upper respiratory tract microflora, such as Staphylococcus species, when analyzed by several

metagenomic and culture techniques (W. Cookson, SMI ICMI 2011). Here, we observe similar

results when culturing organisms present in sputum from 16 patients. In healthy humans,

190 Moraxella, Rothia, Streptococcus, Neisseria, and Corynebacterium were all isolated and

identified by 16S rRNA analysis (Table 7). Interestingly, several culturable bacterial species

Table 7 – Microorganisms cultured from COPD, asthma, NID, or healthy patients.

were conserved amongst healthy, symptomatic, COPD, and asthma patients: Rothia

dentocariosa, Streptococcus salivarius, Neisseria subflava, and Streptococcus mitis.

Surprisingly, Moraxella was only identified in healthy individuals, although it been implicated in

COPD exacerbations. In patients suffering from COPD, Gordonibacter, Lactobacillus,

Agrobacterium, and Rhizobium species were all cultured, suggesting species not currently

implicated in COPD could contribute to disease symptoms. Asthma patients had the least

cultured species diversity but were the only group to include Bacillus and Microbacterium

species. In patients with respiratory symptoms but not identifiable disease (NID), the highest

cultural diversity was observed, even higher than either group with identified COPD or asthma;

Arthrobacter and Micrococcus, and Streptococcus pneumoniae were isolated from one patient.

These observations suggest that NID patients may suffer from mild respiratory infections and perhaps not the full onset of an obstructive respiratory disease. Next, strains common to asthma,

COPD, and NID pateitns were identified to determine if particular strains associated with all

191 diseases. Neisseria sp. R-22841, Streptococcus parasanguinis, Streptococcus oral taxon, and

Rothia mucilaginosa were isolated in all obstructive disease states but not healthy individuals, suggesting that these bacteria could contribute to disease symptoms. Interestingly,

Staphyloccocus epidermidis, while not isolated from any healthy or asthma patients, was identified in NID and COPD patients, suggesting that S. epidermidis may also contribute to respiratory symptoms.

After each species was identified, the frequency of each strain amongst patients was determined (Figure 2A). Perhaps not surprisingly, 13 novel, or unidentifiable based on 16S rRNA sequence homology, bacterial isolates from human sputum were obtained from healthy,

NID, COPD, and asthmatic individuals, indicating that there is still much to be learned about the bacteria that inhabit the lower respiratory tract. The next most commonly isolated bacterium was

S. salivarius, again isolated from healthy and diseased individuals, suggesting it may not contribute to respiratory disease. All other cultured bacteria were isolated at similar frequencies.

Next, we compared the average CFU isolated of individual species isolated from individual patients (Figure 2B). Unidentified species and S. salivarius, that had high frequency, were not species not observed to colonize their host to high levels, indicating that the most frequently isolated species may be controlled. However, strains such as Streptococcus cristatus,

Streptococcus sp. ICM12, and S. epidermidis, which were only isolated in diseased or symptomatic patients, were observed to colonize the lower respiratory tract up to 10 times more than frequently isolated bacteria. These observations suggest that the outgrowth of particular species or the inability to maintain particular species may contribute to disease phenotypes.

192

Figure 7-2 Frequency and colonization of cultured host microflora species. A.) The number of times a species was identified in a patient, or the frequency, was graphed according to bacterial species. B.) The average total CFU of each species per patient was graphed according to individual organisms. D. X. Mina helped with this experiment.

193 Bacilli are associated with respiratory disease, while Bacteroidia associate with healthy

patients.

While culture data provides irreplaceable data linking specific bacterial species with disease states, often very few bacteria are culturable, suggesting a more broad approach like sequenced based detection methods should also be used. Here, we utilized a 16S rRNA library

Figure 7-3 Read frequency according to bacterial class. The frequency of which reads belonging to a particular class were identified within the total 16S rRNA 454 sequencing library.

194 technique and 454 sequencing to identify all of the bacteria present in the sputum samples from one healthy, 2 NID, one COPD, and 2 asthma patients, as an initial approach. After each sequence was identified based on its homology to greengenes database, its frequency within each sample according to bacterial class was determined (Figure 3). Strikingly, the frequency of bacteria classified in the class Bacilli were up to 5 times more common in symptomatic patients than in healthy individuals. This Gram-positive bacterial class includes Bacillus, Listeria,

Staphylococcus, Abiotrphia, Aerococcus, Enterococcus, Lactobacillus, Leuconostoc, and

Streptococcus species known to colonize or cause disease in humans as well as numerous other strains not typically isolated from humans. This result suggests that symptomatic patients may have higher bacterial loads or higher pathogen colonization within the respiratory tract, which may contribute to disease symptoms. Interestingly, the healthy human sample had significantly higher Bacteroidia species, suggesting the bacteria from this class may either outcompete bacteria that contribute to respiratory disease or that they may even protect humans from becoming colonized by unwanted species. While most other bacterial classes were isolated from almost every individual, Mollicutes, Spirochaetes, Synergista were not observed in the COPD patient. Together, this suggests that there may be bacterial dysbiosis in the lower respiratory tract, and that these dysbioses may contribute to respiratory disease.

Pseudomonadales and Enterobacteriales are associated with COPD patients.

As we delve deeper and examine the proportion of bacterial families in each patient, we can learn more about which bacterial families may contribute to disease (Figure 4). Every patient has high proportions of Clostridiales, Fusobacteriales, Bacteroidales, Actinobacteridae,

Pasteurellales, Bacillales, Neisseriales, and Lactobacillales families, and interestingly, every patient has unidentifiable bacterial species present. In attempting to identify bacterial families

195 contributing to respiratory disease, we observed that the COPD patient has very little

Figure 7-4 Bacterial community structure and relationships amongst patients. For each patient, the proportion of sequences corresponding to each bacterial phylum was determined and colored according to frequency, with the most frequent being dark pink and white being least frequent. Phylogenetic relationships based on bacterial community structure within a patient are shown above the grid, while phylogenetic relationships calculated according to frequency amongst all the patients are graphed to the left of the grid.

Spirochaetalles, Mycoplasmatales, Synergistales, or Chloroplast bacterial families, which is similar to healthy patients, but has larger proportions of Pseudomonadales and Enterobacteriales than the healthy patient. We also used these findings to compare the total bacterial communities and how they relate amongst all patients (Figure 4). Asthma patients clustered together, and

COPD patient bacterial communities were the least conserved compared to asthma patients, and

196 NID patients clustered amongst asthma patients, suggesting that the microbial communities that contribute to asthma symptoms may also contribute to overall respiratory symptoms not associated with COPD. Together, these data suggest that there are significant differences between bacterial communities associated with diseases and that culture data alone may not be sufficient to determine which bacteria cause or exacerbate disease.

Over 80 bacterial species inhabit the lower respiratory tract.

Although identifying bacterial classes and families associated with disease can have major benefits for understanding community and evolutionary structure, identifying individual species is important for determining optimal treatment strategies. To determine which individual bacterial species inhabit the lower respiratory tracts of COPD, asthma, symptomatic, and healthy patients, 454 sequencing data was used to create a phylogenetic analysis of individual species via

MEGAN (Figure 5) [22]. If individual species could not be identified, the most common family, class, or phylum was determined instead. From this analysis, we identified over 80 different known bacterial species present in the sputum of humans, indicating that the human respiratory tract contains a diverse bacterial community. Haemophilus and Veillonella were the only two species unique to COPD patients, both of which have been previously associated with this disease. Asthma patients, contrary to what was observed with culture data, contained a much more diverse set of unique microbes, including Alcaligenaceae, Planctomycetacia, Megasphaera sp. oral, Solobacterium, Atopobiumrimae, Fredmanniella, Mogibacterium timidum, Clostridium felsineum, Eubacterium nodatum, Deferribacters, and Prevotella oulorum. Interestingly,

Abiotrophia defective, four species of Eubacterium, Actinomyces, and Streptococcus viridians

197 were found in both COPD and asthma patients and not in healthy individuals, indicating these

198 Figure 7-5 Species-based phylogenetic analysis of all patients. Each species identified by 454 sequencing from each patient was assembled into a phylogenetic tree using MEGAN. The frequency of each identified species from COPD (red), NID (blue, yellow), healthy (green), and asthma (pink, aqua) is graphed next to the bacterial name, with the bar height correlating with the number of times the corresponding read was observed. J. Park created this figure.

species could contribute to respiratory symptoms regardless of disease state. Several species

were unique to healthy individuals, including Campylobacter showae, Lautropia sp. oral,

Elkenella, Dialister sp. oral, Catonella sp. oral, and Porphyromonas uenomis. Strikingly, this

healthy human sample also contained several species of Prevotella, including P. intermedia, P.

micans, and P. oralis. Together, these data suggest that the respiratory tract contains a diverse set of organisms and that individual bacterial species may contribute to the disease symptoms observed in COPD and asthma patients.

199

Discussion

This initial study examines both the culturable and unculturable bacterial species

associated with respiratory disease, namely COPD and asthma, and compares them to healthy

individuals. Culture and sequencing methodologies revealed slightly different conclusions,

although species unique to COPD, asthma, and symptomatic patients can all be detected using

any method. Overall, higher bacterial loads in the respiratory tract were associated with more

severe disease symptoms. A metagenomics approach identified that diseased patients were generally found to have more Bacilli organisms. Similar to results in other COPD studies that examine microbial associations with the disease, Pseudomonadales and Enterobacteriales were identified in higher numbers in COPD patients, while asthma patients appeared to have a very diverse collection of bacteria. Furthermore, these results suggest that healthy humans harbor very different microbiota compared to diseased patients; several Moraxella, Bacteroidia, and

Prevotella species were only detected in healthy humans, suggesting they may play a role in

blocking pathogen colonization and probably do not contribute to respiratory symptoms.

Current research suggests that bacterial infection or presence contributes to the disease

symptoms of both COPD and asthma, although which species cause these symptoms and how

they contribute to disease etiology is still not largely understood. Here, we identify numerous

Bacilli species present in both COPD and asthma patients. Within this Gram-positive class,

many common respiratory pathogens exist, such as Bacillus anthracis, Streptococcus

pneumoniae, Listeria monocytogenes, and Enterococcus faecalis [24–27]. S. pneumoniae has

been reported as COPD associated, while asthma is believed to predispose patients to S.

pneumoniae infections [28]. In this study, S. pneumoniae was only isolated from symptomatic

200 patients, while S. viridians S. mutans, S. cristatus, S. ICMI12, and S. epidermidis were all identified from both diseased and symptomatic patients but not healthy individuals, suggesting that various Streptococcus species contribute to respiratory symptoms. However, healthy individuals were not devoid of Streptococcus species; interestingly, S. midis and S. salivarius were identified in diseased and healthy individuals, suggesting that they may not play a role in disease. Often species within one genus have mechanisms of competing or cooperating with species within its own genus, suggesting that these Streptococcus species may compete or synergize within the host [29–31]. Understanding which species, even within one genus, contribute or initiate disease will be paramount in understanding how to effectively treat patients.

This study suggests that metagenomic sequencing approaches may be the only effective way to distinguish species to this extent and begin to understand how species within one genus may or may not contribute to disease. Furthermore, recent evidence suggests that Streptococcus species can provide a disadvantage for Niesseria species, indicating that these Streptococcus species may dictate colonization by subsequent species [32].

One goal of this study was to identify species unique to healthy humans, in hopes that these species may be able to prevent infection or be involved in dysbiosis. In this study, we observed that Moraxella osloensis was the only cultured organism unique to healthy individuals.

Interestingly, Moraxella catarrhalis has been implicated in COPD, so understanding differences between these strains could help elucidate individual bacterial pathogenesis mechanisms behind how the Moraxella species contribute to disease symptoms. Using 454 metagenomic analysis, we observed that Epsilonproteobacteria and Bacteroidia classes were more commonly identified in a healthy individual. When we looked at what species contributed to this observation, we identified Campylobacter showae (Epsilonproteobacteria), Prevotella intermedia, Prevotella

201 micans, and Prevotella oralis (Bacteroidia) from healthy individuals. Interestingly, a recent study attempting to identify bacteria associated with obesity identified that healthy individuals had more Bacteroidia species in their gut microbiome compared to their obese counterparts [33].

Although the gut and respiratory tract have very different environments and possess very different bacterial communities, it appears that Bacteroidia species, such as Prevotella, may contribute to health homeostasis. Future studies determining whether or not these species can protect against subsequent pathogen colonization or are simply displaced during disease establishment will need to be conducted. Furthermore, understanding how Bacteroidia species contribute to immune priming and overall respiratory health will be essential to understanding how the microbiome and immune system interact.

202

Author Contributions:

Authors: Laura S. Weyrich1,2, Jihye Park1,3, Dan X. Mina1, Sarah E. Young1, and Eric T.

Harvill1

1 Department of Veterinary and Biomedical Sciences, The Pennsylvania State University,

115 Henning Bldg, University Park, PA 16802

2 Biochemistry, Microbiology, and Molecular Biology Graduate Program

3 Bioinformatics and Genomics Graduate Program

Conceived and designed experiments: LSW, ETH

Preformed experiments: LSW, DXM, SEY

Analyzed data: LSW, JP, ETH

Wrote manuscript: LSW, ETH

203

References

1. Jones FS (1922) THE SOURCE OF THE MICROORGANISMS IN THE LUNGS OF NORMAL ANIMALS. J Exp Med 36: 317-328.

2. WHO | The world health report 2004 - changing history (2004). Available: http://www.who.int/whr/2004/en/. Accessed 12 Jan 2012.

3. Wilson R (2001) Bacteria, antibiotics and COPD. European Respiratory Journal 17: 995 - 1007.

4. Martínez-Solano L, Macia MD, Fajardo A, Oliver A, Martinez JL (2008) Chronic Pseudomonas aeruginosa Infection in Chronic Obstructive Pulmonary Disease. Clinical Infectious Diseases 47: 1526 -1533. doi:10.1086/593186

5. De Serres G, Lampron N, La Forge J, Rouleau I, Bourbeau J, et al. (2009) Importance of viral and bacterial infections in chronic obstructive pulmonary disease exacerbations. Journal of Clinical Virology 46: 129-133. doi:10.1016/j.jcv.2009.07.010

6. Bari MR, Hiron MM, Zaman SM, Rahman MM, Ganguly KC (2010) Microbes Responsible For Acute Exacerbation of COPD. Mymensingh Med J 19: 576-585.

7. Wilson R (1998) The Role of Infection in COPD. Chest 113: 242S -248S. doi:10.1378/chest.113.4_Supplement.242S

8. Murphy TF, Sethi S, Niederman MS (2000) The Role of Bacteria in Exacerbations of COPD*. Chest 118: 204 -209. doi:10.1378/chest.118.1.204

9. Marin A, Monsó E, Garcia-Nuñez M, Sauleda J, Noguera A, et al. (2010) Variability and effects of bronchial colonisation in patients with moderate COPD. Eur. Respir. J 35: 295- 302. doi:10.1183/09031936.00126808

10. Ko FWS, Ip M, Chan PKS, Fok JPC, Chan MCH, et al. (2007) A 1-year prospective study of the infectious etiology in patients hospitalized with acute exacerbations of COPD. Chest 131: 44-52. doi:10.1378/chest.06-1355

11. Lode H, Allewelt M, Balk S, Roux A, Mauch H, et al. (2007) A Prediction Model for Bacterial Etiology in Acute Exacerbations of COPD. Infection 35: 143-149. doi:10.1007/s15010-007-6078-z

12. Alamoudi OS (2007) Bacterial infection and risk factors in outpatients with acute exacerbation of chronic obstructive pulmonary disease: a 2-year prospective study. Respirology 12: 283-287. doi:10.1111/j.1440-1843.2006.01002.x

204 13. Harju TH, Leinonen M, Nokso-Koivisto J, Korhonen T, Räty R, et al. (2006) Pathogenic bacteria and viruses in induced sputum or pharyngeal secretions of adults with stable asthma. Thorax 61: 579 -584. doi:10.1136/thx.2005.056291

14. Thumerelle C, Deschildre A, Bouquillon C, Santos C, Sardet A, et al. (2003) Role of viruses and atypical bacteria in exacerbations of asthma in hospitalized children: A prospective study in the Nord‐Pas de Calais region (France). Pediatric Pulmonology 35: 75-82. doi:10.1002/ppul.10191

15. Adams SG, Melo J, Luther M, Anzueto A (2000) Antibiotics Are Associated With Lower Relapse Rates in Outpatients With Acute Exacerbations of COPD*. Chest 117: 1345 -1352. doi:10.1378/chest.117.5.1345

16. Chronic Obstructive Pulmonary Disease: Role of Bacteria and... : Drugs & Aging (n.d.). Available: http://adisonline.com/aging/Abstract/2009/26120/Chronic_Obstructive_Pulmonary_Disease __Role_of.2.aspx. Accessed 24 Jan 2012.

