sign in sign up
<<
Home , Cytochalasin, Francisella, Francisella novicida, Francisella tularensis, Haemaphysalis concinna, Operation Whitecoat

A Dissertation

entitled

Characterization of a novel tularensis Factor Involved in Wall Repair by

Briana Collette Zellner

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the

Doctor of Philosophy Degree in

Biomedical Sciences

______Jason Huntley, Ph.D., Major Advisor

______R. Mark Wooten, Ph.D., Committee Member

______Jyl Matson, Ph.D., Committee Member

______Robert Blumenthal, Ph.D. Committee Member

______R. Travis Taylor, Ph.D., Committee Member

______Cyndee Gruden, PhD, Dean College of Graduate Studies

The University of Toledo

December 2019

© 2019 Briana Collette Zellner

This document is copyrighted material. Under copyright law, no parts of this document may be reproduced without the expressed permission of the author.

An Abstract of

Characterization of a Novel Involved in Repair

by

Briana Collette Zellner

Submitted to the Graduate Faculty as partial fulfillment of the requirements for the Doctor of Philosophy Degree in Biomedical Sciences

The University of Toledo December 2019

Francisella tularensis, the causative agent of , is one of the most dangerous bacterial known. F. tularensis has a low infectious dose, is easily aerosolized, and induces high morbidity and mortality; thus, it has been designated as a

Tier 1 . Studies to identify and characterize F. tularensis envelope proteins are important to help understand the molecular mechanisms by which F. tularensis, and other intracellular pathogens cause , and may lead to the development of new therapeutics. In previous studies, we demonstrated that the F. tularensis disulfide bond formation protein ortholog, DsbA, is required for virulence and, more importantly, identified >50 DsbA substrates, half of which are annotated as hypothetical proteins or proteins with unknown functions. In the current study, we selected one of these unstudied DsbA substrates, FTL1678, for detailed analysis. Using bioinformatics,

FTL1678 was found to contain a putative L,D-carboxypeptidase A (LdcA) domain, indicating a potential role in (PG) remodeling, which likely is required for the intracellular lifestyle of F. tularensis. Unlike prototypic LdcA homologs, F. tularensis LdcA does not localize to the cytoplasm. An FTL1678 mutant was completely

iii attenuated in a mouse pulmonary model, with decreased colonization and inability to disseminate to livers or spleens. Mutant attenuation was confirmed through complementation with wild-type (WT) FTL1678, as well as the jejuni

LdcA homolog Pgp2, and both fully-restored virulence to WT levels. Importantly, immunization with this mutant provided significant protection against pulmonary challenge with fully-virulent F. tularensis strain SchuS4 (in the BSL3). Membrane integrity testing revealed differences in cell wall permeability between WT F. tularensis

Live Strain (LVS) and ΔFTL1678 and electron microscopy analysis of

ΔFTL1678 showed increased outer membrane thickness. In addition, through enzymatic assays, FTL1678 was shown to have L,D-carboxypeptidase and L,D-endopeptidase activities, cleaving peptidoglycan pentapeptides to tetrapeptides and tripeptides. These studies have revealed a new F. tularensis virulence factor and have highlighted the importance of the F. tularensis envelope in protecting the bacterium during infection.

iv

This work is dedicated to my parents, Kent and Kathy Zellner. Your continued love and support over the past four and a half years have been a crucial component of my success.

You have always been my biggest cheerleaders in life, even when I wasn’t sure of myself, and I can’t thank you enough for always being there for me.

v

Acknowledgements

I would first like to thank my advisor, Dr. Jason F. Huntley for all of his support and guidance. I consider myself very lucky to have had the privilege of working in his laboratory for the past four years. My time in the Huntley lab, and his mentorship has made me a better scientist not only in a technical sense, but also as an independent and critical thinker. I would also like to thank my committee members, Drs. Jyl Matson, R.

Mark Wooten, Robert Blumenthal, and R. Travis Taylor, for all their advice on my project over the years and for being available and willing to help me whenever I needed it.

Next, I would like to thank the past and current members of the Huntley lab, Dr.

Wilma Wu, Nicole Bearss, Alison Brandel, Kayla Uveges, Nick Mitchell, Brenden Tully, and Alex McCartney for all your help and friendship through this process. I wouldn’t have made it through without you all.

Finally, I would like to thank all of our collaborators, Dr. Dominique Mengin-

Lecreulx (University of Paris), Drs. Joe Dillard and Ryan Schaub (University of

Wisconsin-Madison), Dr. Erin Gaynor (University of British Columbia), and Drs.

William Gunning and Robert Booth (University of Toledo) for their contributions to my dissertation research.

vi

Table of Contents

Abstract ...... iii

Acknowledgements ...... vi

Table of Contents ...... vii

List of Tables...... xii

List of Figures ...... xiii

List of Abbreviations ...... xv

1 Introduction and Literature Review ...... 1

1.1 Francisella tularensis ...... 1

1.1.1 A Brief History of Tularemia ...... 1

1.1.2 F. tularensis as a Biological Weapon ...... 3

1.1.3 F. tularensis , Subspecies, and Genetics ...... 4

1.1.4 Vectors, , and Clinical Disease ...... 7

1.1.5 F. tularensis Invasion of and Survival in Host Cells ...... 9

1.1.6 F. tularensis Pathogenesis ...... 11

1.1.7 Immune Responses to F. tularensis ...... 12

1.1.7.1 Innate Immunity ...... 12

1.1.7.2 Adaptive Immunity ...... 14

vii 1.2 Virulence Factors of F. tularensis ...... 17

1.2.1 Francisella ...... 17

1.2.2 Additional Francisella Systems ...... 19

1.2.3 F. tularensis Capsule ...... 20

1.2.4 F. tularensis LPS ...... 21

1.2.5 F. tularensis Envelope Proteins ...... 22

1.2.5.1 Outer Membrane Proteins ...... 22

1.2.5.2 Periplasmic Proteins ...... 24

1.2.5.3 Inner Membrane Proteins ...... 25

1.2.6 Disulfide Bond Formation Protein A and Substrates ...... 27

1.3 F. tularensis ...... 28

1.3.1 History of Tularemia Vaccines ...... 28

1.3.2 Models ...... 30

1.3.3 Immune Correlates of Protection ...... 32

1.3.4 Subunit, Killed Whole-Cell, or Live-Attenuated? ...... 34

1.3.5 Live-Attenuated Vaccines Derived from LVS vs. SchuS4 ...... 36

1.4 Gram-Negative Bacterial Peptidoglycan ...... 38

1.4.1 Peptidoglycan General Structure ...... 38

1.4.2 Peptidoglycan Synthesis ...... 39

1.4.3 Peptidoglycan Recycling ...... 40

1.5 Goals and Significance of my Dissertation Studies ...... 41

2 A Periplasmic L,D-Carboxypeptidase is Important for Cell Shape, Membrane

Integrity, and Virulence in F. tularensis ...... 44

viii 2.1 Abstract ...... 45

2.2 Introduction ...... 46

2.3 Results ...... 48

2.3.1 FTL1678 Contains a Putative L,D-Carboxypeptidase Domain ...... 48

2.3.2 FTL1678 Exhibits L,D-Carboxypeptidase and L,D-Endopeptidase

Activities ...... 52

2.3.3 FTL1678 Controls Bacterial Morphology ...... 55

2.3.4 Deletion of FTL1678 Affects Membrane Integrity and Permeability

...... 58

2.3.5 FTL1678 and Ldc Activity are Required for F. tularensis Virulence

...... 60

2.3.6 FTL1678 and Ldc Activity, but not the Ldc Catalytic Triad, are

Required for F. tularensis virulence ...... 63

2.3.7 FTL1678 is Required for F. tularensis Replication in

...... 63

2.3.8 ΔFTL1678 Protects Mice Against Type A F. tularensis Infection .. 66

2.3.9 ΔFTL1678 Does not Cause Tissue Damage...... 68

2.4 Discussion 71

2.5 Materials and Methods ...... 76

2.5.1 Bacterial Strains and Culture Conditions ...... 76

2.5.2 Sequence Alignments and Bioinformatic Predictions ...... 77

2.5.3 Generation of F. tularensis Gene Deletion Mutants ...... 78

2.5.4 FTL1678 Complementation in trans...... 79

ix 2.5.5 C. jejuni Pgp2 Complementation in trans ...... 80

2.5.6 Mouse ...... 81

2.5.7 Membrane Integrity Testing ...... 82

2.5.8 Electron Microscopy ...... 83

2.5.9 Spheroplasting and Sucrose Density Gradient Centrifugation ...... 84

2.5.10 Immunoblotting ...... 85

2.5.11 Infections of Mouse Bone Marrow-Derived Macrophages and

J774A.1 Macrophages ...... 86

2.5.12 Expression and Purification of Recombinant FTL1678 and

FTT0101 ...... 87

2.5.13 Enzymatic Assays for FTL1678/FTT0101 Activity ...... 88

2.5.14 Peptidoglycan Precursors and Muropeptides ...... 89

2.5.15 Statistics ...... 90

2.6 Supplemental Material ...... 91

3 and Tularemia: Do we Know What we Don’t Know? ...... 101

3.1 Abstract ...... 102

3.2 Introduction ...... 102

3.3 Tularemia-Associated Species, Tick Infection Rates, and Geographic

Locations ...... 105

3.4 Francisella-like Endosymbionts ...... 111

3.5 Transstadial Transmission of F. tularensis in Ticks ...... 112

3.6 F. tularensis-Tick Interactions ...... 114

3.7 Conclusions ...... 118

x 4 Type IV Pili and virulence in F. tularensis ...... 120

4.1 Introduction to Type IV Pili ...... 120

4.2 Explanation of Second Research Project ...... 123

4.3 Results and Future Directions ...... 126

5 Summary and Future Directions of FTL1678 Project ...... 130

References ...... 135

A Reproduction of Figure 3-1 and Table 3.1 ...... 183

xi

List of Tables

2.1 Substrate specificity and specific activity of F. tularensis FTL1678 and

FTT0101 ...... 55

2.2 Sensitivity of F. tularensis WT LVS and ΔFTL1678 to , detergents,

and dyes ...... 61

2.S1 Bioinformatic analyses of FTL1678 ...... 92

2.S2 Sensitivity of F. tularensis WT LVS, ΔFTL1678, FTL1678 trans-complement,

and Pgp2 trans-complement to antibiotics, detergents, and dyes ...... 93

2.S3 Bacterial strains and plasmids used in this study ...... 94

2.S4 Primers used in this study ...... 95

3.1 Ticks associated with tularemia ...... 110

4.1 Bioinformatic analyses of FTL1695 ...... 125

xii

List of Figures

2 – 1 Amino acid alignment of bacterial L,D-carboxypeptidases ...... 51

2 – 2 FTL1678 is OM-associated ...... 53

2 – 3 F. tularensis ΔFTL1678 has altered cell morphology ...... 57

2 – 4 Susceptibility of F. tularensis WT LVS and ΔFTL1678 to stressors ...... 59

2 – 5 FTL1678 and Ldc activity are required for F. tularensis virulence ...... 63

2 – 6 FTL1678 is required for F. tularensis replication in macrophages ...... 67

2 – 7 ΔFTL1678 protects against fully-virulent Type A F. tularensis SchuS4 ...... 68

2 – 8 ΔFTL1678 does not induce pathology ...... 71

2 – 9 Peptidoglycan synthesis and recycling pathways remain to be defined in F.

tularensis ...... 77

2 – S1 FTL1678 contains a putative L,D-carboxypeptidase domain ...... 95

2 – S2 TolB is OM-localized ...... 96

2 – S3 FTL1678 is required for correct septation ...... 97

2 – S4 FTT0101 is not required for virulence ...... 98

2 – S5 Catalytic triad is not essential for F. tularensis virulence ...... 99

2 – S6 FTL1678 trans-complement and C. jejuni Pgp2 trans-complement restore WT F.

tularensis LVS phenotype ...... 100

3– 1 U.S. geographic distribution of ticks associated with human tularemia ...... 101

4 – 1 FTL1695 contains a putative Type IV pili domain ...... 123

xiii 4 – 2 ΔFTL1695 does not have an inherent growth defect ...... 126

4 – 3 FTL1695 plays a role in F. tularensis virulence ...... 128

xiv

List of Abbreviations

ATP ...... Adenosine Triphosphate ABC Transporter ...... ATP-binding Cassette Transporter

BHI ...... Heart Infusion BSL2 ...... 2 BSL3 ...... Biosafety Level 3

CDC ...... Centers for Disease Control CDM ...... Chamberlain’s Defined Medium CFU ...... Colony Forming Unit CTAB ...... Cetyltrimethyl Ammonium Bromide CHAPs ...... 3-Cholamidopropyl Dimethylammonio 1-Propanesulfonate

DC ...... Dendritic Cell DMEM ...... Dulbecco’s Modified Eagle Medium DsbA ...... Disulfide Bond Formation Protein A DsbB ...... Disulfide Bond Formation Protein B DsbC ...... Disulfide Bond Formation Protein C DsbD ...... Disulfide Bond Formation Protein D

EEA-1 ...... Early Endosome Antigen-1

FCP ...... Francisella-containing FLE ...... Francisella-like Endosymbionts FPI ...... Francisella Pathogenicity Island Ft...... Francisella tularensis

GlcNAc ...... N-acetylglutamic acid Ge ...... equivalent h ...... hour/s HMM PBP...... High Molecular Mass Binding Protein hMDM ...... human Monocyte-Derived

IFN-γ ...... i.d...... Intradermal

xv IgA ...... Immunoglobulin A IgG ...... Immunoglobulin G IgM ...... Immunoglobulin M IM...... Inner membrane i.n...... Intranasal i.p...... Intraperitoneal IPTG ...... Isopropyl β-D-Thiogalactopyranoside i.v...... Intravenous

KDO...... Keto-3-deoxyoctulsonic acid

LAMP-1 ...... Lysosomal-Associated Membrane Protein-1 LAMP-2 ...... Lysosomal-Associated Membrane Protein-2 LD50 ...... Median Lethal Dose Ldc ...... L,D-Carboxypeptidase LMM PBP ...... Low Molecule Mass Penicillin Binding Protein LPS ...... LT ...... Lytic Transglycosylase LVS ...... Live Vaccine Strain LVSG ...... LVS gray colony variant meso-A2pm ...... meso- MHC ...... Major Histocompatibility Complex MR ...... Mannose Receptor mBMDM ...... murine Bone Marrow-Derived Macrophage mRNA ...... messenger Ribonucleic Acid MurNAc ...... N-acetylmuramic acid

NF-κB ...... Nuclear Factor kappa B NK cell ...... Natural Killer cell NO ...... Nitric Oxide NOD-1 ...... Nucleotide-Binding Oligomerization Domain-Containing Protein 1

OM ...... Outer Membrane OMP ...... Outer Membrane Protein

PAI ...... Pathogenicity Island PAL ...... Peptidoglycan-Associated Lipoprotein PBL ...... Peripheral Blood Lymphocyte PBP ...... Penicillin Binding Protein PG ...... Peptidoglycan qRT-PCR...... quantitative Real-Time PCR

RANTES ...... Regulated (on) Activation Normal T Cell Expressed (and) Secreted ROS ...... Reactive Oxygen Species

xvi RNS ...... Reactive Nitrogen Species s.c...... subcutaneous SOE-PCR...... Splicing Overlap Extension-Polymerase Chain Reaction sMHA...... supplemented Mueller Hinton Agar sMHB...... supplemented Mueller Hinton Broth SR-A ...... Scavenger Receptor A

TCT ...... Tracheal Cytotoxin TBD ...... Tick-Borne Disease TEM ...... Transmission Electron Microscopy TFP ...... Type IV Pili TLR-4 ...... Toll-like Receptor 4 TMD ...... Transmembrane Domain TNF-α ...... Tumor Necrosis Factor α T1SS ...... Type 1 Secretion System T2SS ...... Type 2 Secretion System T6SS ...... Type 6 Secretion System

UDP ...... Uridine-Diphosphate

WCL ...... Whole Cell Lysate WT ...... Wild-type

xvii Chapter 1

Introduction and Literature Review

1.1 Francisella tularensis

1.1.1 A Brief History of Tularemia

Francisella tularensis was first described in 1910 by George W. McCoy as causing a -like disease of ground squirrels in San Francisco, California (1).

McCoy and Charles Chapin later isolated and cultured coccobacilli organisms from a outbreak in Tulare County, California in 1912, and hence named the organism

Bacterium tularense (2). McCoy and Chapin observed that the organism was experimentally transmissible to other , including guinea pigs and rats, and that infection resulted in enlarged lymph nodes and necrotic livers and spleens. The first microbiological identification of tularemia in occurred in 1913 at the University of Cincinnati in Cincinnati, Ohio by Dr. William B. Wherry (3), although several reports dating back as early as the 1890s described a tularemia-like illness (4). Wherry recognized an eye infection similar to the disease seen in ground squirrels in California, collected conjunctival scrapings from the patient, and injected them into a guinea pig, which later died and was confirmed positive for B. tularense.

1 The first terminal case of B. tularense infection was investigated by Edward

Francis in 1919. Francis was sent to Salt Lake City, Utah to study a new disease of man, which many were referring to as “ Fly ”. More than 70 cases of the disease had been reported over a three-year period in that area, and it was thought that the infection initiated with a (Chrysops discalis) bite (5). Francis isolated B. tularense from 7 humans, observed that inoculation of guinea pigs with either patient blood or was fatal and caused necrotic lesions throughout livers and spleens, and cultures from these lesions grew on coagulated egg yolk (6). The disease was hereafter known as tularemia.

The eastern United States saw its first reported case of tularemia in 1921, where the term “rabbit fever” was coined. Since then, tularemia has become well-known as rabbit fever due to the association of rabbit hunters/trappers with tularemia and F. tularensis aerosolization if an infected rabbit is mistakenly run over by a lawnmower.

Although transmission by infected rabbits was most prevalent in the early 1920s, a

“glandular type of tick fever” started appearing around 1923, perhaps indicating the first incidences of tularemia transmitted by ticks (6). F. tularensis was first isolated from a naturally-infected tick, andersoni Stiles, in 1924 (7). By the year 1928, 455 cases of tularemia had been reported in the United States, 24 of which were fatal, highlighting the need for tularemia research and the development of treatments (8). It was not until 1947 that B. tularense was renamed Francisella tularensis, in honor of Dr.

Edward Francis and his extensive contributions to tularemia research.

2 1.1.2 F. tularensis as a Biological Weapon

F. tularensis is one of the most dangerous bacterial pathogens known due to an infectious dose of less than 10 organisms (Type A) and mortality rates of up to 60% if untreated (9). Due to the extreme virulence of this , and the relative ease with which it can be aerosolized, F. tularensis is classified as a Tier 1 Select Agent by the U.S.

Centers for Disease Control (CDC), highlighting that it poses the greatest risk of potential misuse as a biological weapon, with significant risks for mass casualties, devastating economic losses, disruptions to normal activities, and public panic. One study estimated that an attack with aerosolized F. tularensis over an area of 100,000 people would result in 82,500 cases of pneumonic or typhoidal tularemia, resulting in 6,188 deaths, and $5.4 billion in medical expenses (10). Due to the short of F. tularensis, the study suggested that treatment would need to be started within 72 h of a suspected attack to minimize mass casualties and other major impacts. However, because tularemia often presents with non-specific flu-like symptoms, a correct diagnosis is unlikely to be made quickly, highlighting the importance of developing an anti-

Francisella vaccine.

The potential use of F. tularensis as a biological weapon dates back to the 1930s when, along with other biowarfare agents, the Japanese were studying its weaponization

(11). Additionally, a former Soviet Union physician and scientist, Ken Alibek, claims that the Soviets stockpiled weaponized F. tularensis at a facility known as Biopreparat and that the thousands of tularemia cases during World War II may have been due to intentional releases (12). In the 1950s and 1960s, the U.S. also was developing means to disseminate F. tularensis aerosols. However, by 1973 all U.S. offensive weapons

3 programs ceased due to an executive order. Today, the U.S. only conducts research on and prepares for rapid responses to biological attacks, including decontamination, treatments, and vaccines (13). As such, understanding how F. tularensis causes disease and identifying the individual molecules that are required for virulence of this pathogen are very important for developing novel therapeutics and vaccines.

1.1.3 F. tularensis Species, Subspecies, and Genetics

F. tularensis, the causative agent of tularemia, is a highly-infectious, Gram- negative, (14). After several genera changes, from Bacterium, to

Pasteurella, to , F. tularensis was formally placed in the genus Francisella in

1947 (9). Francisella, the only genus in the family Francisellaceae, is a member of the

Gammaproteobacteria subclass (15, 16). The Francisella genus includes four separate species: F. tularensis, F. novicida, F. philomiragia, and F. noatunensis (17), although only F. tularensis is a major concern for human pathogenesis. F. noatunensis is primarily a pathogen and does not cause disease in humans (18). F. novicida and F. philomiragia often are found in or near water, are not associated with -borne disease, and are considered opportunistic pathogens in that they only cause rare disease in immunocompromised individuals.

Despite the low virulence of F. novicida in humans, it is sometimes used as a surrogate to study F. tularensis pathogenesis (19). F. novicida (U112) was first isolated from Ogden Bay in Salt Lake City, Utah in 1950 and, upon initial examination of experimentally-infected guinea pigs, was thought to be F. tularensis (20) . However, after further testing, the isolated bacterium was determined to be a separate species from

4 F. tularensis, based on its inability to agglutinate using anti-F. tularensis , its ability to ferment sucrose, and its lower pathogenicity in animals. More recently, one group proposed that F. novicida should be formally changed to a subspecies of F. tularensis (i.e., F. tularensis subsp. novicida), based on high genetic similarity and quinone and polar lipid comparisons between F. novicida and various Francisella species

(21). However, eleven leaders in the tularemia field immediately objected to the proposed taxonomic change, outlining that differences in biosafety levels, risks of airborne infections, growth and metabolic requirements, and extensive whole genomic analyses do not support F. novicida being an F. tularensis subsp. (22). The classification of F. novicida as a separate species remains controversial and, although a variety of nomenclatures are used in publications, there is considerable evidence documenting phenotypic and genotypic differences between F. tularensis and F. novicida (19), highlighting the need to distinguish between the two species.

F. tularensis is further divided into three subspecies: subsp. tularensis (Type A), subsp. holarctica (Type B), and subsp. mediasiatica. F. tularensis subsp. tularensis is the most virulent with an infectious dose of <10 CFU and is located solely in North

America, whereas subsp. holarctica is associated with a milder form of the disease and has an infectious dose of <103 CFU (9). F. tularensis subsp. mediasiatica has a similar virulence as subsp. holarctica, however, it is geographically restricted to central Asia only, whereas F. tularensis subsp. holarctica is found all throughout the northern hemisphere, resulting in a higher rate of human and animal (23). F. tularensis subsp. tularensis is most often associated with ticks and rabbits, whereas F. tularensis subsp. holarctica is associated with semiaquatic , hares, and water sources such as

5 streams, lakes, rivers, and ponds (4). Additionally, as genotyping methods have improved over the years, researchers have been able to differentiate between Francisella species and very closely related species, including tick endosymbionts such as persica. Tick endosymbionts of F. tularensis will be discussed in Chapter 3. Improved genotyping also led to the identification of two distinct subpopulations of F. tularensis subsp. tularensis, A1 and A2, which can be further subdivided into A1a, A1b, and A2a, all of which vary in geographic distribution and virulence (23). A1 subpopulations are generally found in the eastern United States whereas A2 subpopulations are located mostly in the western United States, with a few exceptions. A comprehensive CDC study of F. tularensis isolates and human disease over a 40-year period found that A1b strains were more fatal (24% mortality) than A1a (4% mortality) or A2 strains (0% mortality)

(24). Although much less is known, 11 subclades of F. tularensis subsp. holarctica also have been recently identified (25).