17. ROWE BH, EDMONDS ML, SPOONER CH, CAMARGO CA (n.d.) Evidence-based treatments for acute asthma. Respiratory care 46: 1380-1390.

18. Droste JHJ, Wieringa MH, Weyler JJ, Nelen VJ, Vermeire PA, et al. (2000) Does the use of antibiotics in early childhood increase the risk of asthma and allergic disease? Clinical & Experimental Allergy 30: 1548-1553. doi:10.1046/j.1365-2222.2000.00939.x

19. Graham V, Lasserson TJ, Rowe BH (n.d.) Antibiotics for acute asthma.. Available: http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD002741/abstract. Accessed 24 Jan 2012.

20. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596-9. doi:msm092 [pii] 10.1093/molbev/msm092

21. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, et al. (2009) Introducing mothur: Open-Source, Platform-Independent, Community-Supported Software for Describing and Comparing Microbial Communities. Appl. Environ. Microbiol. 75: 7537- 7541. doi:

10.1128/AEM.01541-09

22. Huson DH, Auch AF, Qi J, Schuster SC (2007) MEGAN analysis of metagenomic data. Genome Res 17: 377-86. doi:gr.5969107 [pii] 10.1101/gr.5969107

23. Shahnawaz A, SM Saleem, Mushtaq A Lone Shah Sonaullah Tariq Bhat, Gh. Bhat, Gh. Nabi Dhobi (2003) Bacteriological Profile in Acute Exacerbation of Chronic Obstructive Pulmonary Disease (COPD). JK-Practitioner 10: 185-187.

205 24. Inglesby TV, Henderson DA, Bartlett JG, Ascher MS, Eitzen E, et al. (1999) Anthrax as a Biological Weapon. JAMA: The Journal of the American Medical Association 281: 1735 - 1745. doi:10.1001/jama.281.18.1735

25. Patchett RA, Kelly AF, Kroll RG (1991) Respiratory activity in Listeria monocytogenes. FEMS Microbiology Letters 78: 95-98.

26. Chen K-Y, Hsueh P-R, Liaw Y-S, Yang P-C, Luh K-T (2000) A 10-Year Experience With Bacteriology of Acute Thoracic Empyema*. Chest 117: 1685 -1689. doi:10.1378/chest.117.6.1685

27. Whitney CG, Farley MM, Hadler J, Harrison LH, Lexau C, et al. (2000) Increasing Prevalence of Multidrug-Resistant Streptococcus pneumoniae in the United States. N Engl J Med 343: 1917-1924. doi:10.1056/NEJM200012283432603

28. Talbot TR, Hartert TV, Mitchel E, Halasa NB, Arbogast PG, et al. (2005) Asthma as a Risk Factor for Invasive Pneumococcal Disease. N Engl J Med 352: 2082-2090. doi:10.1056/NEJMoa044113

29. Hibbing ME, Fuqua C, Parsek MR, Peterson SB (2010) Bacterial competition: surviving and thriving in the microbial jungle. Nat Rev Micro 8: 15-25. doi:10.1038/nrmicro2259

30. Joo J, Gunny M, Cases M, Hudson P, Albert R, et al. (2006) Bacteriophage-mediated competition in Bordetella bacteria. Proc Biol Sci 273: 1843-8.

31. Griffin AS, West SA, Buckling A (2004) Cooperation and competition in pathogenic bacteria. Nature 430: 1024-1027. doi:10.1038/nature02744

32. Shakhnovich EA, King SJ, Weiser JN (2002) Neuraminidase Expressed by Streptococcus pneumoniae Desialylates the Lipopolysaccharide of Neisseria meningitidis and Haemophilus influenzae: a Paradigm for Interbacterial Competition among Pathogens of the Human Respiratory Tract. Infect Immun 70: 7161-7164. doi:10.1128/IAI.70.12.7161- 7164.2002

33. Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, et al. (2005) Obesity alters gut microbial ecology. Proceedings of the National Academy of Sciences of the United States of America 102: 11070 -11075. doi:10.1073/pnas.0504978102

206

Chapter 8 : Teaching Ethical Aptitude to Graduate Student Researchers

207 Even though experimentation dates back to the ancient Greeks, effectively teaching the ethical principles that underlie that research is relatively new. Responsible Conduct of Research

(RCR) training for any undergraduate, graduate, or post-doctoral researcher conducting National

Science Foundation (NSF) sponsored research was mandated in 2007, after section 7009 of the

America Creating Opportunities to Meaningfully Promote Excellence in Technology, Education,

and Science (COMPETES) Act was passed [1]. This non-standardized training was then later

implemented by the National Institute of Health (NIH) [2,3]. RCR training commonly utilizes

case studies to teach students about the nine core domains: human and animal experimentation,

mentoring, collaboration, peer review, data management and ownership, publication practices,

authorship, conflicts of interest, and research misconduct [4–6]. Additionally, graduate students

in biomedical research fields are also usually required to undergo animal experimentation

training under the oversight of the university’s Institutional Animal Care and Use Committee

(IACUC).

Although IACUC and RCR training should be greatly applauded, there are major caveats

to the training approaches used today. First, to meet RCR requirements, most universities have

implemented semester long RCR courses and online modules as part of their first year graduate

student curriculum, nearing 30 hours total instruction [2,3]. Alternatively, some universities

have implemented brief weekly discussions, totaling no more than 12 hours yearly training that

equates to 36-48 hours of total instruction over the course of a PhD education [7]. If this is

broken up evenly amongst the nine suggested RCR training domains, only three to four hours of

training is invested into each area. For example, if we look at the first core domain, human and

animal experimentation, half that time would be attributed to animal research (1.5 hours) and

even a smaller portion attributed to experimental methodologies, design, and conduct of animal

208 research (30 minutes each). Is 30 minutes enough time to convey comprehension and the ability

to fully discuss ethical quandaries within each area, especially if we are using case studies, role

playing, or a lecture to get the ethical implications of experimental design across?

Secondly, a current and very common RCR training method involves the use of case studies as educational tools [8]. Case studies can be an effective method to teach students what not-to-do, start discussion, or relate ethical dilemmas to the research they conduct; however, inevitably scientists are faced with new questions that novel experimental design, technology, and changing interests induce. Historical or previously resolved cases may not develop the creativity or problem solving skills necessary for students to make decisions when new problems arise, forcing students to make judgment calls only based on their background or personal previous experience.

Lastly, much of current training suggests that researchers simply “know the rules” and abide by the regulations set forth. Many of the principles, especially associated with human or animal experimentation, are served in a “checklist” format, asking the researcher in a yes or no design if they met particular criteria or heeded certain regulations. For example, even the most basic training teaches animal researchers to abide by the 3 R’s: reduction, refinement, and replacement, suggesting researchers reduce the animal numbers by refining experimental design and replacing animal experimentation where possible [9]. Even this is easily refined to three questions that can be answer yes or no in a check-list format. Although a “checklist” format can be a highly useful tool and serves as a means of ensuring ethical structure, it does not teach the individual student why he/she is following the guidelines, simply that he/she must follow them.

Even though the commitment to teaching effective RCR exists at most universities, the problems with this training may be rooted in the manner and method in which it is taught. To

209 avoid many of these dilemmas, graduate ethics training would need to go beyond basic training

modules, monthly seminars, or simple case studies. Today, a fast paced, ever changing,

competitive environment forces RCR training to teach students to design experiments only based

on the scientific question, as long as a very objective-based “ethical checklist” can also be satisfied. The guiding principles behind how many of these “checklist items” came to be, why they exist, or more importantly, whether or not they are appropriate in particular situations are overlooked. Especially in human or animal research, graduate students are trained to follow basic guidelines, without ever being pushed to think deeper about the ethical implications and standards by which those guidelines were conceived. Recently, Seiler et al. suggested we find better ways to implement RCR training by promoting “a deeper appreciation of RCR by shifting the focus away from wanting to simply ‘know the rules’” [10]. The question is, how do universities effectively do this?

I propose that graduate students should be taught foundational ethical theories, as a means for teaching not only ethical standards, but inspiring a more nuanced, creative approach to conducting scientific research. For example, the three classical bioethical theories, including virtue ethics (Aristotle), consequentialist theory (Bentham and Mill), and deontological theory

(Kant), would teach students how the ethical and moral standards came to be and potentially give insight into why scientist should be bound to them. Another example may suggest that students completing animal research are taught Cartesian philosophy, to understand the grounding and basis behind current ethical treatment of animals. Teaching any of these theories would pair sociological rational with scientific methods, facilitating increased perception and cognitive capabilities. Very basic ethics principles will better teach researchers how to make decisions in challenging situations, gauge when animal research is justifiable or appropriate, or even weigh

210 how their experimental findings will impact individuals or society as a whole. This could be achieved by implementing a bioethical theory or methods course into the required curriculum of each incoming graduate student class. This, taught alongside current RCR training modules, would allow graduate students to connect the dots behind current responsible conduct codes and the ethical principles that underlie each “checklist” rule. Most importantly, this process will engage students to develop higher standards in ethical scrutiny and encourage a better understanding the overall impacts and ethical implications of their research.

Here, I propose an example in which to illustrate my point. In veterinary and biomedical research, bacterial or viral genetic mutants are often constructed as means to study an individual gene or protein. After mutant construction, rigorous testing usually proceeds to determine if the mutation is medically relevant, contributes to disease, persistence, immunity, survivability, host interactions, etc. This type of investigation almost always requires animal experimentation.

During a recent comprehensive exam, a student was asked how he/she should proceed after mutant construction, either with option A.) to complete animal experimentation first to prove actual relevance to medical health or B.) to subject the mutant to a battery of in vitro work to determine if in vivo work should proceed. The immediate obvious answer was option B, to complete numerous in vitro experiments before proceeding with animal experimentation, because this satisfied one of the 3 R’s they had been taught in RCR training.

The problem is that neither of these options is very obviously wrong or better than the other one. Option A, completing animal experimentation first has numerous benefits. Relevance to medical benefits is verified much earlier, before time, money, and resources, often paid by

American taxpayers, are invested into investigating a phenotype. Additionally, the phenotypes observed in vivo can help gauge the best in vitro experiments to later biochemically analyze the

211 functional mechanism. One the other hand, option B, completing in vitro or in silico experimentation first, has alternative benefits. This approach may decrease initial mouse numbers and help predict mechanistic functions, or at least narrow hypothesis down to testable ideas.

By today’s RCR training methods, many graduate students assume the first approach using in vivo experimentation initially is “wrong,” because any animals are being used in the first stage of experimentation, when, in reality, laboratories sometimes use this approach for good reason. In these types of experiments, a gene is mutated because there is sufficient evidence for investigation, already suggesting there is rational to proceed with in vivo experimentation.

Interestingly, both of the proposed experimental paths – to identify in vivo phenotypes first or do animal experimentation only after multiple in vitro tests – can satisfy the ethical requirements put forth by The Guide, Animal Welfare act, and PHS Policy, depending on the procedure, experimental design, and overall goal of the experiment. Many laboratories have a common methodology they follow to conduct experiments, generally choosing option A or B, usually depending upon how the principle investigator or major advisor was trained. However, the ultimate purpose or goal of each experiment needs to be examined for the individual researcher to gauge the proper steps and action for that particular experiment. For example, few would argue that a potential SARS vaccine candidate strain should have only first been tested in vitro during the outbreak in 2003. Situational information exists behind every experiment and, with the help of rigorous ethical theory, can properly guide students how to make the best decisions during research. The purpose of teaching such principles is not only to heighten scientist’s awareness of ethical principles, but to additionally transform RCR research into an additional means to trigger high level thought processes and learning. Combining humanities

212 methodologies that inform the field of bioethics with scientific methods creates a more

competent, higher caliber scientist capable of not only completing experiments necessary to push

knowledge forward, but who also posses the ability to decipher and understand the far-reaching impacts of their work.

Why do we teach some of the best scientific minds in the country a naïve view of research ethics that simplistically suggests that complex issues are black or white, yes or no, right or wrong, or can be simply checked off a list? Because of today’s research complexities, students need to not only know what the best option is, but also how to rationalize and understand why a decision is best. We need to go beyond a moral checklist and teach researchers the ethical theories underlying their research, allowing them to rationalize better, more applicable experiments and overall train better, more successful scientists. Most obviously, better understanding the ethical issues and implications of one’s research can minimize corrupt scientific practices and breaches in ethical behavior. Secondly, encouraging scientists to go beyond their scientific expertise and think creatively through engaging methodologies and concepts from the humanities, will most certainly compliment their scientific training, reminding scientists that much of their work goes beyond each experiment and impacts society at large, affecting individual lives. Better training may increase the applicability of science and help decrease the gap between basic science and the applications of that research. Lastly, numerous recent publications have highlighted the lack of trust and interest society has in today’s scientific findings. Bettering scientific training will fundamentally address the crisis of trust, helping to address the gap between science and society. Restoring public faith in science must start with the scientists themselves, by creating a higher caliber of scientists through increased rational and

213 logic training, better ethics training, and increased awareness that guide scientists to return to the reason they were called to science in the first place – to create knowledge and better society.

214

Author Contributions:

Laura S. Weyrich1,2,3

1Biochemistry, Microbiology, and Molecular Biology Graduate Program

2Veterinary and Biomedical Science Department, The Pennsylvania State University

3Science, Technology, and Society Department, The Pennsylvania State University

215

References

1. Institute of Laboratory Animal Research, Commission on Life Sciences, National Research Council (1996) Guide for the Care and Use of Laboratory Animals. Washington, D.C.: The National Academies Press. p.

2. Government and Professional Resources : Animal Welfare Information Center (1966) Animal Welfare Act. p. Available: http://awic.nal.usda.gov/nal_display/index.php?info_center=3&tax_level=4&tax_subject=1 82&topic_id=1118&level3_id=6735&level4_id=11092&level5_id=0&placement_default= 0. Accessed 16 Oct 2011.

3. Office of Laboratory Animal Welfare: PHS Policy on Humane Care and Use of Laboratory Animals (n.d.). Available: http://grants.nih.gov/grants/olaw/references/phspol.htm?print=yes&. Accessed 16 Oct 2011.

4. van Zutphen B (2007) Invited International Perspective - Education and Traingin for the Care and Use of Laboratory Animals: An Overview of Current Practices. ILAR 48. Available: http://dels-old.nas.edu/ilar_n/ilarjournal/48_2/html/v4802vanZutphen.shtml. Accessed 21 Oct 2011.

5. 110th Congress (2007) H.R.2272 -- America Creating Opportunities to Meaningfully Promote Excellence in Technology, Education, and Science Act.. Available: http://thomas.loc.gov/cgi-bin/query/D?c110:5:./temp/~c110PswqSC::#. Accessed 21 Oct 2011.