Most F. tularensis subsp. tularensis and F. tularensis subsp. holarctica strains only can be handled under biosafety level 3 (BSL3) conditions to prevent infection of laboratory workers and release of the bacterium, although there are attenuated strains which can be handled under BSL2 conditions. One of the most commonly-used BSL3 lab strains of F. tularensis subsp. tularensis is SchuS4, an A1a strain, which was originally isolated from a human infection in Ohio. The F. tularensis LVS is an attenuated Type B strain that was originally developed in the former Soviet Union in the

1950s (26). LVS is not a licensed vaccine in the U.S. because the correlates of protection remain under study, the basis of its attenuation is unknown, it retains mild virulence in humans, and it only provides partial protection against virulent Type A challenge in

6 humans (27). However, LVS often is used as a surrogate to study F. tularensis virulence because the LVS genome is 99.3% identical to the SchuS4 genome, and LVS can be manipulated under BSL2 conditions (28).

1.1.4 Vectors, Transmission, and Clinical Disease

Tularemia is a zoonotic disease that infects over 300 different species, including humans (29). Due to the extremely broad host range of this pathogen, it is unclear what the true reservoir host/hosts are and how the are maintained in the environment, although F. tularensis often is associated with rodents and lagomorphs (30). In the first half of the , >90% of all reported cases of tularemia were due to exposure to infected rabbits or hares (31). , including ticks, deer flies, mosquitoes, and horse flies, also play an important role in transmission of tularemia because they feed on infected animals and subsequently bite humans, transferring F. tularensis (4, 32). As noted above, tick-borne tularemia was first identified in 1923 (7) and has since become the main transmission of tularemia in the United States, accounting for approximately half of all tularemia cases (24, 33).

Tularemia manifests in several different clinical forms and each is dependent upon the route of infection. Once infection occurs, the incubation period varies from one to 14 days, but symptoms typically present between three and five days (CDC, 2018).

The most common form of tularemia is ulceroglandular, which typically results from handling infected animal carcasses or from the bite of an infected arthropod such as ticks or deer flies (9, 34). Ulceroglandular tularemia was common among market men in the early half of the 20th century (6) and remains a problem for hunters and trappers.

7 Symptoms of ulceroglandular tularemia start appearing three to six days after infection and often resemble the flu, causing fever, malaise, , , and an which forms at the infection site that persists for months (9). Once inside the body, F. tularensis travels through the to regional lymph nodes, causing , and can then disseminate to other tissues such as the , liver, spleen, kidneys, or the central nervous system (9). Glandular tularemia is typically characterized by large, swollen, painful lymph nodes. Clinical signs of glandular tularemia are very similar to ulceroglandular, except that ulcers are not present.

Oculoglandular tularemia causes many of the same flu-like symptoms, although the initial site of infection is the conjunctiva, which usually occurs by transfer of bacteria from hands to the eyes. Ingestion of contaminated food or water can lead to oropharyngeal tularemia, which, in addition to the symptoms listed above, also can cause a painful sore throat, , and swollen cervical lymph nodes. It is well-known that

F. tularensis can survive in fresh and brackish water, and many outbreaks of waterborne tularemia have been reported in and Asia (32). Typhoidal tularemia is described as septicemic tularemia without lymphadenopathy or ulcers.

The most serious form of the disease, pneumonic tularemia, can occur as a result of any of the other forms of tularemia that go untreated and spread to the lungs or as a result of inhalation of the bacteria (9). The symptoms of pneumonic tularemia include fever, coughing, chest pain, and difficulty breathing. Tularemia can be difficult to diagnose because it is a rare disease, the symptoms can be variable, and the symptoms resemble more common illnesses such as the flu. As noted above, there currently is no licensed vaccine for tularemia, but it is treatable with antibiotics if correctly and quickly

8 diagnosed. Tularemia is typically diagnosed with blood tests and cultures and can be treated with , , , or (CDC, 2018).

1.1.5 F. tularensis Invasion of and Survival in Host Cells

F. tularensis is a facultative intracellular pathogen that readily infects macrophages (4) and, although F. tularensis can be found in the extracellular plasma of bacteremic mice (35), intracellular replication is an essential component of F. tularensis pathogenesis. Studies have demonstrated that intranasal infection of mice with LVS or

SchuS4 results in alveolar macrophages being infected first, accounting for 70% of all infected cells during the first 24 hours (h) (36). Phagocytic uptake of F. tularensis by macrophages occurs via either a novel pseudopod loop mechanism (37) or by recognition by several different receptors (e.g., mannose receptor [MR], complement receptor CR3, scavenger receptor A [SR-A], Fcγ receptor), depending on the serum opsonization state of the bacteria (38, 39). The MR is involved with non-opsonic uptake of F. tularensis

(40), while CR3 (40), SR-A (41) and Fcγ (42) are associated with serum-opsonized uptake. Whereas opsonization of F. tularensis results in increased uptake of the bacteria, phagosomal escape and intracellular replication are decreased, suggesting that F. tularensis may control exposure of certain bacterial ligands to be preferentially internalized by non-opsonized conditions (42). Once phagocytosed by a macrophage, F. tularensis is contained within a Francisella-containing phagosome (FCP), where it enters the normal endocytic pathway and acquires markers of early endosomes (e.g., EEA-1) and late endosomes (e.g., CD63, LAMP-1, LAMP-2, and Rab7). However, the FCP never fuses with lysosomes, as FCPs do not accumulate lysosomal markers such as

9 cathepsin D (38, 43, 44). Studies have shown that under non-opsonized conditions, FCPs do become acidified prior to phagosomal disruption and F. tularensis escape to the macrophage cytosol (45).

Francisella phagosomal escape is a well-documented process in human and murine macrophages, although escape kinetics have caused controversy in the field (43-

47). For example, one study demonstrated F. tularensis LVS macrophage phagosomal escape by two hours post-infection (46), whereas a second study noted approx. 50% of

LVS still contained within a phagosome six hours post-infection (43). In yet another study, LVS demonstrated rapid phagosomal escape, occurring within one hour of infection (44). SchuS4 also exhibits rapid phagosomal escape from macrophages, with

50% of bacteria free in the cytosol by 30 min post-infection (48). However, discrepancies in phagosomal escape kinetics may be attributable to technical differences of the assays used in individual studies. Nevertheless, a major contributor to phagosomal escape appears to be complement or opsonization, as rapid phagosomal escape occurs under nonopsonic conditions (44-46, 48) and slower escape occurs under opsonic conditions (43). Finally, phagosomal escape is paramount for F. tularensis replication, evidenced by several escape-deficient mutants that are unable to replicate in macrophages

(47-50).

F. tularensis replicates within the cytosol until host resources are depleted, the host cells lyse, and bacteria are released into the extracellular environment, where they can infect more cells (51). Francisella has also been shown to reenter the endocytic pathway via host cell autophagy (44). While many studies use in vitro macrophage infection models, F. tularensis can infect a wide range of host cells including fibroblasts,

10 endothelial cells, erythrocytes, polymorphonuclear neutrophils, hepatocytes, muscle cells, alveolar Type II cells, and dendritic cells, making it an interesting intracellular pathogen to study (4, 29, 38, 52).

1.1.6 F. tularensis Pathogenesis

Regardless of the infection route, F. tularensis infection results in a bacteremic phase, which allows for dissemination of the bacteria to distal organs such as the lungs, liver, spleen, and lymph nodes, causing a systemic infection. Histopathologic analysis of

10 human tularemia cases revealed several commonalities: 1. Lymph nodes were grossly necrotic and several presented with granulomatous lesions; 2. Lungs were characterized by necrotizing with abundant neutrophils in alveolar walls and spaces, accompanied by edema and often supperative pleuritis; and 3. Livers and spleens were filled with microabscesses (53). The patients, exposed to tularemia via several different routes, experienced a wide range of symptoms including chills, malaise, fever, , lesions, and even gastrointestinal symptoms. Similar pathology is seen in animal models – SchuS4 aerosol infection caused necrotizing bronchopneumonia, encompassing most of the lung by day four post-infection in rabbits, as well as liver and splenic lesions in rabbits (54) and monkeys (55).

11 1.1.7 Immune Responses to F. tularensis

1.1.7.1 Innate Immunity

Despite rapid and robust replication of F. tularensis in host cells, the bacterium does not induce a strong inflammatory response upon infection (56, 57). In fact, there is a lack of cytokine and chemokine release during the first 48 h of F. tularensis infections.

A transcriptional profiling study indicated a rise in TNF-α and IFN-γ mRNA transcripts in the lung between two and four days post-aerosol infection with virulent Type A

Francisella (58). In addition, serum/distal organ levels of RANTES, IL-6, and IL-1β begin to rise between three and four days post-infection (59). This late onset of cytokines and chemokines resembles a cytokine storm, which ultimately results in severe and organ failure (60). However, proinflammatory cytokines, including IFN-γ, TNF-α, and

IL-12, are critical for controlling F. tularensis infections. For example, TNF-α and IFN-γ depletion resulted in a reduced time-to-death in LVS-infected scid (61) and C3H/HeN mice (62). Conversely, certain anti-inflammatory mediators have been shown to impair host immune responses and promote survival of the pathogen, including TGF-β and IL-

10 (63).

Macrophages, dendritic cells (DCs), and natural killer (NK) cells play important roles early in pulmonary F. tularensis infections. One study noted that F. tularensis LVS induces partial activation of macrophages and DCs, leading to upregulation of MHC class

II and CD86, but does not result in proinflammatory cytokine secretion. Rather, upregulation of TGF-β, an anti-inflammatory cytokine occurs, allowing the bacteria to establish infection (64). These data suggest that macrophages and DCs prolong F. tularensis infections rather than actively clearing the bacteria. In addition, a second study

12 provided evidence for alternative activation of macrophages following an initial proinflammatory response, resulting in decreased expression of IL-12 p70, TNF-α, and

IL-1β, and increased expression of the anti-inflammatory mediators IL-4, IL-13, and

TGF-β (65). Similarly, fully-virulent F. tularensis SchuS4 actively suppresses activation of lung macrophages and DCs, and prevents the secretion of TNF-α, IL-12, and IL-1β during the first 24 to 48 h of infection (57). Finally, NK cells, which also are major producers of IFN-γ, do not release this cytokine until at least 72 h after LVS infection

(66).

Although F. tularensis resists killing by, and does not stimulate an oxidative burst in neutrophils (67), neutrophils are clearly important in responding to Francisella infections. Following respiratory infection, neutrophils account for up to 50% of

Francisella-infected cells in the lungs approx. three days post-infection (36) and are recruited, in part, by IL-17A (68). Neutrophil depletion from WT C57BL/6 mice allowed unrestricted LVS growth in lungs, livers, and spleens, and caused a normally sublethal intradermal (i.d.) infection to become rapidly lethal in these mice (69), as well as in scid mice (61). However, this same effect was not seen in response to LVS aerosol challenge, as bacterial burdens were comparable in neutropenic and immunocompetent BALB/c mouse lungs and spleens (70), suggesting that neutrophil function in F. tularensis infection varies from tissue to tissue.