6. DuBois JM, Schilling DA, Heitman E, Steneck NH, Kon AA (2010) Instruction in the responsible conduct of research: an inventory of programs and materials within CTSAs. Clin Transl Sci 3: 109-111. doi:10.1111/j.1752-8062.2010.00193.x

7. Kon AA, Schilling DA, Heitman E, Steneck NH, DuBois JM (2011) Content Analysis of Major Textbooks and Online Resources Used in Responsible Conduct of Research Instruction. AJOB Prim Res 2: 42-46. doi:10.1080/21507716.2011.564263

8. Horner J, Minifie FD (2011) Research ethics II: Mentoring, collaboration, peer review, and data management and ownership. J. Speech Lang. Hear. Res. 54: S330-345. doi:10.1044/1092-4388(2010/09-0264)

9. Horner J, Minifie FD (2011) Research ethics I: Responsible conduct of research (RCR)-- historical and contemporary issues pertaining to human and animal experimentation. J. Speech Lang. Hear. Res. 54: S303-329. doi:10.1044/1092-4388(2010/09-0265)

216 10. Horner J, Minifie FD (2011) Research Ethics III: Publication Practices and Authorship, Conflicts of Interest, and Research Misconduct. Journal of Speech, Language & Hearing Research 54: S346-S362. doi:10.1044/1092-4388(2010/09-0263)

11. Peiffer AM, Laurienti PJ, Hugenschmidt CE (2008) Fostering a Culture of Responsible Lab Conduct. Science 322: 1186. doi:10.1126/science.322.5905.1186b

12. Macrina F (2005) Scientific integrity : text and cases in responsible conduct of research. 3rd ed. Washington D.C.: ASM Press. p.

13. Russell WMS (1959) The principles of humane experimental technique. Methuen.

14. Anderson LC (2007) Institutional and IACUC Responsibilities for Animal Care and Use Education and Training Programs. ILAR 48. Available: http://dels- old.nas.edu/ilar_n/ilarjournal/48_2/html/v4802Anderson.shtml. Accessed 21 Oct 2011.

15. Seiler SN, Brummel BJ, Anderson KL, Kim KJ, Wee S, et al. (2011) Outcomes Assessment of Role-Play Scenarios for Teaching Responsible Conduct of Research. Accountability in Research 18: 217-246.

217

Chapter 9 : Is there a solution to microbiome patenting?

218

The human microbiome project (HMP), an 8.2 million dollar National Institute of Health initiative to identify microbial communities that inhabit the human body and determine how they contribute to disease, has incited discussion about the ethical and legal implications surrounding isolation and identification of an individual’s microbes [1]. Although numerous ethical and legal issues, including patient rights, data sharing, sampling invasiveness, and impacts on social networks, have recently come to light, Hawkins and O’Doherty recently raised new ethical questions regarding biobanking, patenting, and ultimately ownership of isolated microbes, in their manuscript entitled, “Who owns your poop?: insights regarding the intersection of human microbiome research and the ELSI aspects of biobanking and related studies” [2,3]. These authors begin to delve into ownership of microbes contained on or within the human body, an issue previously, for the most part, undiscussed.

Microbial ownership and, ultimately, the ability to patent organisms within the microbiome boil down to deciding whether or not the microbiome is part of ones being or self, an argument briefly discussed by Hawkins and O’Doherty.

“Currently it is not clear whether these microorganisms should be considered part of or separate from the human body. Arguments can be made for both, but ultimately it seems that the dichotomy of human versus non-human and self versus non-self inevitably breaks down in this context.”

To evaluate this question in great detail, we must first define if microbes are essential to human life. The microbiome has been shown to contribute to immune system development, digest food, produce vitamins, and protect against disease, suggesting our lives would be much worse without them [4]. Whether or not we could physically survive without them is still unclear, especially with increasing biomedical advancements. Secondly, the microbiome can be defined as a tool used for identification, similar to fingerprints or genetic sequence. Recent research has

219 suggested individuals could be identified based on their microbial contents, or that even their sociological or travel habits could be elucidated by microbiological study of an individual [5].

Lastly, we must determine if the microbiome is stable or if a subset of the organisms present remains permanent. Although leading research suggests that many of the organisms identified in several longitudinal studies are transient, the HMP will reveal a much larger data set, when this question can finally we thoroughly addressed [6,7]. To further compound this, we currently lack a solid description defining if or for how long a research sample belongs to an individual after it has been removed, which leads very large open windows for who can then subsequently claim ownership [2]. For example, the immortal story of Hennrietta Lack’s cancer cells paints the perfect example of how corporate ownership resulted in profit from an individual’s sample, and how that sample no longer belonged to the individual after it was “donated” [8]. The fact of the matter is that until more research is done identifying the core microbiome, determining which bacterial communities contribute to which diseases, and identifying species vital to cellular processes, we cannot be sure whether or not our microbiome is indeed part of our autonomous being.

Even though we are currently unsure if the microbiome is synonymous with self, numerous groups have already initiated patenting microbes as probiotics, or organisms which may be administered to potentially increase our health. We are currently trying to identify microbes that will treat inflammatory bowel disease symptoms, prevent obesity, protect against the common cold, explain autism, or cure diabetes, all diseases in which the microbiome has now been linked [9–15]. Administering a microbial cocktail to cure a disease is already widely accepted amongst the general public, in the use of probiotic organisms present in yogurt. Novel therapies where microbes are given to a patient are currently under investigation, and ownership

220 over the bacteria central to these therapies will become a crucial societal issue, especially in the

United States where patenting is highly utilized business strategy. Additionally, novel antibiotics

could be identified by studying how microbiota compete with invading pathogens, suggesting

that microbial ownership could also translate into antibiotic production rights. The microbiome

is teaming with unidentified organisms, ready to be studied and exploited for novel anti-bacterial tools.

However, in order to allow patenting of microbiome microorganisms, several goals must be met. The rights of the individual patient or sample-giver must be considered, especially if their microbiome is part of their autonomous being. On the other hand, corporate interests must be met to encourage industrial science, and ensure research funding to fuel a huge pharmaceutical industry. However, government funded and academic researchers must not be limited by patenting. Lastly, current laws and regulations must be abided and maintained. So, how do we follow these rules and develop a patenting system that works for all parties involved?

In order to examine this issue, three different scenarios have been investigated as potential solutions.

Scenario #1: Corporate patenting.

As the common saying goes, “everything under the sun can be patented.” Currently, patenting of microbial organisms is allowed in the United States, as long as the microbe has been

“human-made” or manufactured, suggesting that simple possession of the organisms does not give one the right to patent it. A landmark United State Supreme Court case, Diamond vs.

Chakrabarty, determined that a live, human-made microorganism is patentable, indicating that work must have been completed on the organism to modify it away from its natural state [16,17].

221 Furthermore, individual microbial proteins have also been patented, most recently as potential novel cancer therapeutics [18].

Using these current standings, an individual does not have rights to his/her microbes and certainly would not be able to patent their own microorganisms. However, corporate interests capable of modifying organisms they obtained during research studies would have rights to certain organisms, regardless of where they were obtained. This requires corporate entities to invest effort, time, and funds to modify the organism to some extent; however, how much the bacteria must be modified is still open for debate. Does inserting several novel genes into an organism making it sensitive to antibiotics constitute enough manufacturing for a microbe that was biologically discovered to fight obesity? Or do yet unidentified human microbes posses novel genetic elements that could easily be exploited or claimed as novel or unique? If so, a patenting frenzy could emerge, and grey lines are easily established with the potential to cost industry millions of dollars in legal fees. Nevertheless, numerous biotechnology firms have argued that patents are essential to increase innovation [19].

Currently, research subjects donate their samples and sign a release waiver, commonly relinquishing any ownership to the microbes. However, this scenario also begins to raise questions behind why a research subject would sign up to even donate samples. Are patients being amply compensated in a just and fair way? Is the government protecting the rights of citizens by allowing corporations to sell microbes that may constitute part of one’s being or identity? Should the subject be given a portion of the profits earned from patenting one of their microbes? Furthermore, this may raise questions concerning the motives of science, or at least what the public would perceive as motives. New systems of compensating patients, such as a

222 probiotic tax, could be installed to ensure patients are justly compensated, although we are

currently far from this possibility.

Corporate ability to patent one individual microbiome microbe additionally raises

questions as to whether or not entire bacterial communities are patentable. Currently, fecal

transplants are being used to treat patients suffering from Clostridium difficile infections of the

gut [20]. This procedure requires an entire cocktail of microbes to be inserted into the intestine,

not just one individual organism. Would companies be allowed to patent whole communities or

subsets of communities? Furthermore, would companies have to buy rights to microbes other

companies owned in order to develop new therapies? Even more troubling, would academic and

government funded researchers, not working for profit, be allowed to investigate microbes that

were patented? The recent debacle of human gene patenting proved this to be an issue, as

patenting of the BCRA1 genes stifled research in numerous British labs that were working

towards a breast cancer cure and generally discouraged researchers from trying to construct

diagnostic tests [21,22].

Scenario #2: Allowing individuals rights to their microbiome.

Although almost unimaginable, we can envision a world when each individual is capable

of patenting their own microbes, citing that each microbe, or even a subset or microbes, are

patentable because work was done to create an environment where these microbes flourished.

Traveling to Asia, Europe, and Africa has allowed an individual to collect a unique subset of

microbes that can all live congruently and potentially have evolved novel genomic sequences or

obtained unique genome sequences. Furthermore, individuals could be taught necessary

molecular cloning skills to manipulate their microbes in a way that was deemed patentable.

223 However, ownership over one’s own microbes would significantly change several current

U.S. laws. One example would be the sewage rights. Currently, cities own sewage piped into a facility for cleaning, purification, etc. Needless to say, if each individual owned their microflora, namely their gut microflora, precious property would be flushed away from the home, gone forever [4]. If individuals owned their own microbes, the definition of property rights in the U.S. would have to change. People’s property would be deposited with every touch, sneeze, scratch, run, tear, and bathroom trip they took. A simple touch of another person’s skin may allow a person to steal personal property, forcing new laws to guide very simple, day to day social norms. Although robotics or creating chemical compounds that secure our microflora could account for these changes, ultimately, allowing microbes to be considered personal property could change the way interact with other humans, a world where this existed is hard to imagine.

Treating human microflora as property and allowing patenting would ultimately allow people to sell and profit from microorganisms that have been around for millions of years, a tactic that seems as ridiculous as patenting trees or yoga poses. Furthermore, this scenario raises questions about the ability to sell biological items on the market. Currently, in the U. S., citizens are not allowed to sell vital organs, such as kidneys, lungs, and hearts, by the National Organ

Transplant Act (NOTA) of 1984, although they can be donated if the life is terminated through accidental means. If microorganisms are indeed part of the one’s identifiable self and essential to life, then they should fall under NOTA and not be commercially available to anyone.

Allowing individuals to profit from their microbiome could easily create harmful situations for humans, as they attempt to obtain microorganisms through potentially highly invasive procedures. The selling and purchasing of microflora would need to be tightly regulated, as it already is for corporations and research institutions. Furthermore, determining who has rights to

224 core-microbiome microbes, or microorganisms shared amongst all humans, is another issue that would need to be resolved.

This scenario additionally raises problems for corporate entities. With the incentive for ingenuity and product creation, how will biotech companies thrive and be successful? Will they have the overhead needed to purchase patentable microbes from individuals? Perhaps companies could increase patenting in areas like bacterial genes, instead of whole microbes, or alternatively patent the procedures and materials used to administer these organisms.

Scenario #3: No microbiome patenting allowed.

Because of the problems identified in scenarios 1 and 2, we can conceive a third scenario where no ownership or microbiome patenting is allowed. This approach would rectify many of the issues so far addressed in this study. Corporations would still be able to obtain the microbes for use in research and therapeutics, while not sacrificing the rights of the individuals.

Individuals would still have the advantage of new disease treatments without selling a part of their body or feeling as though the altruism of science was missing. New property laws for an individual’s microbes would not have to be rewritten, and academic and government researchers could continue to contribute to this ever growing field. Furthermore, individuals or companies who did not possess the capabilities or opportunities to patent microbes would not be disadvantaged, although this is far from the world we live in today.

On the other hand, this final scenario does not allow any one individual to profit from these microorganisms and may not allow current industrial microbiome research to be amply compensated for their investments. The potential invention of microbes as disease prophylactics or probiotics could be hindered, as the incentive and ultimate benefit for companies to invest in

225 creating these products may be gone. However, corporations could still file for patents regarding the process or procedures in which microbes are administered to humans or the matrixes in which these microbes are grown, stored, or mixed for delivery. Because this is a largely uninvestigated area, there is potentially plenty of room for novel patents. Are we, as a society, understanding of this proposition, or would we reject this scenario because of the market driven, and perhaps capitalistic, economy that currently drives these interactions?

A Potential Solution

The familiar market-driven, and perhaps even capitalistic, interactions between biological science and industry have not always been commonplace. Before 1970, President Nixon had not yet declared war on cancer which funneled much of the research dollars into biomedical research

(1971), the Bayh-Dole Act of 1980 had not shifted the legal landscape allowing academic research institutions to patent intellectual property, and the Diamond vs. Chakrabarty case (1980) had not yet determined that biological microbes were patentable objects [17,23,24]. Historical occurrences, such as these, dramatically changed the relationship between science and industry and created the environment we have today, a landscape ripe for prospering biotechnology industries. Interestingly, scientists and researchers being trained today may be unaware of how political and legislative decisions have molded many of the resolutions that are an integral part of science today. The need to discuss whether or not microbes isolated during microbiome research are patentable stems from political and legislative decisions such as these. Discussing microbiome patents gives us a unique opportunity to examine the relationships between science and industry, and the once, perhaps gone, purely altruistic intentions of biomedical research.

226 Robert Merton examined a set of scientific norms in 1942 that governed all science and determined that there were four major principles guiding how scientific research ought to be conducted: communalism, universalism, disinterestedness, and skepticism. Communalism, suggesting that all scientific results are common property of the scientific community, an ideal we are far from today in the 21st century. Perhaps the solution to the patenting of microbiome microbes is dependent upon the idea that all human isolated organisms are common property, and that no one should have rights to them, so that we can promote individual rights, encourage scientific efforts in both academic and industrial fronts, and maintain profits of corporate entities without sacrificing the opportunities of others. However, how we utilize political and legislative action to ensure this scenario becomes a way of life amongst scientific communities remains to be seen.

227

Author Contributions:

Laura S. Weyrich1,2,3

1Biochemistry, Microbiology, and Molecular Biology Graduate Program

2Veterinary and Biomedical Science Department, The Pennsylvania State University

3Science, Technology, and Society Department, The Pennsylvania State University

228

References

1. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, et al. (2007) The Human Microbiome Project. Nature 449: 804-810. doi:10.1038/nature06244

2. Hawkins AK, O’Doherty KC (2011) “Who owns your poop?”: insights regarding the intersection of human microbiome research and the ELSI aspects of biobanking and related studies. BMC Med Genomics 4: 72. doi:10.1186/1755-8794-4-72

3. McGuire AL, Colgrove J, Whitney SN, Diaz CM, Bustillos D, et al. (2008) Ethical, legal, and social considerations in conducting the Human Microbiome Project. Genome Research 18: 1861 -1864. doi:10.1101/gr.081653.108

4. Zimmer C (2011) Our Microbiomes, Ourselves. The New York Times. Available: http://www.nytimes.com/2011/12/04/opinion/sunday/our-microbiomes-ourselves.html. Accessed 1 Feb 2012.

5. Tims S, Wamel W, Endtz HP, Belkum A, Kayser M (2009) Microbial DNA fingerprinting of human fingerprints: dynamic colonization of fingertip microflora challenges human host inferences for forensic purposes. International Journal of Legal Medicine 124: 477-481. doi:10.1007/s00414-009-0352-9

6. Costello EK, Lauber CL, Hamady M, Fierer N, Gordon JI, et al. (2009) Bacterial Community Variation in Human Body Habitats Across Space and Time. Science 326: 1694 -1697. doi:10.1126/science.1177486

7. Caporaso JG (2011) Moving pictures of the human microbiome. Genome Biology 12. doi:10.1186/gb-2011-12-5-r50

8. Skloot R (2010) The Immortal Life of Henrietta Lacks. First Edition. Crown. p.

9. Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, et al. (2005) Obesity alters gut microbial ecology. Proceedings of the National Academy of Sciences of the United States of America 102: 11070 -11075. doi:10.1073/pnas.0504978102

10. Blaser MJ, Falkow S (2009) What are the consequences of the disappearing human microbiota? Nat Rev Micro 7: 887-894. doi:10.1038/nrmicro2245

11. Cerf-Bensussan N, Gaboriau-Routhiau V (2010) The immune system and the gut microbiota: friends or foes? Nat Rev Immunol 10: 735-744. doi:10.1038/nri2850

12. Shultz LD, Ishikawa F, Greiner DL (2007) Humanized mice in translational biomedical research. Nat Rev Immunol 7: 118-130. doi:10.1038/nri2017

229 13. Chow J, Mazmanian SK (2010) A pathobiont of the microbiota balances host colonization and intestinal inflammation. Cell Host Microbe 7: 265-276. doi:10.1016/j.chom.2010.03.004

14. Guarner F, Malagelada J-R (2003) Gut flora in health and disease. The Lancet 361: 512- 519. doi:16/S0140-6736(03)12489-0

15. Tlaskalová-Hogenová H, Štěpánková R, Hudcovic T, Tučková L, Cukrowska B, et al. (2004) Commensal bacteria (normal microflora), mucosal immunity and chronic inflammatory and autoimmune diseases. Immunology Letters 93: 97-108. doi:10.1016/j.imlet.2004.02.005

16. Holden C (1978) Court Rules GE May Patent New Microorganism. Science 199: 1184- 1185. doi:10.1126/science.199.4334.1184-a

17. Krueger KG (1981) Building a better bacterium: genetic engineering and the patent law after Diamond v. Chakrabarty. Columbia Law Rev 81: 159-178.

18. Fialho AM, Chakrabarty AM (2007) Recent patents on bacterial proteins as potential anticancer agents. Recent Pat Anticancer Drug Discov 2: 224-234.

19. Pollack A (2011) Gene Patent in Cancer Test Upheld by Appeals Panel. The New York Times. Available: http://www.nytimes.com/2011/07/30/business/gene-patent-in-cancer- test-upheld-by-appeals-panel.html. Accessed 2 Feb 2012.

20. Jerry Grant, Arizona Man With C. Diff, Finds Relief After Fecal Transplant (2011) Huffington Post. Available: http://www.huffingtonpost.com/2011/11/01/jerry-grant-fecal- transplant-c-diff_n_1069401.html. Accessed 2 Feb 2012.

21. Holman CM (2007) Impact of Human Gene Patents on Innovation and Access: A Survey of Human Gene Patent Litigation, The. UMKC L. Rev. 76: 295.

22. Laura B (2001) Raising the bar for gene patents. Current Biology 11: R115-R116. doi:10.1016/S0960-9822(01)00054-9

23. Affairs) O (Office of P (n.d.) Prepared Remarks for David J. Kappos.. Available: http://www.uspto.gov/news/speeches/2010/Kappos_Bayh_Dole_Act_Remarks.jsp. Accessed 2 Feb 2012.

24. The 40-year war on cancer - CNN.com (n.d.) CNN. Available: http://www.cnn.com/2011/12/23/health/forty-year-war/index.html. Accessed 2 Feb 2012.

230

Chapter 10 : Summary and Significance

231

Overall Summary and Significance

The initial goal of this work was to understand how a common respiratory pathogen

interacts with the host microbiome. A major virulence factor associated with bacterial

competition in Pseudomonas, Vibrio, and Serratia is the Type VI Secretion System. We identified this system in Bordetella bronchiseptica and then investigated how it contributed to the ability of B. bronchiseptica to cause disease, identifying roles in immunomodulation, persistence, and antigen presentation manipulation. We additionally identified a sRNA that regulated a novel T6SS effector. Once that work was completed, we were able to investigate how B. bronchiseptica might utilize the T6SS to displace upper respiratory tract in mice, and in the process, uncovered a novel mechanism in which cytotoxic T-cells produce anti-microbials to assist an invading pathogen. This led to investigating how a closely related human pathogen,

Bordetella pertussis, interacted with microflora, where we observed that murine microflora are capable of inhibiting B. pertussis colonization, identifying a potential new disease prophylactic.

Next, we investigated pathogens associated with Chronic Obstructive Pulmonary Disease

(COPD) and asthma, and even though we were unable to identify Bordetella species in any human patient, we did detect novel bacterial species associated with disease and healthy patients.

Lastly, the ethical implications of using host isolated microorganisms as patentable entities was investigated, along with current issues associated with training biomedical researchers.

Together, this dissertation not only identifies several functions of a novel bordetellae virulence factor, elucidates novel interactions between respiratory pathogens and nasal cavity flora, and begins to identify the microbiome of the human respiratory tract, but it goes beyond biomedical research to investigate the ethical implications and questions presented by it.

232 Type VI Secretion System: A Novel Virulence Determinant in Bordetella

Although the T6SS had been implicated in virulence in numerous other systems, whether

or not this secretion system impacted the ability of Bordetella to cause disease had not been

investigated. By asking the simple question of whether or not the T6SS was a virulence factor in

B. bronchiseptica, we uncovered four major, novel mechanisms behind how the T6SS

contributes to pathology, immune modulation, persistence, and competition with host microflora.

Furthermore, we identified a small RNA that putatively regulates a T6SS effector.

Pathology. We determined that the T6SS is required by B. bronchiseptica to kill immune

phagocytic cells, such as macrophages and dendritic cells (Chapter 1 and 2). The ability to cause

death of epithelial cells, T-cells, or B-cells has yet to be investigated, although we hypothesize it

probably has deleterious effects on each of these cell types. To better understand the mechanism

behind how the T6SS causes cell death, we analyzed the proteome of murine macrophages

infected with B. bronchiseptica or a T6SS deficient mutant (Chapter 1). Proteins associated with intracellular bacterial survival (pyruvate kinase isozymes M1/M2) [1], eukaryotic intracellular signaling (voltage dependent anion selection channel) [2], structural mimicry (guanine nucleotide binding protein subunit beta 2) [3], and apoptosis (voltage dependent anion selection channel, transcription factor E2F7 and isocitrate dehydrogenase NADP fragment) were all identified in this experiment, suggesting that the T6SS is capable of inducing a programmed cell death mechanism [4–6]. However, the secreted proteins associated with these observations have yet to be identified. Our lab is currently using bioinformatic prediction, protein over-expression,

mass spectrometry, and western blotting to identify individual proteins responsible for these

phenotypes.

233 Immune modulation. We hypothesized that the ability of the T6SS to cause cell death may contribute to overall cellular damage and differential immune response, so we investigated pathology, cytokine production, and overall colonization. We determined that the T6SS contributes to overall cellular damage in the lower respiratory tract, causing alveolar cuffing, decreased cell recruitment, and swelling (Chapter 1). To investigate the mechanisms behind this pathogen beyond direct cell killing, we monitored the cytokines released in macrophages and at the site of infection. The T6SS stimulated IL-1β, IL-6, and IL-17 production in vitro and IL-6 and IL-17 in vivo suggesting that the T6SS contributes to a Th17, which was not shown to be required for the clearance of B. bronchiseptica in our laboratory (O. Y. R., unpublished data).

This suggests that the T6SS is capable of shunting the immune response in a way that is inefficient at clearing B. bronchiseptica, allowing this pathogen to persist longer during infection. Why IL-1β production was not observed in vivo is still puzzling, although new research in our laboratory suggests that the T6SS is required to cause lethal infection in IL-1R-/- mice within 3 days post inoculation (A. Karanikas, unpublished data), suggesting that IL-1β may be essential very early during B. bronchiseptica infection. Additionally, the premature form of

IL-1β is cleaved by caspase-1, which is also involved in programmed cell death in macrophages, a mechanism that we hypothesize is induced by the T6SS [7]. Thus, we can hypothesize that the

T6SS-mediate induction of IL-1β in vivo may be affected by the other cell death mechanisms happening concurrently in vivo. Understanding additional interactions with the T6SS and the immune system will undoubtedly lead to an increased understanding of how pathogens escape clearance and immune attack.

Antibody production modulation. Lastly, we observed that the T6SS was required for persistence in vivo. Although we could associate this with the ability of the T6SS to stimulate an

234 unproductive Th17 response, we additionally wanted to determine if antibody production was

manipulated. Our laboratory previously showed that antibodies are required for the clearance of all three classical bordetellae in vivo [8–11]. Perhaps not surprisingly, we observed that a T6SS

deficient B. bronchiseptica mutant that was unable to persist in the lower respiratory tract

induced much higher antibody titers in vivo, suggesting that the T6SS is required to dampen

antibody production (Chapter 2). Because we observed the T6SS to affect both T cell-

independent and dependent antibody production and because we already knew there were

interactions between this system and antigen presenting cells, we chose to monitor antigen

presentation modulation as a mechanism to reduce antibody production. For example, in

Salmonella, effectors of the T3SS have been shown to dampen peptide antigen presentation to T-

cells and B-cells, which has downstream repercussions on the ability of these adaptive cells to

produce antibodies [12,13]. Ultimately, we determined that the T6SS was capable of

dampening peptide antigen presentation (MHC-II) and lipid antigen presentation (CD1d) in vitro.

In vivo, the T6SS-mediated effects on MHC-II surface expression were not detectable, but

modulation of CD1d was observed on macrophages, dendritic cells, and B-cells. We went

further to show that CD1d was required for the T6SS-dependent modulation of antibody

production, and that the T6SS also affected NK T-cell accumulation in the lungs, the cell that receives CD1d antigen presentation from macrophages and dendritic cells. Together, these results suggested that the B. bronchiseptica T6SS modulates CD1d surface expression on antigen presenting cells, decreasing their ability to communicate with and recruit adaptive immune cells, thereby manipulating antibody production. This result established that the T6SS has at least two mechanisms by which to persist in the murine respiratory tract, by inducing a Th17 immune response and by dampening antibody production.

235 Nasal cavity persistence. Nasal cavity colonization was additionally decreased in mice

infected with a T6SS mutant of B. bronchiseptica compared to wild type bacteria. Antibodies or

IL-17 are not known to contribute to nasal cavity persistence, so we hypothesized a novel

mechanism or interaction contributed to this observation. In the Mougous lab, the Pseudomonas

T6SS was shown to deliver antimicrobial proteins that kill other bacterial species; this has

additionally been confirmed in Vibrio and Serratia [14–17]. We hypothesized that this defect in

nasal cavity colonization may be due to the ability to compete with microflora in vivo, even

though the ability of the T6SS to interact with microflora in vivo has not yet been investigated in any system, let alone Bordetella. Work described in Chapters 3 and 4 elucidate the contributions of a T6SS to interspecies and immune mediated competition in vivo.

Regulation. Lastly, we investigated regulation of the T6SS. We determined by

quantitative reverse transcriptase PCR that the T6SS was not BvgAS regulated, making it the

first non-BvgAS regulated factor that significantly contributes to pathology, persistence, and immune modulation in Bordetella (L. S. Weyrich, unpublished data). This single observation

suggests that there is much to learn about how bordetellae cause disease, and ultimately, how

disease modifiers are regulated in Bordetellae. Our laboratory has since begun investigating

environmental factors that alter transcription, alternative sigma factors that contribute to

virulence, and small RNA molecules as novel regulators of virulence in bordetellae. Using next

generation RNA sequencing, we were able to identify 181 new sRNA molecules in B.

bronchiseptica strain RB50 and established that these were conserved amongst B. pertussis and

B. parapertussis (Chapter 6). Using target prediction software, we determined that large protein

systems, such as the T3SS and flagellum in B. bronchiseptica, are the primary messenger RNA

targets of these molecules. Interestingly, two sRNAs, bbsrna28 and bbsrna70 were identified to

236 target a putative T6SS effector. The gene encoding this putative effector, tssJ, was previously observed to be up-regulated in response to serum, which correlated with an increased cell death by B. bronchiseptica (L. S. Weyrich and S. E. Hester, unpublished data). These results suggest that alternative regulatory systems and environmental factors regulate the T6SS during infection and additionally highlight how little is known about virulence factor regulation in the classical bordetellae. Furthermore, manipulating or utilizing sRNAs as novel antibacterial molecules against Bordetella infections is a novel, burgeoning field.

Novel Competition Mechanisms between Respiratory Pathogens and Host Microflora

Most infectious disease research is done only in the context of the pathogen and host.

Host microflora are commonly overlooked, as most researchers assumed microflora are negligible when pathogens are administered in very large doses, most likely overwhelming the microflora. However, infections do not usually occur in such a manner in nature, suggesting we understand very little about mechanisms that pathogens use to initially colonize a host. To investigate these questions, we created a new model system in which low doses and low volumes are administered to investigate how interactions with microflora contribute to the ability of a pathogen to cause disease, colonize a host, and stimulate the immune system. In this system, we identified that the broad host range, zoonotic pathogen B. bronchiseptica, can displace microflora, while the host-restricted, human pathogen B. pertussis, is inhibited by host organisms.

T3SS, T6SS, and BvgAS mediate microflora displacement. A pathogen must first overcome microflora to establish itself and cause disease. By infecting mice with 100 CFU of B. bronchiseptica in 10 µL, we were able to determine that this pathogen can displace culturable

237 microflora from the upper respiratory tract in mice as mechanism for colonization. Because

immune-mediated responses can generally be slower than direct bacterial competition, we

investigated the innate ability of B. bronchiseptica to out-compete individual microbes. Using bacterial mutants deficient in a wide range of virulence factors, we identified that the T3SS,

T6SS, and BvgAS are required to out-compete different microflora species. While we had earlier hypothesized that the T6SS would be required to displace nasal cavity flora, the T3SS and

BvgAS systems were unexpected. Thus, we hypothesize that the T3SS can additionally secrete antimicrobial factors that are potentially BvgAS regulated. This raises interesting implications regarding how bordetellae have evolved to co-regulate virulence factors that target the host and the microflora within the host. Alternatively, we could hypothesize that the T6SS and T3SS are linked, potentially by regulation, in a manner that when one is disrupted, so is the other.

Regulation of the T3SS and T6SS has been shown to be linked in other bacteria, suggesting this may be the most plausible explanation [18–20]. However, the B. bronchiseptica factors that either of these systems secrete remain unknown. A homolog of the T6SS-associated antimicrobial homolog identified by Mougous et al. in Pseudomonas does not exist in B. bronchiseptica (J. Park, unpublished data), suggesting that B. bronchiseptica has novel antimicrobial proteins that can be used in disease treatments against Klebsiella or Staphylococcus

species [21]. We also hypothesize that B. bronchiseptica has numerous additional novel

competition mechanisms, such as the use of bacteriophage, to deter other unculturable

competitor species that it might encounter in vivo or in the environment [22].

Immune modulation to displace flora. Although we identified a mechanism required for

microbe-to-microbe competition in vitro, these conclusions were based on the fact that host

microorganisms survived better when incubated with mutants compared to wild type. Several

238 host microflora strains were able to at least persist in the presence of the wild type strain in vitro,

suggesting additional in vivo mechanisms may assist B. bronchiseptica in displacing flora. To

test this, numerous immunodeficient mice were used to determine that T-cells and two T-cell associated cytokines, IFN-γ and IL-17, were required for B. bronchiseptica to fully displace flora

in vivo (Chapter 4). By monitoring which cell types became activated in the nasal cavity, we

observed that CD8+ single positive and CD4+CD8+ double positive T-cells had increased in number in the first 24 hours of infection. Double positive T-cells are a relatively newly discovered T-cell subset that was originally believed not to exist because of how T-cells were thought to mature in the thymus [23]. CD4+CD8+ double positive cells were found to be tissue

resident, memory T-cells with cytotoxic like functions, similar to CD8+ cells [24,25]. Because

both double positive cells and single positive CD8+ T-cells have shared functions, we focused on their release of anti-microbial proteins, such as granulysin and granzyme [26]. However, granulysin has not been identified in mice, suggesting that release of granzyme was the most likely candidate protein to assist B. bronchiseptica in displacing flora [27–29]. Host microflora isolated during our earlier studies were additionally found to be sensitive to granzyme B while B. bronchiseptica was found to be resistant, suggesting a novel mechanism behind how respiratory pathogens may be able to manipulate the immune system to outcompete flora. B. bronchiseptica enters the upper respiratory tract and immediately competes with flora using its T6SS, T3SS, and

BvgAS regulated factors. B. bronchiseptica then recruits and stimulates cytotoxic memory T-

cells, potentially by the same virulence factors that are directly competing with flora, to release

granzyme B, which in turn kills host microflora while leaving B. bronchiseptica unharmed.