13 1.1.7.2 Adaptive Immunity

B cells and antibodies respond to F. tularensis infections but the true role of humoral immune responses in protecting against tularemia is unknown. Antibodies have become an important diagnostic tool, as IgM, IgG, and IgA antibody titers substantially increase approximately two weeks after infection, peak around 1-2 months, and have been detected up to 11 years later in humans naturally infected with tularemia (71).

Indeed, tularemia clinical diagnosis is based on a four-fold increase in serum antibodies between acute and convalescent sera (60). Furthermore, although a large portion of B cell contribution is antibody-mediated, B cells also have roles in production of cytokines and chemokines, antigen presentation, and recently were discovered to permit replication of Francisella (72). Older literature provides many examples of antibody responses following natural infection with tularemia and after vaccination, mostly with LVS (63).

Most antibody responses to Francisella infection are directed against lipopolysaccharide

(LPS), though other proteins also are recognized (73). Humans vaccinated with LVS also develop IgM, IgG, and IgA serum antibodies approx. two weeks after vaccination that persist for at least one and a half years (74, 75). However, anti-Francisella serum antibodies have not been predictive of protection against virulent strains.

Passive transfer of anti-LVS serum, purified anti-Francisella LPS antibodies, or other bacterial components has demonstrated partial protection against subsequent F. tularensis challenge. For example, passive transfer of sera from LPS-immunized mice to naïve mice protected against intraperitoneal (i.p.) LVS challenge and immunization of naïve mice with LPS also protected against LVS challenge, but not fully-virulent SchuS4 challenge (76). Similarly, a second study found that passive transfer of immune serum to

14 naïve mice protected against intranasal LVS challenge (77). Finally, immunization with

Francisella outer membrane proteins (OMPs) resulted in robust production of IgM, IgG, and IgA antibodies, and conferred partial protection (50%) against F. tularensis SchuS4 challenge (78). Interestingly, a recent study demonstrated that SchuS4 is capable of binding host plasmin, a serine protease that can degrade opsonizing antibodies, possibly explaining the resistance of highly-virulent Francisella to antibody treatment (79). In summary, although antibodies alone are not sufficient to protect against fully-virulent F. tularensis, they do have a significant role in host responses to Francisella infection.

As noted above, B cells and antibodies are only partially protective against F. tularensis, therefore optimal protection requires cell-mediated immune responses.

Although IFN-γ and TNF-α are sufficient to control the early stages of F. tularensis LVS infection in scid mice, those mice eventually succumb to infection. However, when those mice were reconstituted with spleen cells from WT BALB/c mice, they were able to survive i.d. LVS infection long-term(61). Similarly, nu/nu mice (deficient for T cells) survived i.d. LVS challenge only until day 30 post-infection. Reconstitution with CD4+ and CD8+ T cells rescued nu/nu mice, demonstrating the importance of T lymphocytes in protection against Francisella (80). To better understand the specific roles of T cells in protecting against F. tularensis, many studies have individually depleted CD4+, CD8+, or other T cell subsets. For example, mice depleted of CD4+, CD8+, or both subsets of T cells were able to control LVS growth following intravenous (i.v.) challenge, but could not fully resolve the infection (81). The inability to clear LVS from T cell depleted mice, resulted in chronic infection (mice survived for approx. eight weeks post-infection), showing that while T cells are not required to prevent bacterial growth, they are

15 important for complete destruction of bacteria from the tissues and long term survival of infected mice. Furthermore, isolated CD4+, CD8+, or Thy1+CD4−CD8− T cell subsets are sufficient to control LVS growth in BMMs, likely through the production of Th1-type cytokines, including IFN-γ, TNF-α, and IL-2 (82).

Although T cells generated in response to parenteral LVS infection are capable of controlling bacterial replication, virulent Type A aerosol infection results in thymus atrophy and CD4+ CD8+ depletion (83, 84). The exact mechanism of this depletion is not fully-understood, however, it represents a key pathogenic feature of highly-virulent

Francisella. Vaccine-induced protection against secondary pulmonary challenge with virulent Francisella requires both CD4+ and CD8+ T cells, as depletion of either significantly reduced survival of mice (85). Similar to mice, humans also develop CD4+ and CD8+ T cell responses to Francisella infection that demonstrate an effector memory phenotype (86). While T cell-specific mechanisms for protection against F. tularensis remains an active area of study by many research laboratories, it is clear that T cell production of TNF-α, IFN-γ, and IL-17A are important (87). For all the reasons listed above, vaccine development efforts against F. tularensis recently have focused on generating protective T cell responses and determining the correlates of protection.

16

1.2 Virulence Factors of Francisella tularensis

1.2.1 Francisella Pathogenicity Island

Pathogenicity islands (PAI) are genetic elements present on bacterial chromosomes that typically are acquired by , are only found in pathogenic strains, and are absent in nonpathogenic strains (88). PAIs contain genes that are required for bacterial virulence and identification of PAIs can lead to the discovery of novel virulence factors. A Francisella Pathogenicity Island (FPI) of about 30 kb (16 to

19 genes), has been described in F. tularensis and was demonstrated to be important for intramacrophage survival and F. tularensis virulence (89). The FPI is surrounded by transposable elements, has a lower GC content than the rest of the F. tularensis genome, and, interestingly, is duplicated in F. tularensis subsp. tularensis and F. tularensis subsp. holarctica. However, only one copy of the FPI is present in F. novicida (89), highlighting that F. novicida is distinct from F. tularensis. In addition, FPI copy number directly correlates with differences in virulence between F. tularensis and F. novicida, further supporting that F. novicida should remain its own species. However, the presence of two copies of the FPI in F. tularensis subsp. made genetic studies of the FPI difficult.

As such, most FPI studies used F. novicida because gene deletions were easier to make

(90-92).

The FPI encodes a genetically-distinct Type 6 secretion system (T6SS) that is important for phagosomal disruption, intramacrophage growth, and Francisella virulence

(93). T6SS is a contractile phage-like apparatus that extends across the inner membrane

(IM) and outer membrane (OM) of Gram-negative bacteria, to inject bacterial effector

17 proteins into host cells (94). Despite differences in gene composition among F. tularensis and canonical T6SS, high structural similarity of the F. novicida apparatus has been demonstrated (93). The F. novicida T6SS apparatus is composed of an outer sheath made up of IglA/IglB (F. novicida gene loci FTN1324/FTN1323) heterodimers that are stacked upon one another in a helical configuration (95), a rigid inner tube that is suspected to be composed of IglC (FTN1322), and a needle-like tip structure made up of

VgrG (FTN1312) and PdpA (FTN1309) (canonical tip structures only contain VgrG)

(96). Importantly, Type 6 secretion has not been observed in virulent F. tularensis subsp.

Most of the FPI genes/proteins have been demonstrated to be important for intracellular growth or phagosomal escape, highlighting that F. tularensis has evolved several mechanisms to increase its efficiency as an intracellular pathogen. For example:

IglC is required for LVS and F. novicida phagosomal escape, induction of in

J774A.1 cells, and virulence in mice (90, 97, 98) (99); IglA is needed for F. novicida intramacrophage growth in J774A.1 cells and virulence in chicken embryos (92); PdpA is essential for F. novicida virulence in mice and growth in murine bone marrow-derived macrophages (mBMDMs) (89, 100); PdpB (FTN1310), PdpC (FTN1319), and PdpD

(FTN1325) are important for F. novicida phagosomal disruption and escape into the cytosol of mBMDMs (101); IglD (FTN1321) is required for cytosolic replication in human monocyte-derived macrophages (hMDMs) (102); and DotU (FTN1316) and VgrG

(FTN1312) are required for LVS phagosomal disruption and intramacrophage survival in

J774A.1 cells and virulence in C57BL/6 mice (50). Finally, IglE (SchuS4 gene loci

FTT1346 and FTT1701), an OM-localized lipoprotein, is required for SchuS4 escape from the phagosome and replication in the cytosol of mBMDMs, and a deletion mutant of

18 4 both copies of SchuS4 IglE was highly attenuated (LD50 > 3.5x10 CFU) in an intranasal mouse model of infection (103). All the F. novicida FPI genes, except pdpE (FTN1320) and anmK (FTN1326), are required for virulence in mice. However, similar studies need to be performed in F. tularensis, utilizing double mutants, to confirm roles in virulence.

1.2.2 Additional Francisella Secretion Systems

F. tularensis also encodes a Type 1 secretion system (T1SS) that is important for bacterial virulence (104). The T1SS, composed of only three proteins, an ATP-binding cassette (ABC) transporter, a membrane fusion protein, and an outer membrane protein, spans from the cytoplasm, across the envelope, to the extracellular environment (105).

Many Gram-negative bacteria use the T1SS to secrete virulence factors, such as , and as drug efflux pumps, to confer resistance to antibiotics, detergents, and dyes (106).

The most well-characterized T1SS is the system (Hly) in E. coli which includes TolC as the outer membrane component (105, 107). F. tularensis encodes two

TolC orthologs, tolC (FTT1724 or LVS gene loci FTL1865) and ftlC (FTT1095,

FTL1107), both of which are outer membrane-localized and are involved in multidrug resistance (108, 109). Importantly, only tolC, but not ftlC, appeared to play a role in LVS and SchuS4 virulence in mice (109, 110), was required for LVS to delay macrophage cell death and induction of apoptosis (111), and was involved in suppressing host immune responses (112). In contrast, a recent study demonstrated that deletion of ftlC from

SchuS4 resulted in partial attenuation in mice (113). Finally, a third tolC homolog, silC

(FTT1258, FTL0686), was recently demonstrated to contribute to multidrug resistance,

19 resistance to oxidants and silver, and LVS virulence in mice and RAW264.7 macrophages (114).

1.2.3 F. tularensis Capsule

Gram-negative bacterial capsules are large structures, often composed of , associated with the OM, that surround the bacteria and protect them from immune recognition and complement-mediated (104, 115). In 1977, a crude

F. tularensis capsule preparation was found to be composed of mannose and rhamnose

(116). The capsular material, further characterized in 2010, has a molecular weight between 100 and 250 kDa, is composed of a tetrasaccharide repeat that is identical to the

O-antigen subunit of LPS, and was detected in 14 different Type A and Type B strains

(117). Importantly, the purified capsule did not contain other materials characteristic of the LPS, such as keto-3-deoxyoctulsonic acid (KDO) or (117). Due to the high molecular weight and composition of the F. tularensis capsule, it is designated as a

Group 4 capsule, along with several other Gram-negative species including cholerae, enterica, and coli (115).

An acapsular mutant of LVS was demonstrated to be more sensitive to complement-mediated killing and IgM opsonization than wild-type LVS (118), bound more complement component C3 than LVS, and was unable to replicate in mice (119)

(120). Additionally, SchuS4 acapsular mutants were found to be serum sensitive and attenuated for intramacrophage growth (121). Thus, capsule is important for F. tularensis virulence due to its functions in preventing complement-mediated lysis and antibody opsonization.

20 1.2.4 F. tularensis LPS

LPS makes up the outer leaflet of the outer membrane of Gram-negative bacteria, serving as a permeability barrier, and is composed of lipid A, core oligosaccharide, and the O-antigen (122). LPS is highly immunogenic, particularly through recognition by

Toll-like receptor 4 (TLR4), stimulating production of proinflammatory cytokines and leading to bacterial clearance (123). However, many Gram-negative pathogens, including pestis, pylori, and F. tularensis, have developed lipid A modifications that prevent recognition of LPS and induction of innate immune responses

(123). In contrast to the prototypic lipid A from E. coli, which has six acyl chains (hexa- acylated) that are 12 to 14 carbons in length, and is phosphorylated at the 1’ and 4’ positions, F. tularensis lipid A has four acyl chains (tetra-acylated) that are 16 to 18 carbons in length, and has an α-linked galactosamine at the 1’ position and no phosphate at the 4’ position (124). Additionally, up to 95% of Francisella lipid A is present in a free form, lacking KDO, the core oligosaccharide, and the O-antigen (124, 125). These modifications to F. tularensis lipid A lead to reduced stimulation of TLR4, allowing for the bacteria to evade immune recognition and clearance (126).