Several facets of this mechanism remain unknown. First, the effector or antimicrobial

proteins secreted by either the T3SS or T6SS or an unidentified system regulated by BvgAS are

239 not known. Future work investigating the secretome and transcriptomic changes in response to

host microflora may elucidate some of these proteins. Next, the factor that stimulates cytotoxic

T-cells and ultimately the mechanism behind how B. bronchiseptica stimulates these cells remains unknown. Experiments stimulating cell cultured T-cells and identifying bacterial proteins associated with the cells could identify factors that bind to T-cells. Because of the speed in which these T-cells were stimulated, we hypothesize that B. bronchiseptica possesses a super- antigen, similar to ones identified in Staphylococcus aureus, which directly stimulates T-cells perhaps by binding MHC-I or CD1d [30,31]. Lastly, the factor that prevents granzyme- susceptibility in B. bronchiseptica remains undetermined. Future experiments in our laboratory hope to indentify this protein by determining which endogenous B. bronchiseptica proteins can bind to granzyme B in vitro, which can then be identified by mass-spectrometry.

B. pertussis colonization is inhibited by murine microflora. When B. pertussis was

administered in low doses and at low volumes, no colonization was observed. B. pertussis was

only able to colonize if a broad spectrum antibiotic, such as enrofloxacin, was delivered

beforehand (Chapter 5). This suggested that murine nasal cavity flora were inhibiting low dose,

low volume B. pertussis colonization. Because this observation could have dramatic public

health ramifications, we investigated how each isolated murine nasal cavity flora might out-

compete B. pertussis in vitro. By co-culturing B. pertussis with microflora organisms, we

determined that Staphylococcus species, common to both mice and humans, were able to prevent

B. pertussis growth in vitro, while a Klebsiella species was capable of killing B. pertussis. If

these microbes were capable of blocking B. pertussis growth in vitro, then we hypothesized that

they would also inhibit B. pertussis colonization in vivo. To test the potential of these murine

organisms to be utilized as probiotics to prevent of B. pertussis, we treated mice with antibiotics

240 and then infected them with murine microflora isolates. Both Staphylococcus species and

Klebsiella were capable of blocking or significantly limiting B. pertussis in the mouse model.

This observation suggests that microorganisms currently isolated from healthy animals or

humans may provide a novel solution to blocking B. pertussis colonization, although the

mechanism behind how these host organisms out-compete B. pertussis will still need to be

elucidated.

Evolution of competition microbes in the classical Bordetella. Unlike the ability B. bronchiseptica possesses to colonize mice at 5 CFU and displace flora, B. pertussis was unable to colonize mice when administered at low doses. Interestingly, B. parapertussis was able to colonize the murine nasal cavity at low doses but was unable to displace the flora. This raises interesting questions regarding pathogen evolution and how they have adapted to overcome microbes present in their niche environments. B. pertussis and B. parapertussis are believed to have evolved from a B. bronchiseptica-like projenitor species. B. pertussis lost 35% of its genome compared to B. bronchiseptica, while B. parapertussis appears to have lost only 20% (J.

Park, Submitted to PLoS Pathogens). Nonetheless, these pathogens are still highly related, and several groups have argued they be classified as subspecies [32]. These observations suggest that the genes B. pertussis has lost over time, probably due to host adaptation, may also have limited its ability to effectively out-compete murine flora. Maintaining genes necessary for murine competition may not have been essential as B. pertussis evolved as a host-restricted human pathogen, indicating a particular metabolic cost for maintaining genes required for interspecies microbial competition. B. parapertussis has also become more host restricted, but an individual clade is found to infect sheep, suggesting that B. parapertussis may have needed to maintain genes essential to microbial competition to remain successful in hosts beyond humans.

241 Interestingly, B. pertussis has lost the locus encoding a T6SS, and B. parapertussis has several

pseudogenes within the T6SS encoding locus (J. Park, unpublished data), suggesting that

evolution of the T6SS locus in Bordetella may have contributed to these phenotypes. Even if B.

pertussis and B. parapertussis have lost T6SS functional and the ability to displace flora, they

still remain highly successful in humans, while B. bronchiseptica is rarely observed in humans.

This suggests that a different set of mechanisms is required to compete with human microflora

species or that these pathogens may be able to infect the host by bypassing the upper respiratory

tract and only having to compete with lower respiratory tract microflora that are starkly different

than nasal cavity flora. Furthermore, it may partially explain why B. bronchiseptica is capable of

infecting such a broad host range, if it can utilize both innate and immune-mediating

mechanisms to displace flora. Future studies investigating how B. pertussis and B. parapertussis

effectively compete or displace flora within their natural hosts will be essential to understanding

evolution in terms of microbial ecology and host restriction.

Microbial competition as novel disease treatments and preventions.

Penicillin was initially discovered when Alexander Flemming observed that Penicillim

notatum was producing a molecule that was capable of killing bacteria, thus leading to modern

day antibiotics. The interaction between microflora species can be used in a similar manner. If

the proteins secreted by the T6SS or T3SS that limit microflora growth can be identified, they

could be used as novel antibiotics. These T6SS-dependent factors investigated here were shown to be active against Staphylococcus species, suggesting these factors may be particularly relevant to finding novel antibiotics against super bugs, such as methycilin resistant Staphylococcus aureus (MRSA). Another potential novel antimicrobial strategy will be to determine the factor

242 that stimulates cytotoxic T-cells. Any pathogen susceptible to granzyme B could be targeted by

therapy where cytotoxic T-cells are stimulated. Recent work has shown that granzyme B is

released systemically in the host and does not have a cytolytic effect on cell types unless

administered with perforin from cytotoxic T-cells, suggesting it may be a highly effective

antimicrobial molecule to run surveillance for other pathogens [29,33–36]. Understanding how

and if other pathogens evolved resistance to granzyme B or granulysin and how to stimulate

cytotoxic T-cells to just release granzyme B without additional perforin release will also be

essential in understanding the effectiveness of these molecules as novel therapies.

Observations raised within these studies also suggest that microorganisms isolated from

various hosts may protect against invading pathogens by out-competing them; essentially, microbes found to protect hosts against pathogens could be used as respiratory tract probiotics.

For example, Staphylococcus or Klebsiella species could be administered in a nasal spray daily or in conjunction with antibiotic treatment to help prevent B. pertussis colonization. However, there are several areas of research that must happen before respiratory tract probiotics can be implemented. In order for these probiotic organisms to effectively establish themselves in the upper respiratory tract, we first had to use antibiotic treatment to displace other organisms. It is plausible that a B. bronchiseptica-like T-cell stimulation mechanism could be utilized in place of antibiotics or that organisms with the ability to colonize the upper respiratory tract without antibiotic treatment will be identified in the future. Secondly, changing the microbiome present to protect against one pathogen could perhaps render the host more susceptible to an alternative pathogen. More research to understand how microbial communities contribute to overall disease susceptibility by direct competition and immune modulation will definitely need to be completed before respiratory tract probiotics can be implemented.

243 The simple observation that antibiotic treatment lowered the ID50 of B. pertussis has major public health ramifications. This study suggests that antibiotic treatment in the clinical setting may predispose individuals to B. pertussis infection. Interestingly, any stress factor or bacterial colonization that changes the upper respiratory tract microbiome could potentially increase the chance of B. pertussis infection. This observation may additionally suggest another possibility, beyond the lack of immunity, to why newborn infants are susceptible to whooping cough infections; the microbiome in their upper respiratory tract may not yet be established, allowing for pathogens like B. pertussis to easily colonize and cause infection. Our laboratory is currently collecting upper respiratory tract samples from newborn infants and ones infected with

B. pertussis to identify differences in the bacterial communities in these children. Understanding how respiratory tract colonization after birth contributes to disease susceptibility will undoubtedly lead to novel therapies for childhood diseases. Furthermore, the observation that B. bronchiseptica was able to displace flora up to 70 days post-inoculation has additional ramifications for veterinary health. Currently, live attenuated administration of intra-nasal B. bronchiseptica as a vaccine against canine kennel cough is currently recommended as the best preventative treatment. This vaccine strategy may be successful because B. bronchiseptica is capable of displacing flora and potentially other B. bronchiseptica strains from the upper respiratory tract. Understanding how other B. bronchiseptica strains are capable of overcoming this intranasal vaccine may lead to understanding additional novel mechanisms behind interspecies bacterial competition.

244 Contributions to the Human Respiratory Microbiome

Most of the work presented here was completed utilizing a murine mouse model.

However, in Chapter 7, culture and metagenomic analysis of human lower respiratory tract organisms was examined in healthy individuals or patients suffering from asthma, COPD, or unidentified respiratory symptoms. Although additional sequencing is needed, our initial results suggest that lower respiratory tract flora differ significantly from what is currently known about upper respiratory tract inhabitants. While Staphylococcus species are commonly identified in the nasal cavity, these species were not detected in the lower respiratory tract of healthy individuals, suggesting that pathogens may need an entirely different set of microbial competition mechanisms to establish themselves in the lower respiratory tract. Furthermore, our study identifies that respiratory disease can be associated with a dysbiosis in the lungs. Bacteroidia bacteria were identified in a healthy individual, while more Bacilli, Pseudomonales, and

Enterobacteriales were associated with COPD and asthma patients. Studies investigating the microbiome in obesity and inflammatory bowel disorder patients have identified similar findings, but this is the first study examining similar interactions in the upper respiratory tract [37–40].

Additional sequencing will determine if this observation is true over a larger sampling size and will undoubtedly identify more novel organisms yet unidentified in the lower respiratory tract.

Further work will also need to determine if these microbes contribute to the onset of disease or whether or not diseased lungs simply provide a niche environment for these microbes.

Ethical Implications of Biomedical and Microbiome Research

The current biomedical research field has grown exponentially over the past thirty years, developing copious new medical tools and strategies to combat disease. With this expansion,

245 increased pressure on researchers to produce bigger and better solution has occurred. Although the NIH and NSF recently recognized ethical training as an essential part of scientific training, proper bioethics training significantly lags behind research progress. Ethics is the foundation of quality biomedical research, and without bioethics training, infractions, dubious exploits, falsifications, and human harm are easily invited into scientific research. In Chapter 8, we aimed to provide an argument for teaching ethical methodologies and theories to young researchers early in their training to ensure that scientific progress proceeds in an ethical and beneficial manner. Teaching the theories behind the ‘check-list’ ethics researchers are currently taught will give them the freedom to make ethically sound decisions independently.

Chapter 9 was used as an example of how scientists can analyze ethical problems within their field of research. Currently, many of the ethical debates include historians, lawyers, or even a few ethicists, but rarely, scientists, which should be a pivotal portion of scientific policy and regulatory debates. The research in this dissertation suggested that microbes isolated from mice can be used as a prophylactic treatment to prevent whooping cough. The current human microbiome project is isolating thousands of microbial species in that study, implicating that many of these strains could be used similarly to how murine species are being utilized in this study. This raises immediate questions about microbial ownership, self autonomy, property rights, patenting law, and corporate ethics. Decisions made regarding the microbe ownership could restrict the use of organisms in academic research. Adding scientific theory into ethical debates can strengthen arguments and provide a fresh look from the front lines of research. In order for science to advance, scientists, politicians, and ethicists must harmoniously communicate and work together to ensure ethically robust progress.

246

References

1. Williams JM, Chen G, Zhu L, Rest RF (1998) Using the yeast two‐hybrid system to identify human epithelial cell proteins that bind gonococcal Opa proteins: intracellular gonococci bind pyruvate kinase via their Opa proteins and require host pyruvate for growth. Molecular Microbiology 27: 171-186. doi:10.1046/j.1365-2958.1998.00670.x

2. Shoshan-Barmatz V, Israelson A, Brdiczka D, Sheu SS (2006) The Voltage-Dependent Anion Channel (VDAC): Function in Intracellular Signalling, Cell Life and Cell Death. Current Pharmaceutical Design 12: 2249-2270. doi:10.2174/138161206777585111

3. Stebbins CE, Galán JE (2001) Structural mimicry in bacterial virulence. Nature 412: 701- 705. doi:10.1038/35089000

4. Li J, Ran C, Li E, Gordon F, Comstock G, et al. (2008) Synergistic Function of E2F7 and E2F8 Is Essential for Cell Survival and Embryonic Development. Developmental Cell 14: 62-75. doi:10.1016/j.devcel.2007.10.017

5. Kim SY, Lee SM, Tak JK, Choi KS, Kwon TK, et al. (2007) Regulation of singlet oxygen- induced apoptosis by cytosolic NADP+-dependent isocitrate dehydrogenase. Molecular and Cellular Biochemistry 302: 27-34. doi:10.1007/s11010-007-9421-x

6. Lang F, Lang PA, Lang KS, Brand V, Tanneur V, et al. (2004) Channel-induced apoptosis of infected host cells?the case of malaria. Pfl�gers Archiv European Journal of Physiology 448: 319-324. doi:10.1007/s00424-004-1254-9

7. Dinarello C (1996) Biologic basis for interleukin-1 in disease. Blood 87: 2095 -2147.

8. Kirimanjeswara GS, Mann PB, Harvill ET (2003) Role of antibodies in immunity to Bordetella infections. Infect Immun 71: 1719-24.

9. Kirimanjeswara GS, Mann PB, Pilione M, Kennett MJ, Harvill ET (2005) The complex mechanism of antibody-mediated clearance of Bordetella from the lungs requires TLR4. J Immunol 175: 7504-11.

10. Wolfe DN, Kirimanjeswara GS, Harvill ET (2005) Clearance of Bordetella parapertussis from the lower respiratory tract requires humoral and cellular immunity. Infect Immun 73: 6508-13. doi:73/10/6508 [pii] 10.1128/IAI.73.10.6508-6513.2005

11. Kirimanjeswara GS, Agosto LM, Kennett MJ, Bjornstad ON, Harvill ET (2005) Pertussis toxin inhibits neutrophil recruitment to delay antibody-mediated clearance of Bordetella pertussis. J Clin Invest 115: 3594-601.

247 12. Lapaque N, Hutchinson JL, Jones DC, Méresse S, Holden DW, et al. (2009) Salmonella regulates polyubiquitination and surface expression of MHC class II antigens. Proc. Natl. Acad. Sci. U.S.A. 106: 14052-14057. doi:10.1073/pnas.0906735106

13. Kirveskari, He, Leirisalo‐Repo, M äki ‐ Ikola, Wuorela, et al. (1999) Enterobacterial infection modulates major histocompatibility complex class I expression on mononuclear cells. Immunology 97: 420-428. doi:10.1046/j.1365-2567.1999.00803.x

14. Hood RD, Singh P, Hsu F, Güvener T, Carl MA, et al. (2010) A type VI secretion system of Pseudomonas aeruginosa targets a toxin to bacteria. Cell Host Microbe 7: 25-37. doi:10.1016/j.chom.2009.12.007

15. Jani AJ, Cotter PA (2010) Type VI secretion: not just for pathogenesis anymore. Cell Host Microbe 8: 2-6. doi:10.1016/j.chom.2010.06.012

16. MacIntyre DL, Miyata ST, Kitaoka M, Pukatzki S (2010) The Vibrio cholerae type VI secretion system displays antimicrobial properties. Proceedings of the National Academy of Sciences 107: 19520 -19524. doi:10.1073/pnas.1012931107

17. Murdoch SL, Trunk K, English G, Fritsch MJ, Pourkarimi E, et al. (2011) The Opportunistic Pathogen Serratia marcescens Utilizes Type VI Secretion To Target Bacterial Competitors. J. Bacteriol. 193: 6057-6069. doi:10.1128/JB.05671-11

18. Rao PSS, Yamada Y, Tan YP, Leung KY (2004) Use of proteomics to identify novel virulence determinants that are required for Edwardsiella tarda pathogenesis. Molecular Microbiology 53: 573-586. doi:10.1111/j.1365-2958.2004.04123.x

19. Cascales E (2008) The type VI secretion toolkit. EMBO Rep 9: 735-41. doi:embor2008131 [pii] 10.1038/embor.2008.131

20. Parsons DA, Heffron F (2005) sciS, an icmF Homolog in Salmonella enterica Serovar Typhimurium, Limits Intracellular Replication and Decreases Virulence. Infect. Immun. 73: 4338-4345.

21. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, et al. (2011) Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475: 343-347. doi:10.1038/nature10244

22. Joo J, Gunny M, Cases M, Hudson P, Albert R, et al. (2006) Bacteriophage-mediated competition in Bordetella bacteria. Proc Biol Sci 273: 1843-8.