As expected, of genes/proteins involved in lipid A modification result in

Francisella attenuation. Experimental deletion of naxD, preventing addition of galactosamine to lipid A, resulted in F. novicida and F. tularensis SchuS4 attenuation in mice and macrophages (127). Mutation of flmF2, also preventing modification of lipid A with galactosamine, attenuated F. novicida in pulmonary and subcutaneous mouse models of infection (128). Finally, an F. novicida lpxD1 mutant, deficient in acyl chain

21 additions to lipid A, also was avirulent in mice (129). In summary, the atypical F. tularensis LPS plays an important role in infection and immune evasion.

1.2.5 F. tularensis Envelope Proteins

Because of the designation of F. tularensis as a Tier 1 Select Agent, many researchers have focused on identifying and characterizing novel F. tularensis virulence factors. The F. tularensis envelope, consisting of the inner membrane, , and the outer membrane, contains many proteins that have been demonstrated to be required for virulence (104). F. tularensis envelope proteins that have been previously studied are described in the following sections, based on their subcellular compartment.

1.2.5.1 Outer Membrane Proteins

Bacterial OMPs often serve as virulence factors of and can be involved with a number of important processes including host cell invasion and immune evasion (104, 130-132). Outer membrane protein A (OmpA) is one of the most well characterized OMPs among Gram-negative bacteria and functions in adhesion, serum resistance, host cell invasion, peptide resistance, and host cell activation

(133). OMPs also play very important roles as immunostimulatory antigens in a number of vaccines (78, 108, 131, 134). The surface localization of OMPs highlights their importance in directly interacting with host cells, playing important roles in various pathogenic processes, and serving as immunostimulatory molecules to promote . As such, many studies have aimed to identify and characterize bacterial

OMPs.

22 Over 20 different OMPs have been identified in F. tularensis and many have been shown to be virulence determinants (104). FslE (FTT0025c, FTL1863), an OM receptor involved in iron uptake, is unique to Francisella, and has been demonstrated to be essential for growth of SchuS4 in iron-limited media (135). A second

OMP involved in iron acquisition, FupA (FTT0918, FTL0439), functions in uptake of siderophore-bound iron and ferrous iron (siderophore-independent), and is required for

SchuS4 virulence in mice and in J774A.1 macrophages (136-138). While a SchuS4

ΔfupA mutant was only partially attenuated (60% survival) in mice, a SchuS4

ΔfslEΔfupA mutant was completely attenuated, indicating that these proteins play synergistic roles in iron acquisition (138).

Global transcriptional profiling of SchuS4-infected mBMDMs and mutagenesis studies identified DipA (FTT0369c, FTL1306), a surface-exposed protein, as an important determinant for intracellular replication and virulence in mice (48). Deletion of dipA resulted in autophagic capture of replication-deficient bacteria in the macrophage cytosol (139). DipA was also shown to bind to another surface-exposed protein, FopA

(FTT0583, FTL1306), which contains an OmpA domain. Although the specific functions of DipA and FopA are not clear, it has been hypothesized that both proteins form a membrane complex that is essential for intracellular replication and survival (139).

One of the major reasons that F. tularensis is such a successful intracellular pathogen is its impressive ability to suppress host immune responses, including preventing activation of macrophages and DCs, as well as preventing secretion of the proinflammatory cytokines TNF-α, IL-12, and IL-1β (57). One study identified an

OmpA-like protein of F. tularensis (FTT0831c, FTL0325) that was required for LVS and

23 SchuS4 intracellular survival in mBMDMs. Deletion mutants of these orthologs,

ΔFTT0831c and ΔFTL0325, stimulated high levels of TNF-α production and increased

NFκB signaling in macrophages (140). ΔFTL0325 was later shown to be completely attenuated in a mouse intranasal infection model, highlighting the importance of immune suppression in F. tularensis pathogenesis (141).

Other F. tularensis OMPs that have roles in virulence include: FsaP (FTT0119,

FTL1658), which is required for LVS adherence to lung epithelial cells (142); Flpp3

(FTT1416c, FTL0645), which was found to be required for LVS survival in mouse lungs

(143); FmvB (FTT0602c, FTL0867c), which contributes to LVS virulence in mice (118); and IglE and DsbA, which are discussed below. Thus, OMPs are important mediators of

F. tularensis intracellular survival, replication, and pathogenesis.

1.2.5.2 Periplasmic Proteins

The OM and IM layers of Gram-negative bacterial cell envelopes are separated by the periplasmic space. The periplasm has several functions, including protein folding and transport, protection of the cytoplasm from toxins, regulation of PG synthesis and recycling, regulation of , and responses to envelope stressors and protein damaging molecules (144). Despite the importance of the above listed functions, very few periplasmic proteins have been identified and studied in F. tularensis. As an intracellular pathogen, F. tularensis is exposed to the harsh environment of phagocytes and, in order to survive, need mechanisms to overcome host antibacterial mechanisms such as reactive oxygen species (ROS) and reactive nitrogen species (RNS). A CuZn- containing superoxide dismutase C, SodC (FTT0879, FTL0380), was demonstrated to protect F. tularensis from ROS but not RNS, and was required for intracellular survival

24 in IFN-γ stimulated macrophages, but only played a minor role in virulence of mice

(145). Recently, a predicted periplasmic protein with homology to penicillin-binding proteins, DacD (FTT1029, FTL1060), was shown to be necessary for F. tularensis virulence in mice (146, 147). Additionally, a mutant of DacD was more sensitive to high salt (3% NaCl), low pH (4.0), and high temperature (42°C), but was more resistant to oxidative stress induced by CuCl2 (147).

1.2.5.3 Inner Membrane Proteins

Bacterial inner membrane proteins typically have roles in energy production, lipid biosynthesis, and protein secretion although, similar to periplasmic proteins, not many have been studied in F. tularensis (104, 148). QseBC is a bacterial two-component system, where QseB is the response regulator and QseC is the inner membrane-bound sensor kinase that functions as an adrenergic receptor to recognize bacterial autoinducers

(quorum-sensing molecules), as well as eukaryotic epinephrine and norepinephrine (149).

Francisella QseC (FTT0094c, FTL1762) was determined to be essential for LVS survival in macrophages and for production of the high molecular mass O- of LPS

(150). In addition, QseC was found to be upregulated in SchuS4, and a small molecule inhibitor of QseC, LED209, attenuated intracellular survival of SchuS4 in an i.n. mouse model of infection (151). Another two-component system, KdpDE, which has been shown to be important for regulating potassium, responding to bacterial stressors, and playing roles in quorum sensing in other bacteria, also has been implicated in Francisella virulence (152). Whereas a transposon mutant of F. novicida KdpD was demonstrated to be attenuated in a mouse model of infection (153), KdpD exists as a truncated

25 pseudogene in SchuS4 and LVS, indicating that it is not required for virulence in the human pathogenic F. tularensis strains. The exact role of KdpDE in potassium transport, osmolarity, or stress response in F. tularensis is unknown.

RipA (FTT0181c, FTL1914) was identified as a virulence factor of F. tularensis

LVS from a transposon mutant library, demonstrating an inability to replicate in alveolar epithelial cells (154). ripA mutants escape from the phagosome of J774A.1 macrophages but cannot replicate in the cytosol, are attenuated for replication in mouse lungs (154), and stimulate increased production of pro-inflammatory cytokines IL-1β, IL-18, and

TNF-α (155). FvfA (FTT0924, FTL1286), a protein of unknown function that has no known homology to proteins outside the Francisella genus, localized to the outer leaflet of the inner membrane (facing the periplasm), was reported to be required for intracellular growth of LVS and SchuS4 in J774A.1 macrophages, and for virulence of

SchuS4 in mice (156). Additionally, a SchuS4 FvfA mutant, ΔFTT0924, had increased sensitivity to β-lactam antibiotics and osmotic stress, indicating envelope permeability defects (156). The authors of that study proposed that FvfA plays a role in maintaining the integrity of PG, though no follow up studies have confirmed this role.

26

1.2.6 Disulfide Bond Formation Protein A and Substrates

Prototypic disulfide bond formation protein A (DsbA) is a periplasmic protein involved in forming disulfide bonds between consecutive cysteine residues in newly formed proteins as they translocate through the inner membrane, allowing for correct protein folding (157-159). DsbA functions as an oxidoreductase with a thioredoxin active site CXXC motif (160, 161). DsbA is an extremely powerful oxidant and, once the

CXXC motif cysteines become oxidized, is very unstable and readily transfers the disulfide bond to an unfolded protein, causing subsequent reduction of DsbA (162). An inner membrane-bound protein, DsbB, is responsible for re-oxidizing DsbA (163). Two additional proteins, DsbC and DsbD, are required to recognize and reduce mismatched disulfide bonds and form correct disulfide bond patterns, which are required for proper protein folding and function (164-166).

F. tularensis DsbA (FTT1103, FTL1096) was first identified from a SchuS4 transposon mutant library screen identifying intracellular replication-deficient mutants

(167). Deletion of FTT1103 resulted in an inability of the mutant to escape from the phagosome and replicate in J774A.1 macrophages, and attenuation in mice (168). As opposed to prototypic periplasmic DsbA (108), F. tularensis DsbA is unusual in that it is

OM-localized and is a bifunctional enzyme, possessing oxidoreductase and isomerase activities (169, 170).

Although several studies have identified F. tularensis DsbA as an essential virulence factor (168, 169, 171, 172), its disulfide bond formation activity suggests more of a housekeeping role in maintaining protein structure and function. Our lab previously

27 developed a molecular trapping assay in which we identified over 50 substrates of DsbA which we hypothesized to be virulence factors of F. tularensis (170). Several of the identified DsbA substrates already were established virulence factors of F. tularensis but the majority of the identified DsbA substrates were hypothetical (unstudied) proteins.

Our lab is currently studying the function of several of these substrates and determining their roles in virulence.

1.3 F. tularensis Vaccines

1.3.1 History of Tularemia Vaccines

F. tularensis LVS was first developed in the former Soviet Union between the years of 1934 and 1944 (173). The original vaccine strain, named “Moscow,” was developed by repeated sub-culturing of F. tularensis subsp. holarctica on media containing F. tularensis antiserum or by drying. This strain was tested in human volunteers and was shown to be protective up to six months after immunization, with no adverse side effects (173, 174). However, after immunization of several thousand people, this strain was lost. New attenuated strains were prepared and, in 1956, strains 15 and

155 were sent from the Gamaleia Institute in Russia to the Medical

Research Institute of Infectious Disease (USAMRIID) (174). One of these strains was selected and serial-passaged to achieve further attenuation and safety and was then named the LVS. Scarification, the chosen method for administration of LVS, is performed by first reconstituting dried LVS with water, cleaning the shoulder of the person with alcohol, administering 3-4 drops of LVS suspension onto the skin, making several small scratches through the drops, and, finally rubbing the vaccine into the scratches (174).

28 A complicating factor of LVS is the appearance of two colony phenotypes, known as blue-gray variants. The blue colony variant appears to be immunogenic, while the gray colony variant does not (174). Colony variants also have been observed in F. tularensis subsp. tularensis, where more-virulent strains appeared as smooth, blue colonies and less-virulent strains appeared as rough (non-smooth), gray, or buff colonies.