23. Federico A Z (1999) Extrathymic CD4/CD8 double positive T cells. Veterinary Immunology and Immunopathology 72: 55-66. doi:10.1016/S0165-2427(99)00118-X

24. Nascimbeni M, Pol S, Saunier B (2011) Distinct CD4+CD8+ Double-Positive T Cells in the Blood and Liver of Patients during Chronic Hepatitis B and C. PLoS ONE 6: e20145. doi:10.1371/journal.pone.0020145

248 25. Nascimbeni M, Shin E-C, Chiriboga L, Kleiner DE, Rehermann B (2004) Peripheral CD4+CD8+ T cells are differentiated effector memory cells with antiviral functions. Blood 104: 478 -486. doi:10.1182/blood-2003-12-4395

26. Harty JT, Tvinnereim AR, White DW (2000) CD8+ T Cell Effector Mechanisms in Resistance to Infection. Annual Review of Immunology 18: 275-308. doi:10.1146/annurev.immunol.18.1.275

27. Stenger S, Hanson DA, Teitelbaum R, Dewan P, Niazi KR, et al. (1998) An antimicrobial activity of cytolytic T cells mediated by granulysin. Science 282: 121-125.

28. Krensky AM, Clayberger C (2009) Biology and clinical relevance of granulysin. Tissue Antigens 73: 193-198. doi:10.1111/j.1399-0039.2008.01218.x

29. Thiery J, Walch M, Jensen DK, Martinvalet D, Lieberman J (n.d.) Isolation of Cytotoxic T Cell and NK Granules and Purification of Their Effector Proteins.. Available: http://onlinelibrary.wiley.com/doi/10.1002/0471143030.cb0337s47/abstract. Accessed 13 Jan 2012.

30. Krakauer T (1999) Immune response to staphylococcal superantigens. Immunol. Res. 20: 163-173.

31. Ulrich RG (2000) Evolving superantigens of Staphylococcus aureus. FEMS Immunol. Med. Microbiol. 27: 1-7.

32. Parkhill J, Sebaihia M, Preston A, Murphy LD, Thomson N, et al. (2003) Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35: 32-40. doi:10.1038/ng1227 ng1227 [pii]

33. Browne KA, Blink E, Sutton VR, Froelich CJ, Jans DA, et al. (1999) Cytosolic Delivery of Granzyme B by Bacterial Toxins: Evidence that Endosomal Disruption, in Addition to Transmembrane Pore Formation, Is an Important Function of Perforin. Molecular and Cellular Biology 19: 8604 -8615.

34. Lauw FN, Simpson AJ, Hack CE, Prins JM, Wolbink AM, et al. (2000) Soluble granzymes are released during human endotoxemia and in patients with severe infection due to gram- negative bacteria. J. Infect. Dis. 182: 206-213. doi:10.1086/315642

35. Chun-yan G, Bo H, Hong C, Hong-lei J, Xiu-zhen H (2009) Anti-perforin neutralizing antibody reduces myocardial injury in viral myocarditis. Cardiol Young 19: 601-607. doi:10.1017/S104795110999182X

36. Shafer W, Pohl J, Onunka V, Bangalore N, Travis J (1991) Human lysosomal cathepsin G and granzyme B share a functionally conserved broad spectrum antibacterial peptide. J. Biol. Chem. 266: 112-116.

249 37. Blaser MJ, Falkow S (2009) What are the consequences of the disappearing human microbiota? Nat Rev Micro 7: 887-894. doi:10.1038/nrmicro2245

38. Cerf-Bensussan N, Gaboriau-Routhiau V (2010) The immune system and the gut microbiota: friends or foes? Nat Rev Immunol 10: 735-744. doi:10.1038/nri2850

39. Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, et al. (2005) Obesity alters gut microbial ecology. Proceedings of the National Academy of Sciences of the United States of America 102: 11070 -11075. doi:10.1073/pnas.0504978102

40. Chow J, Mazmanian SK (2010) A pathobiont of the microbiota balances host colonization and intestinal inflammation. Cell Host Microbe 7: 265-276. doi:10.1016/j.chom.2010.03.004

250

Appendix A: SIPHT Predicted Small RNAs

SIPHT sRNAName sRNAstart end sRNAtype BLASTscore

Candidate_1_NC_002927 282124 282316 candidate 210 Candidate_2_NC_002927 436163 436213 candidate 271 Candidate_3_NC_002927 462484 462671 candidate 1239 Candidate_4_NC_002927 482342 482512 candidate 117 Candidate_5_NC_002927 690614 690707 candidate 911 Candidate_6_NC_002927 701641 701767 candidate 1330 Candidate_7_NC_002927 1265355 1265452 candidate 332 Candidate_8_NC_002927 1413006 1413065 candidate 745 Candidate_9_NC_002927 1558070 1558363 candidate 339 Candidate_10_NC_002927 1587615 1587688 candidate 490 Candidate_11_NC_002927 1620675 1620746 candidate 390 Candidate_12_NC_002927 1669731 1669792 candidate 1185 Candidate_14_NC_002927 1806019 1806118 candidate 221 Candidate_15_NC_002927 1925801 1926307 candidate 2217 Candidate_17_NC_002927 2103468 2103499 candidate 1646 Candidate_18_NC_002927 2117294 2117353 candidate 542 Candidate_20_NC_002927 2248399 2248514 candidate 686 Candidate_21_NC_002927 2281083 2281146 candidate 885 Candidate_22_NC_002927 2293026 2293225 candidate 207 Candidate_23_NC_002927 2314182 2314361 candidate 682 Candidate_25_NC_002927 2613364 2613398 candidate 486 Candidate_26_NC_002927 3371111 3371186 candidate 222 Candidate_28_NC_002927 3921112 3921199 candidate 529 Candidate_31_NC_002927 4341981 4342020 candidate 892 Candidate_32_NC_002927 4494724 4494854 candidate 244 Candidate_33_NC_002927 4500362 4500407 candidate 605 Candidate_34_NC_002927 4536491 4536700 candidate 202 Candidate_35_NC_002927 4654104 4654186 candidate 471 Candidate_36_NC_002927 5272308 5272472 candidate 205 Candidate_37_NC_002927 5242459 5242556 candidate 316 Candidate_38_NC_002927 4752603 4752732 candidate 205 Candidate_39_NC_002927 4657456 4657567 candidate 227 Candidate_40_NC_002927 4640591 4640891 candidate 204 Candidate_41_NC_002927 4600135 4600249 candidate 1384 Candidate_42_NC_002927 4450455 4450604 candidate 886

251 Candidate_43_NC_002927 4401767 4401808 candidate 607 Candidate_44_NC_002927 4389511 4389555 candidate 1170 Candidate_45_NC_002927 4187098 4187152 candidate 285 Candidate_46_NC_002927 4161051 4161547 candidate 3819 Candidate_47_NC_002927 4114080 4114479 candidate 3206 Candidate_48_NC_002927 3720527 3720600 candidate 218 Candidate_49_NC_002927 3478379 3478599 candidate 287 Candidate_50_NC_002927 3468751 3468862 candidate 694 Candidate_51_NC_002927 3373657 3373803 candidate 447 Candidate_52_NC_002927 3370090 3370371 candidate 1499 Candidate_53_NC_002927 3361914 3361953 candidate 1536 Candidate_54_NC_002927 3351503 3351657 candidate 225 Candidate_55_NC_002927 2933375 2933407 candidate 721 Candidate_56_NC_002927 2832167 2832248 candidate 406 Candidate_58_NC_002927 2316500 2316594 candidate 511 Candidate_59_NC_002927 2253264 2253334 candidate 460 Candidate_60_NC_002927 2199220 2199276 candidate 552 Candidate_61_NC_002927 2147775 2147934 candidate 199 Candidate_62_NC_002927 2023266 2023330 candidate 311 Candidate_63_NC_002927 1983570 1983618 candidate 268 Candidate_64_NC_002927 1967882 1967958 candidate 1032 Candidate_65_NC_002927 1941567 1941637 candidate 360 Candidate_66_NC_002927 1626156 1626229 candidate 199 Candidate_67_NC_002927 1602837 1602909 candidate 246 Candidate_70_NC_002927 530673 530712 candidate 276 Candidate_71_NC_002927 167406 167461 candidate 216 Candidate_72_NC_002927 96881 96997 candidate 226 Candidate_76_NC_002927 74074 74213 candidate 1160 Candidate_77_NC_002927 278493 278604 candidate 208 Candidate_78_NC_002927 292795 292855 candidate 446 Candidate_79_NC_002927 504092 504182 candidate 269 Candidate_80_NC_002927 807182 807254 candidate 203 Candidate_81_NC_002927 874824 874925 candidate 663 Candidate_82_NC_002927 1022378 1022453 candidate 114 Candidate_83_NC_002927 1032259 1032376 candidate 107 Candidate_84_NC_002927 1032212 1032304 candidate 107 Candidate_85_NC_002927 1108273 1108387 candidate 211 Candidate_86_NC_002927 1170637 1170945 candidate 118 Candidate_87_NC_002927 1279158 1279257 candidate 136 Candidate_88_NC_002927 1425641 1425750 candidate 111 Candidate_89_NC_002927 1533876 1533963 candidate 277 Candidate_90_NC_002927 1815833 1815930 candidate 950

252 Candidate_91_NC_002927 1817602 1817713 candidate 214 Candidate_92_NC_002927 1843278 1843453 candidate 124 Candidate_93_NC_002927 1864279 1864418 candidate 95 Candidate_94_NC_002927 1944018 1944148 candidate 210 Candidate_95_NC_002927 2098668 2098780 candidate 825 Candidate_96_NC_002927 2118037 2118141 candidate 1551 Candidate_97_NC_002927 2141282 2141354 candidate 1000 Candidate_98_NC_002927 2281083 2281163 candidate 885 Candidate_99_NC_002927 2361763 2361824 candidate 1180 Candidate_100_NC_002927 2371536 2371735 candidate 1072 Candidate_102_NC_002927 2553580 2553753 candidate 91 Candidate_103_NC_002927 2553583 2553724 candidate 91 Candidate_104_NC_002927 2558978 2559118 candidate 1153 Candidate_105_NC_002927 2636280 2636323 candidate 217 Candidate_106_NC_002927 2656355 2656510 candidate 119 Candidate_107_NC_002927 2827315 2827382 candidate 890 Candidate_109_NC_002927 2830454 2830575 candidate 213 Candidate_110_NC_002927 2831077 2831157 candidate 500 Candidate_111_NC_002927 2909820 2909891 candidate 475 Candidate_112_NC_002927 3055240 3055321 candidate 530 Candidate_113_NC_002927 3082339 3082492 candidate 880 Candidate_114_NC_002927 3150574 3150783 candidate 106 Candidate_115_NC_002927 3220778 3220898 candidate 214 Candidate_116_NC_002927 3377346 3377431 candidate 565 Candidate_117_NC_002927 3396121 3396208 candidate 460 Candidate_118_NC_002927 3402314 3402451 candidate 825 Candidate_119_NC_002927 3424411 3424640 candidate 217 Candidate_120_NC_002927 3521864 3521977 candidate 206 Candidate_121_NC_002927 3631619 3631720 candidate 198 Candidate_122_NC_002927 3684229 3684408 candidate 99 Candidate_124_NC_002927 3737942 3738004 candidate 1180 Candidate_125_NC_002927 3805157 3805281 candidate 96 Candidate_126_NC_002927 3805199 3805281 candidate 96 Candidate_127_NC_002927 3987776 3987888 candidate 771 Candidate_128_NC_002927 3993179 3993251 candidate 366 Candidate_129_NC_002927 4019725 4019829 candidate 805 Candidate_130_NC_002927 4046998 4047131 candidate 103 Candidate_131_NC_002927 4181258 4181360 candidate 125 Candidate_132_NC_002927 4181194 4181299 candidate 125 Candidate_133_NC_002927 4194244 4194313 candidate 209 Candidate_134_NC_002927 4268210 4268353 candidate 206 Candidate_135_NC_002927 4290724 4290977 candidate 103

253 Candidate_136_NC_002927 4295154 4295311 candidate 115 Candidate_137_NC_002927 4295220 4295311 candidate 115 Candidate_138_NC_002927 4314884 4314965 candidate 137 Candidate_139_NC_002927 4354328 4354413 candidate 89 Candidate_141_NC_002927 4409009 4409169 candidate 288 Candidate_142_NC_002927 4432225 4432287 candidate 267 Candidate_143_NC_002927 4561458 4561618 candidate 98 Candidate_144_NC_002927 4561469 4561577 candidate 98 Candidate_145_NC_002927 4566471 4566647 candidate 226 Candidate_146_NC_002927 4661739 4662004 candidate 441 Candidate_147_NC_002927 4666193 4666348 candidate 203 Candidate_148_NC_002927 4763542 4763652 candidate 876 Candidate_149_NC_002927 4789074 4789228 candidate 81 Candidate_150_NC_002927 4807084 4807187 candidate 98 Candidate_151_NC_002927 4870336 4870414 candidate 370 Candidate_153_NC_002927 4889292 4889360 candidate 201 Candidate_154_NC_002927 4914364 4914404 candidate 225 Candidate_155_NC_002927 4949553 4949666 candidate 202 Candidate_156_NC_002927 5182692 5182775 candidate 290 Candidate_157_NC_002927 5210580 5210647 candidate 365 Candidate_158_NC_002927 5227795 5227859 candidate 405 Candidate_159_NC_002927 5246619 5246777 candidate 88 Candidate_160_NC_002927 1034976 1035018 candidate 440 Candidate_161_NC_002927 1039302 1039389 candidate 461 Candidate_162_NC_002927 10260 10482 candidate 202 Candidate_163_NC_002927 1051815 1051912 candidate 570 Candidate_164_NC_002927 1064513 1064624 candidate 672 Candidate_165_NC_002927 1071924 1072010 candidate 690 Candidate_168_NC_002927 1228562 1228681 candidate 386 Candidate_169_NC_002927 1233807 1234009 candidate 1255 Candidate_170_NC_002927 1243106 1243207 candidate 290 Candidate_171_NC_002927 1329752 1329962 candidate 374 Candidate_172_NC_002927 1339654 1339801 candidate 1566 Candidate_173_NC_002927 1359598 1359638 candidate 217 Candidate_174_NC_002927 1366296 1366407 candidate 276 Candidate_175_NC_002927 1371474 1371643 candidate 207 Candidate_176_NC_002927 1383148 1383214 candidate 214 Candidate_177_NC_002927 1423452 1423532 candidate 201 Candidate_178_NC_002927 1429494 1429727 candidate 204 Candidate_179_NC_002927 1444262 1444499 candidate 216 Candidate_180_NC_002927 1452573 1452718 candidate 199 Candidate_181_NC_002927 1458436 1458521 candidate 767