The strain name SchuS4 arose from a 1951 study identifying 4 smooth phase variants of the F. tularensis isolate from a human ulcer, Schu – SchuS4 and SchuS3 are smooth blue colonies whereas SchuS1 and SchuS2 are smooth buff colonies (175). More recent studies have demonstrated that differences in immunogenic properties of the blue-gray variants are in part due to the lipid A structure of LPS. In comparison to the blue colony variant which did not induce nitric oxide (NO) from rat macrophages, the gray colony variant (LVSG) induced high levels of NO, causing attenuation in macrophages (176).

Additionally, in comparison to WT LVS, LVSG was demonstrated to contain less LPS

O-antigen and did not have a galactosamine modification on its lipid A (177). Due to their higher immunogenicity and protective capacities, the smooth, blue colony phenotypes are used in developing vaccines.

In a series of studies in the 1960s, known as “Operation Whitecoat,” the protective capacity of LVS immunization in humans was investigated. One study aerogenically-immunized prison “volunteers” with either 104, 106, or 108 CFU LVS and reported that, while only 30% of 104-immunized individuals experienced mild symptoms,

90% of individuals immunized with 108 CFU experienced mild typhoidal tularemia (27).

Groups of men that received the higher immunization doses were challenged with

SchuS4 at either 2 months or 14 months post-immunization, with 73% and 50%,

29 respectively, of those vaccinees developing disease that had to be treated with antibiotics

(27). A second study immunized inmates from the Ohio State Penitentiary with either the killed Foshay vaccine or LVS (109 CFU) and subsequently exposed them to aerosolized

SchuS4 (10-50 CFU) (178). In comparison to the 75% of non-vaccinated controls who developed tularemia, 57% of individuals immunized with the Foshay vaccine developed disease, and only 17% of LVS-immunized men experienced tularemia symptoms (178).

While the above LVS immunization studies have demonstrated partial protection against virulent F. tularensis, LVS is not a licensed vaccine in the United States due to safety and efficacy concerns, including an incomplete understanding of its attenuation, induction of mild disease in healthy individuals, and incomplete protection against challenge. These findings highlight the need for more effective vaccines and defined correlates of protection.

1.3.2 Animal Models

When developing and evaluating novel vaccines for efficacy, it is important to consider the model used, as the outcomes of immunization are dependent on multiple variables including F. tularensis strain, dose and route of challenge, and host genetic background (179). Mouse models of F. tularensis infection are routinely used because the reliability of inbred mouse strains, the large number of immunologic tools available to assess mouse immune responses, and mouse infections mimic human infections – mice can be infected through multiple routes, with bacterial dissemination from the site of infection to lungs, liver, spleen, and lymph nodes, regardless of the infection route.

However, it should be noted that there is a discrepancy between humans and mice in

30 lethality of various strains. When given intradermally, LVS establishes a sublethal infection in mice. However, when LVS is administered in low doses by any other route

(e.g., i.v., intraperitoneal [i.p], or respiratory), it can kill mice (60). Importantly, Type A and B F. tularensis, as well as F. novicida, kill mice rapidly when administered by any route. As noted above, Type B infections, although debilitating, are rarely fatal in humans (179). However, death due to Type B F. tularensis has been reported in several cases (180, 181). Conversely, human F. novicida infections are extremely rare and occur only in immunocompromised individuals.

Subcutaneous (s.c.) or i.d. routes of infection or immunization often are used to

6 study vaccine efficacy due to the relatively high LD50 for LVS (LD50>10 ) (179) and the high likelihood that human vaccinations would require similar immunization routes.

3 Intranasal (i.n.) and aerosol routes of infection are much more lethal in mice (LD50<10

CFU for LVS, LD50<10 CFU for SchuS4) and are often used in protection studies because aerosolized F. tularensis is most likely to be used during a attack. F. tularensis can also be given orally to simulate infection by ingestion of

6 contaminated food or water and has an LD50 of approx. 10 CFU for virulent Type A F.

9 tularensis (182). Additionally, LVS oral administration has an LD50 > 10 CFU and has been demonstrated to be protective against F. tularensis Type A challenge (183).

Commonly-used inbred mouse strains for F. tularensis studies are BALB/c,

C57BL/6, and C3H/HeN (184). Intranasal LVS immunization conferred full protection

(100% survival) against aerosol challenge with 200 CFU fully-virulent Type A F. tularensis in BALB/c, but not C57BL/6 mice (100% death by day 12) (185).

Furthermore, BALB/c mice were better protected against respiratory challenge with Type

31 A F. tularensis than C57BL/6 mice after intradermal immunization with an F. tularensis

ΔclpB mutant (186). Conversely, C3H/HeN mice were demonstrated to be better protected than BALB/c mice after i.d. LVS vaccination and subsequent SchuS4 aerosol challenge (187). However, there are a number of conflicting reports about mouse strain susceptibilities to F. tularensis and/or abilities to mount protective immune responses, as one study reported that BALB/c mice immunized i.d. with a SchuS4 ΔclpB mutant were better protected against fully virulent SchuS4 than C3H/HeN mice. In contrast, a second study noted that C3H/HeN mice orally immunized with SchuS4 ΔclpB were better protected against SchuS4 challenge than BALB/c mice immunized the same way (188).

Finally, Fischer 344 rats also have been studied as an immunization and infection model for tularemia because they are less susceptible to LVS and other Type B infections than mice, better mimicking the infectious doses of humans (189).

1.3.3 Immune Correlates of Protection

Immune correlates of protection are specific, measurable indicators that can be used to determine how effective a new vaccine is at inducing a protective immune response. Correlates of protection are especially important for highly-virulent intracellular pathogens, such as F. tularensis, due to the challenges involved with human studies. Until recently, no correlates of protection had been defined for F. tularensis

(190). Fortunately, the FDA “animal rule” allows for efficacy testing of potential human products to be performed and assessed in animals, while safety testing still requires human clinical trials (191).

32 A macrophage-T cell co-culture method recently was developed to determine the correlates of protection for clearing F. tularensis infections. This co-culture method involves overlaying splenocytes from naïve or vaccinated C57BL/6 mice on macrophages infected with F. tularensis LVS (192). Immune cell gene expression was analyzed by qRT-PCR and 16 of 84 total genes examined showed significant correlation with survival against SchuS4 challenge and reductions in LVS numbers in macrophages, including

IFN-γ, IL-6, IL-12rβ2, T-bet, IL-18bp, SOCS-1, GM-CSF, TNF-α, IRF-1, IL-22, IL-27, and CCL7 (192). A similar study was performed in Fischer 344 rats using peripheral blood lymphocytes (PBLs) and splenocytes, and similar results were obtained as in mouse studies (190). Additional genes upregulated in rat PBLs included NOS2, IL-21,

CCL5, LTA, FasL, CXCL9, CCR3, CCL4, and IL-1RA, suggesting that these molecules also may be correlated with protection (190). Several strains of LVS, conferring differing levels of protection against lethal challenge, were used in these studies so that expression levels of the above noted genes could be correlated with good or poor protection. The results from these studies provide a standard for which newly-developed vaccines can be compared to determine relative efficacy.

33

1.3.4 Subunit, Killed, Whole-Cell, or Live-Attenuated?

Subunit vaccines are cell-free and are composed of purified microbial antigens

(193). Purified F. tularensis LPS has been demonstrated to protect against subsequent challenge with LVS and, although serum transfer studies indicate that LPS-induced protection appears to be primarily mediated by a humoral response, cellular immune responses also are important, as IFN-γ-, CD4-, and CD8-deficient mice were not protected against LVS challenge (76, 194, 195). Conversely, immunization with purified

LPS did not protect against SchuS4 challenge (76). Mice immunized with F. tularensis outer membrane protein preparations were partially protected against pulmonary SchuS4 challenge (50% survival) and had increased time-to-death (78). Yet another study tested the ability of F. tularensis LVS membranes (inner and outer) and poly (I:C), encapsulated into PLGA nanoparticles, to protect against i.n. LVS challenge, observing a 50% increased survival rate over naïve mice (196). However, that same membrane-poly (I:C)-

PLGA nanoparticle immunization did not protect against SchuS4 challenge.

The Foshay vaccine, produced approx. 80 years ago, was the first attempt at a killed whole-cell vaccine for tularemia (184). Several formulations of the Foshay vaccine were tested, including heat-killed, phenol-treated, and acetone-treated. However, all formulations conferred minimal protection in laboratory workers and hunters (26,

197). Furthermore, although the Foshay vaccine protected non-human primates against

SchuS4 challenge, the animals experienced local necrotic lesions and lymphadenopathy

(198). A direct comparison of vaccination of guinea pigs with the live blue colony variant of LVS, heat-killed blue colony variant, phenol-killed blue colony variant, and the

34 Foshay vaccine revealed significant differences in protection by day 10 post-challenge with SchuS4 – 100% survival of live blue colony variant-immunized, 30% survival of

Foshay vaccine-immunized, and 0% survival of heat-killed or phenol-killed blue colony variant-immunized (26). Because of the promising results from this study, the development and testing of new and modified live attenuated vaccines continues to this day.

LVS has been classified as an investigational new drug in the U.S. for many years and is the most extensively studied live attenuated vaccine candidate. A 2008 study evaluated the safety of a new LVS formulation in rabbits, demonstrating that no adverse effects were observed after subcutaneous, percutaneous (scarification), or intradermal immunizations of up to 1x109 CFU, and that immunization induced high levels of IgM,

IgG, and IgA (199). Unfortunately, T cell responses were not examined in those immunized rabbits. A follow-up phase I clinical trial evaluated the safety and efficacy in human volunteers, reporting minor adverse reactions and noting that scarification was the best vaccination method for production of antibodies and IFN-γ (75). Despite these studies, LVS still is not a licensed vaccine in the U.S. due to the safety and efficacy concerns highlighted above. As such, many researchers have focused on developing new live attenuated vaccines with at least equal or better protection than LVS.

35

1.3.5 Live-Attenuated Vaccines Derived from LVS vs. SchuS4

Live attenuated vaccines are developed by deleting or mutating a gene(s) of an immunogenic bacterial strain. Many live attenuated vaccines have been developed in F. tularensis LVS, since that strain can be manipulated under BSL2 conditions. If those

LVS-based vaccines are shown to be attenuated and protective against SchuS4 challenge, researchers often test the homologous SchuS4 mutant for attenuation and protective capabilities. However, deletions of homologous genes in LVS and SchuS4 do not always result in the same phenotype (200).

A recent review outlined all of the major F. tularensis live attenuated vaccines that have been developed, including 19 LVS mutants and 20 SchuS4 mutants, where genes involved in metabolism, oxidative stress response, heat response, or capsule and membrane structure had been deleted (200). Depending on the immunization route and dose, mouse strain used, immunization regimen (prime only or prime and boost), and

SchuS4 challenge dose, live attenuated vaccine efficacies have varied greatly. For example, an LVS ΔpurMCD mutant, deficient in purine biosynthesis, administered as a prime only, conferred no protection to BALB/c mice against 100 CFU SchuS4 i.n. challenge. By comparison, a prime-boost regimen of ΔpurMCD conferred 100% protection against SchuS4 challenge (200, 201). Interestingly, prime and boost immunization with the homologous SchuS4 mutant only conferred 71% protection (201), demonstrating that protection and/or attenuation are not always equivalent in LVS and

SchuS4 mutants. Many immunization and challenge studies have used a SchuS4 challenge dose of 100 CFU, as this represents approximately 10× the LD50 of SchuS4.

36 Notably, the LVS ΔclpB mutant was demonstrated to provide protection against 1000

CFU of SchuS4 (100× the LD50) (87, 200).

Other notable LVS-derived vaccines, ΔcapB, ΔFTL0057, ΔFTL0325, and

ΔFTL0291, and SchuS4-derived vaccines, ΔclpB and ΔfipB, all demonstrated 100% protection against SchuS4 i.n. challenge (141, 168, 200, 202, 203). For a new vaccine to be considered for use in humans, it must be safer than, and at least as effective as, LVS.

While it is relatively easy to generate live attenuated vaccines that are safer than LVS, it is much more difficult to control the balance between attenuation and protection.

Concerns for live attenuated vaccines include hypoattenuation, being too virulent to use, and hyperattenuation, inducing poor immunogenic responses. The number of live attenuated vaccines that meet these criteria include LVS ΔcapB (capsule biosynthesis),

LVS ΔpurMCD (purine biosynthesis), LVS ΔsodBFt (superoxide dismutase), LVS:wzy

(O-polysaccharide synthesis), LVS ΔcapB/IglABC (capsule biosynthesis mutant that overexpresses FPI genes), SchuS4 ΔclpB (heat shock protein), and SchuS4

ΔFTT0918ΔcapB (200). However, to achieve sufficient protection against virulent F. tularensis aerosol challenge, these vaccines often must be administered to mice or rats intranasally, a route unlikely to be used in humans. In addition, long-term protection has not been evaluated with many of these vaccine candidates.

37

1.4 Gram-Negative Bacterial Peptidoglycan

1.4.1 Peptidoglycan General Structure

PG (i.e., murein) is an essential component of the cell wall of most bacteria, playing roles in virulence, cell division, maintenance of cell morphology and internal turgor pressure, and serving as an anchor for envelope proteins (204, 205). Gram- negative PG is located in the periplasmic space and serves as a scaffold for covalent OM attachment through Braun’s lipoprotein and noncovalent attachment through Pal and other lipoproteins (206-209). PG is composed of alternating glycan strands, N- acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc), crosslinked by peptide stems. In Gram-negative bacteria, the peptide stem is usually composed of L-

Alanine-γ-D-Glutamine-meso-2,6-diaminopimelic acid (meso-A2pm)-D-Alanine-D-

Alanine, where L-Alanine is attached to the MurNAc residues, and the majority of peptide stems are crosslinked by a peptide bridge between the fourth peptide residue, D-

Alanine, of one stem, and the third peptide residue, meso-A2pm, of another stem (4-3 crosslinks) (204, 210, 211).

Many successful bacterial pathogens have evolved mechanisms to modify the basic PG structure so as to avoid being recognized by host immune mechanisms and to prevent targeting by PG-degrading enzymes such as glycan O-acetylation, N- deacetylation, and N-glycolylation (205, 212, 213). In addition to PG remodeling, Gram- negative bacteria, especially intracellular pathogens, must continuously recycle and synthesize new PG to repair any damage by external insults including ROS, RNS,

38 lysozyme, and NO (214, 215). There are more than 50 proteins, mostly characterized in

E. coli, involved in the tight regulation, synthesis, and recycling of PG (215).

1.4.2 Peptidoglycan Synthesis

PG synthesis begins in the cytoplasm when fructose-6-phosphate, a product of PG recycling, is converted to uridine-diphosphate (UDP)-GlcNAc through a series of reactions catalyzed by GlmS, GlmM, and GlmU (216). UDP-GlcNAc is converted to

UDP-MurNAc by MurA and MurB (217), and several other Mur ligases are then involved in sequential addition of L- and D-amino acids to form the pentapeptide chain of

PG: L-Ala (MurC), D-Glu (MurD), meso-A2pm (MurE), and D-Ala-D-Ala (MurF) (218).

MraY attaches UDP-MurNAc-pentapeptide to a lipid carrier known as undecaprenyl phosphate, to form the molecule lipid I (219), which is then transferred to the IM (220), where MurG attaches UDP-GlcNAc to the lipid I molecule, forming lipid II (221). Lipid

II is flipped into the periplasm by FtsW (222) and MurJ (223), where the MurNAc ends can then be added to GlcNAc of existing PG strands by the transglycosylase domain of high molecular mass penicillin binding proteins (HMM PBPs), and cross-linked to adjacent peptide stems by the transpeptidase domain of HMM PBPs (215). Cross-linkage of newly formed PG, containing pentapeptide stems, to existing tripeptide stems, results in release of the terminal D-Ala, forming tetrapeptide stems (224). Low molecular mass penicillin binding proteins (LMM PBPs) also have D,D-carboxypeptidase activity, which allows them to remove the terminal D-Ala of pentapeptide stems, thus reducing the amount of potential cross links as transpeptidases can no longer act on the resulting tetrapeptide (225). As synthesis is occurring, older layers of PG are constantly being

39 recycled (up to 50% per generation) to provide the substrates for synthesis. Finally, because PG is so dynamic, but also tightly regulated to prevent bacterial cell death, many of the PG synthesis enzymes are effective antibacterial targets (226). For instance, β- lactam antibiotics mimic the D-Ala-D-Ala end of pentapeptide stems, bind PBPs, and prevent cross-linking of adjacent peptide stems (225).

1.4.3 Peptidoglycan Recycling

PG recycling is essential for maintaining cell wall homeostasis as it is involved in providing the substrates for synthesis of new PG and removing damaged PG. As opposed to PG synthesis, which begins in the cytoplasm, PG recycling begins in the periplasm with the action of lytic transglycosylases (LT) (176). LTs act on glycan strands with the same specificity as lysozyme, cleaving the β-1,4-glycosidic linkage between MurNAc and GlcNAc to produce a GlcNAc-1,6-anhydro-MurNAc, which allows for either the insertion of new muropeptides, flagella, or the formation of secretion systems (227). Additional periplasmic enzymes, LMM PBPs, have endopeptidase activity, allowing them to hydrolyze cross links between adjacent peptide stems, and/or

D,D-carboxypeptidase activity, allowing them to cleave off the terminal D-Ala of peptide stems, resulting in tetrapeptide stems (228). A final class of periplasmic-localized PG hydrolases are the amidases, which are responsible for releasing the tetrapeptide stems from the glycan strands (229). Inner membrane-associated AmpG permease facilitates the transport of free tetrapeptide stems, GlcNAc-1,6-anhydro-MurNAc or GlcNAc-1,6- anhydro-MurNAc-tetrapeptide, from the periplasm to the cytoplasm (230).

40 Once in the cytoplasm, a set of three enzymes are responsible for preparing the muropeptide breakdown products to rejoin the synthesis pathway. The NagZ enzyme separates GlcNAc and MurNAc (231), AmpD cleaves any remaining peptides from

MurNAc residues (232), and L,D-carboxypeptidase A (LdcA) cleaves tetrapeptide stems into tripeptides (233). The ability of bacteria to recycle their PG is critical for energy conservation, cell wall homeostasis, and protection from external insults. Mutation of any one of the enzymes in the synthesis or recycling pathways can lead to altered cell morphology and permeability, cell division defects, or loss of virulence (233-236).

Although PG synthesis and recycling pathways have been well-characterized in E. coli, very little is known about these processes in intracellular pathogens such as F. tularensis.

Indeed, within the last year, a putative D,D-carboxypeptidase ortholog, DacD, was identified in F. tularensis and shown to be required for virulence in mice (146, 147). In addition, replication in macrophages and growth in the presence of stress stimuli (e.g., low pH, 42°C, and 3% NaCl) was dependent upon DacD (147). However, functional studies were not performed in either study to verify carboxypeptidase activity, highlighting the need for additional studies to better understand the role of PG modifying proteins in F. tularensis virulence.

1.5 Goals and Significance of my Dissertation Studies

Compared to prototypic E. coli DsbA which only has oxidoreductase activity and is located in the periplasm, previous studies in our lab identified an unusual F. tularensis

DsbA that possesses both oxidoreductase and isomerase activities, and is OM-localized

(170). Although work by our laboratory, and others, has demonstrated ΔdsbA mutants to

41 be fully-attenuated in mice, the disulfide bond formation activity of DsbA suggested that it may play more of a role in maintaining protein structure and function, and is not a “true virulence factor,” because it does not directly interact with the host (168-172). As such, we performed a molecular trapping assay and identified over 50 DsbA substrates, which we hypothesized may be F. tularensis virulence factors. In my dissertation project, I selected one of those previously uncharacterized proteins, F. tularensis FTL1678, based on bioinformatic analysis suggesting that it contains an L, D-carboxypeptidase (Ldc) domain and may be responsible for cleaving tetrapeptide stems of PG to tripeptide stems, aiding in PG recycling. Bioinformatic analyses also predicted that FTL1678 localizes to the periplasm, which is in contrast to prototypic cytoplasmic Ldcs. My dissertation research project was designed to characterize the role that FTL1678 plays in F. tularensis virulence and to determine the function of this hypothetical protein.

Many different PG synthesis and recycling enzymes, including Ldcs, are demonstrated to be required for virulence (205). In particular, E. coli Ldc mutants are unable to survive in stationary phase (237), and H. pylori Ldc mutants have defects in motility and biofilm formation (235, 238), and gonorrhoeae Ldc mutants are unable to stimulate host NOD-1 responses (239). Although

PG modification has been studied in those bacteria, very few studies have examined PG modification in pathogenic intracellular bacteria such as pneumophila,

Burkholderia pseudomallei, or F. tularensis. Given that intracellular pathogens are exposed to cell membrane- and cell wall damaging-molecules while inside the hostile environment of the host cell, they need to be able to quickly remodel and repair damaged envelope proteins and PG in order to survive. To date, only one putative PG

42 synthesis/recycling protein, DacD, has been identified and studied in F. tularensis, although no functional studies were performed to confirm PG synthesis/remodeling activity. The lack of knowledge about these processes highlights the need to identify and characterize PG-modifying enzymes in an effort to better understand F. tularensis pathogenesis. We hypothesized that FTL1678 is an L, D-carboxypeptidase that is required for efficient PG recycling and F. tularensis virulence. In addition, we hypothesized that the predicted periplasmic localization of FTL1678 is an efficiency mechanism to quickly and efficiently repair PG damage. The results from these studies not only provide new information to the F. tularensis field about PG synthesis/ remodeling but also provide insights into intracellular pathogenesis and PG recycling processes among diverse bacterial pathogens.

43 Chapter 2 A Periplasmic L,D-carboxypeptidase is Important for Cell Morphology, Membrane Integrity, and Virulence in Francisella tularensis

(PLoS Pathogens, manuscript under review)

Briana Zellner1, Dominique Mengin-Lecreulx2, William T. Gunning 3rd3, Robert Booth3,

Jason F. Huntley1*

1 Department of Medical and Immunology, University of Toledo, Toledo,

OH, U.S.A.; 2 Institute for Integrative Biology of the Cell (I2BC), CEA, CNRS, Univ.

Paris-Sud, Université Paris-Saclay, 91198 Gif-sur-Yvette, France; 3Department of

Pathology, University of Toledo, Toledo, OH, U.S.A.

*Correspondence: [email protected]

44

2.1 Abstract

F. tularensis is a Gram-negative, intracellular bacterium that causes the zoonotic disease tularemia. Intracellular pathogens, including F. tularensis, have evolved mechanisms to allow survival in the harsh environment of macrophages and neutrophils, where they are exposed to cell membrane-damaging molecules. One mechanism that protects intracellular Gram-negative bacteria from macrophage or neutrophil killing is the ability to recycle and repair damaged PG, – a process that requires over 50 different PG synthesis and recycling enzymes. In addition to PG repair, PG recycling occurs during cell division and plays critical roles in maintaining cell morphology, structure, and membrane integrity. Here, we identified a PG recycling enzyme, L,D-carboxypeptidase

A (LdcA), of F. tularensis that is responsible for converting PG tetrapeptide stems to tripeptide stems. Unlike prototypic LdcA homologs, F. tularensis LdcA does not localize to the cytoplasm and also exhibits L,D-endopeptidase activity, converting PG pentapeptide stems to tripeptide stems. Loss of F. tularensis LdcA led to altered cell morphology and membrane integrity, as well as attenuation in a mouse pulmonary infection model and in primary and immortalized macrophages. Finally, a F. tularensis

LdcA mutant protected mice against virulent Type A F. tularensis SchuS4 pulmonary challenge.