254 Candidate_182_NC_002927 1460768 1460823 candidate 223 Candidate_183_NC_002927 1460746 1460828 candidate 223 Candidate_184_NC_002927 1460851 1460904 candidate 231 Candidate_185_NC_002927 1460935 1460989 candidate 224 Candidate_186_NC_002927 1460913 1460995 candidate 223 Candidate_187_NC_002927 1496915 1497035 candidate 347 Candidate_188_NC_002927 1504253 1504432 candidate 203 Candidate_190_NC_002927 1505414 1505468 candidate 226 Candidate_191_NC_002927 1505500 1505546 candidate 200 Candidate_192_NC_002927 1527674 1527725 candidate 1006 Candidate_193_NC_002927 1557767 1558035 candidate 3043 Candidate_194_NC_002927 1558070 1558184 candidate 339 Candidate_195_NC_002927 1559145 1559183 candidate 211 Candidate_196_NC_002927 158235 158310 candidate 245 Candidate_197_NC_002927 1584768 1584850 candidate 213 Candidate_198_NC_002927 1589111 1589262 candidate 1333 Candidate_199_NC_002927 1640177 1640336 candidate 994 Candidate_200_NC_002927 1669595 1669645 candidate 202 Candidate_201_NC_002927 1755976 1756125 candidate 487 Candidate_202_NC_002927 1760501 1760572 candidate 382 Candidate_203_NC_002927 1761358 1761536 candidate 218 Candidate_204_NC_002927 1775474 1775628 candidate 208 Candidate_205_NC_002927 1782347 1782385 candidate 200 Candidate_206_NC_002927 1805881 1805960 candidate 2203 Candidate_207_NC_002927 1820225 1820380 candidate 108 Candidate_208_NC_002927 1828478 1828768 candidate 233 Candidate_209_NC_002927 1843278 1843367 candidate 124 Candidate_210_NC_002927 1857910 1858103 candidate 1013 Candidate_211_NC_002927 1858472 1858686 candidate 1051 Candidate_212_NC_002927 187238 187317 candidate 699 Candidate_213_NC_002927 189180 189281 candidate 200 Candidate_214_NC_002927 1898293 1898368 candidate 478 Candidate_215_NC_002927 1951092 1951200 candidate 1149 Candidate_217_NC_002927 1951805 1951943 candidate 1317 Candidate_218_NC_002927 1956411 1956525 candidate 1814 Candidate_219_NC_002927 1960679 1960758 candidate 675 Candidate_220_NC_002927 1962856 1963185 candidate 222 Candidate_221_NC_002927 1963097 1963185 candidate 222 Candidate_222_NC_002927 1967752 1967875 candidate 352 Candidate_223_NC_002927 20005 20090 candidate 513 Candidate_224_NC_002927 2025213 2025341 candidate 1031 Candidate_225_NC_002927 2030790 2030968 candidate 219

255 Candidate_226_NC_002927 2038840 2038939 candidate 580 Candidate_227_NC_002927 2050498 2050589 candidate 530 Candidate_228_NC_002927 2058059 2058090 candidate 634 Candidate_229_NC_002927 2058207 2058250 candidate 228 Candidate_230_NC_002927 2058138 2058241 candidate 208 Candidate_231_NC_002927 207805 207848 candidate 288 Candidate_232_NC_002927 2084913 2085037 candidate 213 Candidate_233_NC_002927 2084972 2085037 candidate 213 Candidate_234_NC_002927 2090121 2090200 candidate 467 Candidate_235_NC_002927 2100875 2100966 candidate 520 Candidate_236_NC_002927 2103571 2103712 candidate 204 Candidate_237_NC_002927 210263 210387 candidate 214 Candidate_238_NC_002927 2110470 2110537 candidate 456 Candidate_239_NC_002927 2114350 2114501 candidate 620 Candidate_241_NC_002927 2117889 2117959 candidate 540 Candidate_242_NC_002927 2128201 2128380 candidate 213 Candidate_243_NC_002927 2139718 2139779 candidate 258 Candidate_244_NC_002927 2152917 2152993 candidate 450 Candidate_245_NC_002927 2153514 2153606 candidate 470 Candidate_246_NC_002927 2167897 2168035 candidate 780 Candidate_247_NC_002927 2184321 2184478 candidate 379 Candidate_248_NC_002927 2184321 2184525 candidate 379 Candidate_249_NC_002927 2191725 2191794 candidate 851 Candidate_250_NC_002927 2193992 2194074 candidate 1148 Candidate_251_NC_002927 2196220 2196323 candidate 620 Candidate_252_NC_002927 2201329 2201409 candidate 355 Candidate_253_NC_002927 2202998 2203066 candidate 616 Candidate_254_NC_002927 2206705 2206916 candidate 657 Candidate_256_NC_002927 2216954 2217009 candidate 232 Candidate_257_NC_002927 2217031 2217080 candidate 221 Candidate_258_NC_002927 2230348 2230416 candidate 355 Candidate_259_NC_002927 2237604 2237659 candidate 215 Candidate_261_NC_002927 2237687 2237741 candidate 240 Candidate_262_NC_002927 223925 224156 candidate 946 Candidate_263_NC_002927 2277573 2277693 candidate 1229 Candidate_264_NC_002927 2285096 2285148 candidate 222 Candidate_265_NC_002927 2289609 2289724 candidate 226 Candidate_266_NC_002927 2298263 2298307 candidate 288 Candidate_267_NC_002927 2298145 2298293 candidate 288 Candidate_268_NC_002927 2304552 2304700 candidate 802 Candidate_269_NC_002927 2307696 2307747 candidate 249 Candidate_270_NC_002927 2343616 2343677 candidate 404

256 Candidate_271_NC_002927 2343504 2343679 candidate 404 Candidate_272_NC_002927 2352608 2352683 candidate 263 Candidate_273_NC_002927 2357311 2357392 candidate 440 Candidate_274_NC_002927 2376144 2376230 candidate 472 Candidate_275_NC_002927 2379435 2379501 candidate 415 Candidate_276_NC_002927 2380622 2380705 candidate 485 Candidate_277_NC_002927 2387527 2387623 candidate 532 Candidate_278_NC_002927 2387500 2387567 candidate 532 Candidate_279_NC_002927 2401274 2401327 candidate 235 Candidate_280_NC_002927 2401357 2401410 candidate 231 Candidate_281_NC_002927 2401440 2401493 candidate 231 Candidate_282_NC_002927 2406414 2406547 candidate 377 Candidate_283_NC_002927 2406413 2406443 candidate 377 Candidate_284_NC_002927 2412129 2412229 candidate 611 Candidate_285_NC_002927 2434497 2434596 candidate 218 Candidate_286_NC_002927 2434619 2434673 candidate 224 Candidate_287_NC_002927 2434702 2434847 candidate 224 Candidate_288_NC_002927 2434596 2434763 candidate 237 Candidate_289_NC_002927 2434786 2434847 candidate 224 Candidate_290_NC_002927 2434768 2434846 candidate 224 Candidate_291_NC_002927 2434869 2434930 candidate 224 Candidate_292_NC_002927 2434851 2434929 candidate 224 Candidate_293_NC_002927 2434952 2435013 candidate 224 Candidate_294_NC_002927 2434934 2435012 candidate 224 Candidate_295_NC_002927 2435035 2435089 candidate 225 Candidate_296_NC_002927 2435118 2435179 candidate 237 Candidate_297_NC_002927 2435202 2435247 candidate 233 Candidate_298_NC_002927 2445688 2445793 candidate 292 Candidate_299_NC_002927 2474318 2474425 candidate 206 Candidate_300_NC_002927 2478945 2479017 candidate 300 Candidate_301_NC_002927 2494914 2495105 candidate 1415 Candidate_302_NC_002927 2509755 2509825 candidate 258 Candidate_303_NC_002927 2545348 2545418 candidate 208 Candidate_304_NC_002927 255967 256035 candidate 332 Candidate_305_NC_002927 2566385 2566432 candidate 750 Candidate_306_NC_002927 257358 257462 candidate 494 Candidate_307_NC_002927 2636196 2636323 candidate 217 Candidate_308_NC_002927 2648793 2648836 candidate 275 Candidate_309_NC_002927 2648924 2648985 candidate 435 Candidate_310_NC_002927 2739273 2739377 candidate 197 Candidate_311_NC_002927 2752402 2752556 candidate 249 Candidate_312_NC_002927 2767777 2767823 candidate 279

257 Candidate_313_NC_002927 2775515 2775555 candidate 260 Candidate_314_NC_002927 2802046 2802262 candidate 124 Candidate_315_NC_002927 2826504 2826605 candidate 431 Candidate_316_NC_002927 2832167 2832236 candidate 406 Candidate_317_NC_002927 2920800 2920921 candidate 295 Candidate_318_NC_002927 292794 293099 candidate 446 Candidate_319_NC_002927 3016721 3016812 candidate 507 Candidate_320_NC_002927 3111879 3112110 candidate 1611 Candidate_321_NC_002927 3115537 3115646 candidate 685 Candidate_322_NC_002927 3117843 3117964 candidate 690 Candidate_323_NC_002927 3133040 3133153 candidate 615 Candidate_324_NC_002927 3154982 3155034 candidate 1220 Candidate_325_NC_002927 3158525 3158584 candidate 600 Candidate_326_NC_002927 3170524 3170622 candidate 751 Candidate_327_NC_002927 3174235 3174325 candidate 674 Candidate_328_NC_002927 3184577 3184701 candidate 300 Candidate_329_NC_002927 321225 321339 candidate 725 Candidate_330_NC_002927 3213390 3213572 candidate 901 Candidate_331_NC_002927 3217820 3217873 candidate 857 Candidate_332_NC_002927 3217820 3218035 candidate 857 Candidate_333_NC_002927 3256526 3256643 candidate 639 Candidate_334_NC_002927 3272765 3273035 candidate 1146 Candidate_335_NC_002927 3289849 3289885 candidate 518 Candidate_336_NC_002927 3300120 3300213 candidate 476 Candidate_337_NC_002927 331081 331180 candidate 1600 Candidate_338_NC_002927 3319465 3319766 candidate 273 Candidate_339_NC_002927 3323191 3323268 candidate 1677 Candidate_340_NC_002927 3328055 3328103 candidate 1170 Candidate_341_NC_002927 3370295 3370371 candidate 1499 Candidate_342_NC_002927 3373980 3374104 candidate 225 Candidate_343_NC_002927 3380013 3380088 candidate 362 Candidate_345_NC_002927 33909 33953 candidate 162 Candidate_346_NC_002927 3396972 3397039 candidate 606 Candidate_347_NC_002927 3402714 3402795 candidate 230 Candidate_348_NC_002927 3424371 3424594 candidate 217 Candidate_349_NC_002927 3434385 3434450 candidate 1218 Candidate_350_NC_002927 3469987 3470057 candidate 326 Candidate_351_NC_002927 3478165 3478234 candidate 180 Candidate_352_NC_002927 3478275 3478599 candidate 287 Candidate_353_NC_002927 3478456 3478592 candidate 237 Candidate_354_NC_002927 3485707 3485785 candidate 541 Candidate_356_NC_002927 3524063 3524135 candidate 1394

258 Candidate_357_NC_002927 3534223 3534255 candidate 264 Candidate_358_NC_002927 3542363 3542521 candidate 870 Candidate_359_NC_002927 3544144 3544242 candidate 545 Candidate_360_NC_002927 3564972 3565056 candidate 204 Candidate_361_NC_002927 3565087 3565140 candidate 231 Candidate_362_NC_002927 3565161 3565203 candidate 137 Candidate_363_NC_002927 3580072 3580219 candidate 1242 Candidate_364_NC_002927 3586873 3586999 candidate 205 Candidate_365_NC_002927 3595169 3595246 candidate 473 Candidate_366_NC_002927 3598868 3598926 candidate 878 Candidate_367_NC_002927 3616847 3617163 candidate 218 Candidate_368_NC_002927 3626157 3626226 candidate 415 Candidate_369_NC_002927 3633754 3633891 candidate 205 Candidate_370_NC_002927 3652558 3652656 candidate 119 Candidate_371_NC_002927 365304 365345 candidate 585 Candidate_372_NC_002927 3672907 3672964 candidate 370 Candidate_373_NC_002927 3686334 3686693 candidate 253 Candidate_374_NC_002927 3686399 3686693 candidate 253 Candidate_375_NC_002927 3686301 3686430 candidate 246 Candidate_376_NC_002927 3686465 3686693 candidate 253 Candidate_377_NC_002927 3686531 3686693 candidate 253 Candidate_378_NC_002927 3686597 3686693 candidate 253 Candidate_379_NC_002927 3686301 3686628 candidate 246 Candidate_380_NC_002927 3709234 3709325 candidate 865 Candidate_381_NC_002927 3709347 3709390 candidate 236 Candidate_382_NC_002927 3711534 3711615 candidate 252 Candidate_383_NC_002927 3711574 3711615 candidate 252 Candidate_384_NC_002927 3728030 3728231 candidate 964 Candidate_385_NC_002927 3742374 3742456 candidate 440 Candidate_386_NC_002927 3746945 3747022 candidate 263 Candidate_387_NC_002927 3763535 3763587 candidate 252 Candidate_388_NC_002927 3813581 3813664 candidate 1256 Candidate_389_NC_002927 3814635 3814710 candidate 707 Candidate_390_NC_002927 3818904 3819054 candidate 836 Candidate_391_NC_002927 3893781 3893879 candidate 774 Candidate_392_NC_002927 3903610 3903662 candidate 1305 Candidate_393_NC_002927 38994 39066 candidate 821 Candidate_394_NC_002927 3913722 3913754 candidate 224 Candidate_395_NC_002927 3913722 3913784 candidate 224 Candidate_396_NC_002927 3940422 3940490 candidate 515 Candidate_397_NC_002927 3947209 3947320 candidate 707 Candidate_398_NC_002927 3954065 3954140 candidate 493

259 Candidate_399_NC_002927 395628 395717 candidate 217 Candidate_400_NC_002927 3963929 3963971 candidate 287 Candidate_401_NC_002927 3995201 3995363 candidate 1018 Candidate_402_NC_002927 4009042 4009316 candidate 1254 Candidate_403_NC_002927 4025245 4025308 candidate 411 Candidate_404_NC_002927 4072539 4072631 candidate 270 Candidate_405_NC_002927 4072445 4072568 candidate 377 Candidate_406_NC_002927 4114388 4114479 candidate 3206 Candidate_407_NC_002927 4115941 4116067 candidate 705 Candidate_408_NC_002927 4122955 4123062 candidate 253 Candidate_409_NC_002927 4124613 4124685 candidate 826 Candidate_410_NC_002927 4125704 4125785 candidate 360 Candidate_411_NC_002927 4146147 4146259 candidate 1443 Candidate_412_NC_002927 4149881 4150046 candidate 916 Candidate_413_NC_002927 4160832 4160887 candidate 566 Candidate_414_NC_002927 4184510 4184601 candidate 441 Candidate_415_NC_002927 4199090 4199211 candidate 216 Candidate_416_NC_002927 4223118 4223204 candidate 300 Candidate_417_NC_002927 4234461 4234537 candidate 209 Candidate_418_NC_002927 4234555 4234608 candidate 228 Candidate_421_NC_002927 4243473 4243555 candidate 210 Candidate_422_NC_002927 4257897 4258015 candidate 380 Candidate_423_NC_002927 4310297 4310380 candidate 675 Candidate_424_NC_002927 4327601 4327646 candidate 850 Candidate_425_NC_002927 4341145 4341215 candidate 340 Candidate_426_NC_002927 4397117 4397197 candidate 930 Candidate_428_NC_002927 4429325 4429395 candidate 223 Candidate_429_NC_002927 4429121 4429357 candidate 218 Candidate_430_NC_002927 4429511 4429600 candidate 591 Candidate_431_NC_002927 4429527 4429600 candidate 591 Candidate_432_NC_002927 4434536 4434667 candidate 364 Candidate_433_NC_002927 4445994 4446035 candidate 850 Candidate_434_NC_002927 4456815 4457035 candidate 1140 Candidate_435_NC_002927 4480768 4481202 candidate 253 Candidate_436_NC_002927 4494826 4494888 candidate 329 Candidate_437_NC_002927 4515912 4516038 candidate 831 Candidate_438_NC_002927 4536554 4536632 candidate 262 Candidate_439_NC_002927 4551624 4551740 candidate 228 Candidate_440_NC_002927 4562013 4562062 candidate 676 Candidate_441_NC_002927 4638143 4638323 candidate 931 Candidate_442_NC_002927 4639162 4639215 candidate 1442 Candidate_443_NC_002927 4741289 4741342 candidate 1046

260 Candidate_444_NC_002927 474360 474483 candidate 1048 Candidate_445_NC_002927 4752624 4752817 candidate 245 Candidate_446_NC_002927 4799744 4799780 candidate 305 Candidate_447_NC_002927 482342 482571 candidate 117 Candidate_448_NC_002927 482651 482728 candidate 117 Candidate_449_NC_002927 4839965 4840035 candidate 875 Candidate_450_NC_002927 4880099 4880346 candidate 241 Candidate_451_NC_002927 4883186 4883244 candidate 209 Candidate_452_NC_002927 4925737 4925901 candidate 752 Candidate_453_NC_002927 4938900 4939002 candidate 222 Candidate_454_NC_002927 501171 501259 candidate 370 Candidate_455_NC_002927 504146 504259 candidate 269 Candidate_456_NC_002927 50857 51049 candidate 287 Candidate_457_NC_002927 5146299 5146351 candidate 555 Candidate_458_NC_002927 5147268 5147438 candidate 416 Candidate_459_NC_002927 5147470 5147622 candidate 1920 Candidate_460_NC_002927 515127 515409 candidate 815 Candidate_461_NC_002927 5160864 5161008 candidate 461 Candidate_462_NC_002927 5164713 5164767 candidate 316 Candidate_463_NC_002927 5178201 5178278 candidate 265 Candidate_464_NC_002927 5189400 5189589 candidate 219 Candidate_465_NC_002927 5222977 5223035 candidate 1401 Candidate_466_NC_002927 5264369 5264449 candidate 957 Candidate_467_NC_002927 5266573 5266768 candidate 1055 Candidate_468_NC_002927 5272379 5272472 candidate 205 Candidate_469_NC_002927 529755 529839 candidate 332 Candidate_470_NC_002927 5308938 5309012 candidate 470 Candidate_471_NC_002927 5314585 5314785 candidate 1150 Candidate_472_NC_002927 534419 534493 candidate 420 Candidate_473_NC_002927 537137 537244 candidate 358 Candidate_474_NC_002927 553677 553937 candidate 483 Candidate_475_NC_002927 573907 573961 candidate 405 Candidate_476_NC_002927 58048 58164 candidate 244 Candidate_477_NC_002927 633432 633546 candidate 1430 Candidate_478_NC_002927 6309 6425 candidate 940 Candidate_479_NC_002927 690614 690673 candidate 911 Candidate_480_NC_002927 690614 690731 candidate 911 Candidate_481_NC_002927 68986 69142 candidate 865 Candidate_482_NC_002927 712658 712744 candidate 1047 Candidate_483_NC_002927 721690 721756 candidate 201 Candidate_484_NC_002927 723945 723998 candidate 420 Candidate_485_NC_002927 74831 74978 candidate 842

261 Candidate_486_NC_002927 881682 881824 candidate 198 Candidate_487_NC_002927 896107 896162 candidate 269 Candidate_488_NC_002927 914146 914308 candidate 875 Candidate_489_NC_002927 928502 928590 candidate 122 Candidate_490_NC_002927 928872 928937 candidate 278 Candidate_491_NC_002927 94662 94888 candidate 242 Candidate_492_NC_002927 947995 948088 candidate 386 Candidate_493_NC_002927 950045 950145 candidate 273 Candidate_494_NC_002927 951485 951567 candidate 540 Candidate_495_NC_002927 952868 952924 candidate 883 Candidate_496_NC_002927 962476 962538 candidate 532 Candidate_497_NC_002927 96927 96997 candidate 226 Candidate_498_NC_002927 988694 988823 candidate 2875 ~ mini_ykkC_iter2_NC_002927 862187 862251 mini_ykkC_iter2 485 SRP_bact_NC_002927 1534186 1534264 SRP_bact 316 sucA_NC_002927 3885596 3885734 sucA 641 P9_NC_002927 1359553 1359632 P9 241 tmRNA_NC_002927 3723812 3723944 tmRNA 213

262

Appendix B: TransTerm predicted small RNA terminators.

bbsRNA Start End Strand Terminator Terminator Distance name Start End from sRNA 1 4484877 4484974 - 4479814 4479834 5043 3620838 3620935 - 3616692 3616721 4117 3622470 3622908 - 3616692 3616721 5749 4486509 4486947 - 4479814 4479834 6675 3 4493425 4493727 - 4489140 4489160 4265 4 5147238 5147452 - 5147274 5147300 -62 7 2206723 2206863 - 2206711 2206733 -10 8 3617287 3617457 - 3616692 3616721 566 4481326 4481496 - 4474391 4474411 6915 9 4484503 4484641 - 4479814 4479834 4669 3620464 3620602 - 3616692 3616721 3743 10 4073804 4073886 - 4072545 4072562 1242 11 4401795 4401862 - 4401768 4401797 -2 15 1456010 1456098 - 1455636 1455654 356 18 4397155 4397211 - 4397123 4397148 7 20 4234331 4234488 - 4231686 4231700 2631 21 4636206 4636268 - 4635782 4635807 399 24 170646 170701 - None 25 4148588 4148632 - None 30 2767205 2767372 - None 32 217610 217702 - 207827 207842 9768 34 4032089 4032124 - 4025251 4025270 6819 35 4117374 4117451 - 4116025 4116053 1321 - 4116024 4116054 1320 36 535582 535676 - 529808 529833 5749 37 3131675 3131758 - 3130855 3130868 807 38 2558978 2559049 - 2553584 2553615 5363 - 2553586 2553613 5365 39 2403491 2403561 - 2402247 2402263 1228 - 2402241 2402269 1222 40 1427632 1427685 - 1425711 1425742 1890 - 1425710 1425743 1889 41 3379965 3380072 - 3377398 3377422 2543 42 3819649 3819698 - 3819029 3819048 601 43 26212 26378 - 20061 20084 6128 44 530604 530658 - 529808 529833 771 - 529803 529838 766 45 4235832 4235924 - 4234641 4234655 1177 52 4223871 4223908 - 4223126 4223143 728 54 4178156 4178268 - 4178170 4178184 -28 55 1041354 1041439 - 1039308 1039331 2023 56 3249958 3250122 - None 57 2100138 2100195 - 2092922 2092937 7201

263 58 3314666 3314740 - 3312864 3312885 1781 59 1761266 1761315 - 1760540 1760563 703 60 4223117 4223180 - 4223126 4223143 -26 61 3885556 3885675 - 3885602 3885636 -80 62 5254046 5254082 - 5250924 5250943 3103 63 2721230 2721340 - None 66 4537607 4537692 - 4526166 4526188 11419 - 4526167 4526187 11420 71 1496239 1496458 - None 73 4454106 4454181 - None 76 3324932 3324975 - 3317700 3317741 7191 78 4365504 4365544 - None 80 2110463 2110537 - 2100939 2100960 9503 81 886765 886809 - 884076 884102 2663 82 3687177 3687247 - 3686603 3686622 555 83 2842702 2842905 - None 84 4938829 4938874 - 4938906 4938934 -105 - 4930008 4930034 8795 85 3979063 3979148 - 3979122 3979155 -92 89 4429341 4429390 - 4429332 4429349 -8 93 2111745 2111869 - 2111454 2111470 275 96 62968 63058 - 61628 61647 1321 - 61623 61652 1316 97 3983228 3983268 - 3979122 3979155 4073 98 3978590 3978733 - 3973067 3973091 5499 101 530826 530866 - 529803 529838 988 - 529808 529833 993 103 4479641 4479733 - 4474391 4474411 5230 105 4647582 4647629 - 4640841 4640866 6716 - 4640771 4640815 6767 106 4694954 4695005 - 4687665 4687701 7253 107 4122916 4123096 - 4122566 4122601 315 108 1103656 1103701 - 1092170 1092194 11462 109 3518771 3518841 - 3517580 3517606 1165 110 4912409 4912535 - 4907919 4907957 4452 - 4912578 4912610 -201 112 3164003 3164098 - 3155108 3155132 8871 113 300128 300200 - None 114 2736205 2736332 - None 117 4672037 4672073 - None 118 2702324 2702424 - 2700946 2700963 1361 119 3478166 3478278 - 3478171 3478190 -24 126 2101054 2101098 - 2100939 2100960 94 - 2100934 2100965 89 128 2874729 2874763 - None 130 4114378 4114433 - 4114394 4114419 -41 131 152712 152766 - None 132 97393 97466 - 92436 92459 4934 134 4910616 4910731 - 4907919 4907957 2659

264 136 189040 189125 - 187244 187267 1773 137 1175152 1175223 - 1170907 1170940 4212 140 3373878 3373929 - 3371142 3371163 2715 142 1139209 1139271 - None 144 3616764 3617014 - 3616692 3616721 43 150 4480719 4480959 - 4479814 4479834 885 153 411311 411368 - None 155 3679712 3679748 - 3671542 3671572 8140 156 3184599 3184699 - 3184583 3184602 -3 157 2055598 2055658 - 2055180 2055201 397 1974453 1974550 + None 2 1972480 1972918 + None 5 928735 928936 + 928905 928928 -31 6 1534240 1534372 + 1558150 1558173 23778 1977931 1978101 + 1990455 1990482 12354 1974786 1974924 + None 12 4633004 4633137 + None 13 3601640 3601836 + 3602487 3602508 651 14 3719812 3719946 + 3720554 3720574 608 16 2769741 2769799 + None 17 1761259 1761306 + None 19 5233190 5233463 + None 22 2767242 2767417 + 2767758 2767813 341 23 1054449 1054588 + 1061221 1061243 6633 26 4804958 4805066 + None 27 3604725 3604906 + 3605268 3605298 362 28 2736123 2736589 + 2737469 2737509 880 + 2737471 2737507 882 29 4834896 4834957 + 4835835 4835864 878 31 4827227 4827262 + 4835835 4835864 8573 33 1978220 1978466 + None 46 2117387 2117451 + None 47 2384958 2385014 + 2387533 2387557 2519 48 395761 395820 + 400048 400068 4228 49 3169500 3169577 + 3177230 3177256 7653 50 1359516 1359596 + 1359598 1359626 2 + 1359599 1359625 3 + 1359604 1359620 8 51 5247657 5247747 + 5250917 5250950 3170 + 5250920 5250947 3173 53 3913443 3913504 + 3913733 3913748 229 + 3913757 3913775 253 64 1504333 1504397 + 1505446 1505460 1049 65 2703745 2703879 + 2709770 2709787 5891 67 2379049 2379167 + 2379441 2379456 274 68 2268339 2268431 + 2277667 2277687 9236 69 100532 100732 + None 70 2327272 2327310 + None

265 72 2298090 2298231 + 2298269 2298286 38 74 4695015 4695092 + 4474391 4474411 -220701 75 63114 63175 + 69060 69073 5885 77 2484722 2484757 + None 79 3851536 3851589 + None 86 176964 177042 + 179259 179284 2217 + 179260 179283 2218 87 5151360 5151507 + 5152338 5152367 831 88 4661821 4662015 + 4661978 4661998 -37 + 4661746 4661781 -269 90 321225 321295 + 321306 321331 11 + 321386 321406 91 91 2694629 2694755 + 2700943 2700966 6188 + 2700946 2700963 6191 92 4666337 4666459 + 4667202 4667224 743 94 535509 535680 + 537197 537242 1517 95 901479 901515 + None 99 3679824 3679857 + 3684238 3684271 4381 100 6317 6426 + 11502 11534 5076 102 5272553 5272592 + 5273083 5273120 491 + 5273084 5273119 492 104 2704810 2704920 + 2709770 2709787 4850 111 42201 42417 + 50900 50934 8483 + 50902 50932 8485 + 50903 50931 8486 115 11542 11623 + 20061 20084 8438 116 2632555 2632596 + 2632548 2632582 -48 + 2632547 2632583 -49 120 2171928 2172097 + 2179042 2179057 6945 121 886737 886810 + 898504 898543 11694 122 5337023 5337152 + None 123 4005152 4005220 + 4007745 4007774 2525 + 4007752 4007767 2532 124 3824032 3824103 + 3824929 3824957 826 + 3824935 3824951 832 125 2229521 2229641 + 2229571 2229608 -70 + 2229574 2229605 -67 127 4816284 4816329 + None 129 2378229 2378265 + 2379441 2379456 1176 133 49743 49816 + 50900 50934 1084 + 50902 50932 1086 + 50903 50931 1087 135 2964013 2964055 + None 138 4494700 4494933 + 4494838 4494853 -95 139 3953167 3953226 + 3954071 3954088 845 141 3336834 3336869 + None 143 929038 929085 + None 145 34160 34220 + 39037 39059 4817

266 146 1782322 1782358 + 1791800 1791818 9442 147 42078 42200 + 50900 50934 8700 148 3540524 3540620 + None 149 4888933 4888963 + 4892219 4892246 3256 151 4268230 4268336 + 4268339 4268354 3 152 4191356 4191471 + 4194242 4194271 2771 + 4194243 4194270 2772 + 4194250 4194263 2779 154 3989538 3989609 + 3993185 3993211 3576 158 203877 203945 + 207827 207842 3882

267

Laura S. Weyrich Education 2007 South Dakota State University B. S. Applied and Industrial Microbiology 2012 The Pennsylvania State University Ph.D. Biochemistry, Microbiology, and Molecular Biology Professional Positions 2005 South Dakota State University West Nile Research Team Dr. M. B. Hildreth 2006 Manchester Metropolitan University Research Volunteer Dr. Anna Doran 2006-2007 South Dakota State University Research Assistant Dr. Volker Brözel 2007-2008 The Pennsylvania State University Teaching Assistant Dr. Sillman/Dr. Keating 2007-Present The Pennsylvania State University Research Assistant Dr. Eric Harvill Awards and Grants 2007 Nellie H. and Oscar L. Roberts Fellowship $5,000 2008 National Science Foundation Incentive Award $1,000 2008 College of Agricultural Sciences Travel Award for Cambridge, UK $500 2009 College of Agricultural Sciences Competitive Student Grant $2,000 2009 National Science Foundation Graduate Research Fellowship (NSFGRFP) $90,000 2009 College of Agricultural Sciences Fellowship $1,000 2010 American Society for Microbiology Travel Grant $500 2010 Pennsylvania State University BMMB Department Student Travel Grant $750 2010 Ninth International Bordetella Conference Travel Award $500 2011 Harold K. Schilling Dean’s Graduate Fellowship for Science and Ethics $2,500 2011 Institute for Molecular Evolutionary Genetics Travel Award $300 2011 National Science Foundation International Travel Award $1,000 Select Publications and Presentations 1.) L. Weyrich, S. Lindblom, C. Scofield, A. Rosa, S. Vilain, S. George, R. Kaushik, D. Francis and V. Brözel. Eubacterial Diversity of the Porcine Ileum of Weaned Pigs. SD Academy of Science. 2006. Chamberlain, SD. 2.) L. S. Weyrich, A.T. Karanikas, E.M. Goebel, X. Zhang, and E.T. Harvill. A Genetic Comparison of an Emerging Pathogen Bordetella holmesii with Other Pathogenic Bordetella Species. July 2010. International Bordetella Workshop. Cambridge, UK. 3.) Anne M. Buboltz, Tracy L. Nicholson, Laura S. Weyrich, and Eric T. Harvill. Role of the Type III Secretion System in a Hypervirulent Lineage of Bordetella bronchiseptica. Infect. Immun., 2009. 77(9): p. 3969-3977. 4.) Laura S. Weyrich, Olivier Y. Rolin, Sara E. Hester, Jihye Park, Nicholas Spidale, Mary J. Kennett, Chun Chen, Edward Dudley, and Eric T. Harvill. The Type VI Secretion System: A Novel Virulence Determinant in Bordetella bronchiseptica. American Society of Microbiology National Meeting. May 2010, San Diego, CA. 5.) Laura S. Weyrich, Olivier Y. Rolin, Sarah J. Muse, Jihye Park, and Eric T. Harvill. Host Microflora and Bordetella bronchiseptica Interactions during Respiratory Infection. International Congress of Mucosal Immunology Meeting. July 2011, Paris, France. 6.) Xuqing Zhang, Laura S. Weyrich, Alexia T. Karanikas, Jennie S. Lavine, and Eric T. Harvill. Bordetella pertussis vaccines are ineffective against B. holmesii due to lack of cross-reactive antibodies. EID. Submitted. 7.) Laura S. Weyrich, Olivier Y. Rolin, Sara E. Hester, Jihye Park, Nicholas Spidale, Mary J. Kennett, Chun Chen, Edward Dudley, and Eric T. Harvill. The Type VI Secretion System: A Novel Virulence Determinant in Bordetella bronchiseptica. PLoS One. Submitted. 8.) Laura S. Weyrich, Olivier Y. Rolin, Nathan T. Jacobs, and Eric T. Harvill. CD1d Manipulation by a Type VI Secretion System allows Bacterial Lower Respiratory Tract Persistence. JI. Submitted. 9.) L. S. Weyrich. Teaching Ethical Aptitude to Graduate Student Researchers. PLoS One. Submitted.