45

2.2 Introduction

The Gram-negative bacterial cell wall plays an important role in maintaining cell shape, protecting against external insults, and preventing cell lysis amid fluctuations in internal turgor pressure (215). The cell wall is composed of PG, a network of alternating

N-acetylglucosamine (GlcNAc) and N-acetylmuramic acid (MurNAc) glycan chains that are crosslinked through peptide stems, and lies just outside of the cytoplasmic membrane of most bacteria (225, 240). PG has been shown to be covalently attached to the outer membrane (OM) of Gram-negative bacteria by Braun’s lipoprotein, and noncovalently attached by Pal and other lipoproteins (206-209). A loss of membrane integrity can occur if interactions between PG and attached lipoproteins are disturbed, thus it is extremely important to maintain correct PG architecture.

Synthesis of Gram-negative PG precursors, studied mainly in E. coli, occurs in the bacterial cytoplasm and requires a series of enzymes to build a pentapeptide PG monomer before moving this molecule into the periplasm. Periplasmic PG cross-linking most often occurs between the fourth residue (D-Ala) of newly formed pentapeptide stems to the third residue (meso-A2pm) of existing tripeptide stems (4-3 cross links), resulting in the release of the pentapeptide terminal D-Ala and forming tetrapeptide stems (211, 241). As new PG is formed and incorporated, older layers of PG must be cleaved to maintain consistent sacculus thickness (242). Up to 60% of PG is recycled per generation, which repairs damaged PG and also provides energy during periods of stress or starvation (214, 215,

243). Together, there are more than 50 different enzymes, mostly cytoplasmic, involved in Gram-negative PG synthesis and recycling. PG homeostasis is extremely important, as

46 in any one of these enzymes can result in altered cell morphology, changes in cell membrane integrity, and reduced ability to replicate and divide (215). Importantly, a recent review highlighted the largely neglected field of Gram-negative bacterial PG synthesis and recycling, particularly given the growing threat of and association of PG homeostasis with virulence (205). For example, penicillin-binding protein (PBP) mutants were unable to survive in vitro or in vivo and were more sensitive to complement-mediated killing than wild-type bacteria (244) and PG hydrolase (AmiA) mutants were unable to colonize mouse stomach tissue (245).

PG recycling, mainly characterized in E. coli, begins with periplasmic lytic transglycosylase (LT) cleavage of the β-1,4-glycosidic linkage between MurNAc and

GlcNAc, forming a GlcNAc-1,6-anhydro-MurNAc product, which allows for either insertion of new muropeptides or recycling of GlcNAc-1,6-anhydro-MurNAc (227). Low molecular mass penicillin binding proteins (LMM PBPs) can function as endopeptidases, cleaving the cross-links between adjacent tetrapeptide stems, and/or as D,D- carboxypeptidases, removing the terminal D-Ala of pentapeptides during transpeptidation

(cross-linking) reactions, forming the tetrapeptide (215). Amidases then remove the tetrapeptide stems from GlcNAc-1,6-anhydro-MurNAc, the peptide stems and GlcNAc-

1,6-anhydro-MurNAc are transferred into the cytoplasm via inner membrane permeases, and cytoplasmic L,D-carboxypeptidases (Ldc) cleave the terminal D-Ala from the tetrapeptide, resulting in a tripeptide, and allowing for additional hydrolases to further degrade PG (215). PG is unusual in that it is both highly dynamic (e.g., allowing for bacterial division and molecular transport across the periplasm), yet tightly regulated to

47 prevent membrane collapse and bacterial death. As such, PG recycling enzymes have been speculated to be important virulence determinants in Gram-negative bacteria (205).

Indeed, E.coli Ldc mutants lyse in stationary phase (233) and are more susceptible to β- lactam antibiotics (246), Helicobacter and Campylobacter Ldc mutants have altered cell morphology and defects in motility and biofilm formation (234, 238, 247), and Ldc mutants are unable to stimulate NOD1-dependent responses in the host

(239). However, very little is known about the importance of PG recycling enzymes in the pathogenesis of intracellular pathogens such as Francisella tularensis.

F. tularensis, the causative agent of tularemia, is a Gram-negative, intracellular, coccobacillus that can infect and cause lethal disease in many species, including humans

(13, 29). There are three subspecies of F. tularensis, subsp. tularensis (Type A), subsp. holarctica (Type B), and subsp. mediasiatica, although only subsp. tularensis and subsp. holarctica are virulent for humans (19). F. tularensis poses a severe threat to public health and has been classified as an NIH Category A Priority Pathogen and a CDC Tier 1 Select

Agent due to its low infectious dose (<10 CFU), ease of aerosolization, and high morbidity and mortality rates (up to 60%) (4, 9). Like other intracellular pathogens, F. tularensis has evolved different mechanisms to infect, survive, and replicate within host cells, including macrophages and neutrophils (248). However, this lifestyle exposes the bacteria to reactive oxygen species (ROS), reactive nitrogen species (RNS), antimicrobial peptides, and other cell membrane- and cell wall-damaging molecules (249). Our group previously demonstrated that the F. tularensis disulfide bond formation protein A (DsbA) ortholog repairs damaged outer membrane proteins and known virulence factors. We additionally showed that F. tularensis DsbA, unlike periplasmic DsbA in E. coli and most other Gram-

48 negative bacteria, is outer membrane-bound and is a multifunctional protein with both oxidoreductase and isomerase activities. Finally, using a molecular trapping approach, we identified over 50 F. tularensis DsbA substrates, many of which we speculate are involved in virulence (170).

Here, we determined the function of one of those F. tularensis DsbA substrates – a previously unstudied hypothetical protein containing a putative LdcA domain – and assessed its role in bacterial virulence. Deletion of F. tularensis LdcA resulted in bacteria with altered cell morphology, increased sensitivity to β-lactam antibiotics, yet increased resistance to several stressors (e.g., H2O2, NaCl, low pH). Next, we demonstrated that F. tularensis LdcA exhibits L,D-carboxypeptidase and L,D- endopeptidase activities on pentapeptide and tetrapeptide residues of PG. Finally, we established that F. tularensis LdcA is required for virulence, as mutants were unable to replicate in macrophages or cause disease in mice.

2.3 Results

2.3.1 FTL1678 Contains a Putative L,D-carboxypeptidase Domain

Previous studies by our group and others have shown that F. tularensis DsbA mutants are attenuated in mice (168). However, additional work by our group, demonstrating that DsbA possesses both oxidoreductase and isomerase activities to repair damaged envelope and cell membrane proteins, indicated that other envelope proteins likely are responsible for F. tularensis virulence (170). To identify new F. tularensis virulence factors, we used a molecular trapping approach and identified over 50 F.

49 tularensis DsbA substrates (170). One of those DsbA substrates, FTL1678, is annotated in the F. tularensis genome as a conserved membrane hypothetical protein. Here, a conserved domain search revealed that a large portion of FTL1678 contains a putative Ldc domain, part of the peptidase_S66 superfamily (Fig. 2-S1). Ldc proteins have been studied in a number of Gram-negative bacteria, including E. coli (233, 237, 246, 250),

Pseudomonas aeruginosa (235), N. gonorrhoeae (239), and Campylobacter jejuni (238).

To further explore this conserved domain, amino acid sequences of FTL1678 (F. tularensis subsp. holarctica [Type B] LVS) and FTT0101 (homolog of FTL1678 in F. tularensis subsp. tularensis [Type A] SchuS4) were aligned with LdcA orthologs from E. coli, P. aeruginosa, N. gonorrhoeae, and C. jejuni (named Pgp2). Despite low percentages of amino acid identities among the LdcA orthologs (6.3% to 30.3%; Fig. 1), there was a higher degree of amino acid similarity among LdcA orthologs (13.0% [E. coli and C. jejuni] to

44.7% [E. coli and N. gonorrhoeae]; Fig. 2-1). Notably, the LdcA Ser-Glu-His catalytic triad, previously shown to be required for P. aeruginosa LdcA activity (235), was absent from C. jejuni Pgp2 but was present in all LdcA homologs, including FTL1678 and

FTT0101 (Fig. 2-1).

50

Figure 2-1. Amino acid alignment of bacterial L,D-carboxypeptidases. Clustal Omega amino acid alignment of E. coli LdcA (BAA36050.1), P. aeruginosa LdcA (Q9HTZ1), N. gonorrhoeae LdcA (YP_208343.1), F. tularensis Type B FTL1678, F. tularensis Type A FTT0101, and C. jejuni Pgp2 (WP_002856863). Percent identities (pid), compared to E. coli LdcA, are indicated. Black shading indicates similar residues. Red shading indicates the catalytic triad.

51 E. coli and P. aeruginosa Ldcs have been localized to the bacterial cytoplasm (233,

235). However, C. jejuni Pgp2 is unusual in that it contains a signal peptide and has been speculated to be periplasmic (238). In addition, N. gonorrhoeae LdcA was found to be periplasmic and outer membrane-associated (239). As noted above, we previously demonstrated that FTL1678 is a DsbA substrate (170), indicating that FTL1678 is located in the F. tularensis envelope (i.e., in the inner membrane [IM], periplasm, or outer membrane [OM]). Bioinformatic analyses of FTL1678 indicated that it is a periplasmic protein due to the presence of a signal peptide but absence of OM or lipoprotein signatures

(Table 2.S1). To experimentally confirm FTL1678 localization, we generated an F. tularensis strain with 6 histidine-tagged FTL1678, then performed spheroplasting, osmotic lysis, and sucrose density gradient centrifugation to separate IM and OM fractions and probe for protein subcellular localization. Immunoblotting of whole-cell lysates

(WCL), OM fractions, and IM fractions demonstrated that the OM control protein, FopA

(108), only was present in WCL and OM fractions (but not IM fractions; Fig. 2-2) and the

IM control protein, SecY (108), only was present in WCL and IM fractions (but not OM fractions; Fig. 2-2). By comparison, FTL1678 only was detected in WCL and OM fractions, demonstrating OM-association and indicating that FTL1678 may be periplasmic.

To further explore this possibility, we examined localization of the periplasmic protein

TolB, which is well-known in Gram-negative bacteria to bind PG with the peptidoglycan associated lipoprotein, Pal (251, 252). Notably, we previously demonstrated that F. tularensis Pal is OM-localized (108) , similar to other Gram-negative bacteria. Here, F. tularensis TolB was detected in OM fractions, but not IM fractions, confirming its OM-

52 association and providing further evidence that FTL1678 is OM-associated and may be a periplasmic protein (Fig. 2-S2).

Figure 2-2. FTL1678 is OM-associated. Spheroplasting, osmotic lysis, and sucrose density gradient centrifugation were performed to separate inner membranes (IM) and outer membranes (OM) from F. tularensis ΔFTL1678 trans-complemented with a 6× histidine-tagged FTL1678. Whole-cell lysates (WCL), OM fractions, and IM fractions were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotting was performed using antisera specific for the OM control protein FopA (αFopA), IM control protein SecY (αSecY), or histidine-tagged FTL1678.

2.3.2 FTL1678 Exhibits L,D-carboxypeptidase and L,D-endopeptidase

Activities

To confirm the predicted Ldc domain in FTL1678 and FTT0101, recombinant

FTL1678 and FTT0101 were expressed and affinity purified from E. coli. As a control, lysates from E. coli containing the empty vector (pPROEX HTb) also were affinity- purified. Recombinant FTL1678, lysate from the , or buffer alone were incubated with various PG precursors and PG intermediates (Table 2.1) to determine substrate specificity and specific activity. The vector control and buffer alone did not demonstrate activity against any of the PG substrates (data not shown). When FTL1678 was incubated with various PG substrates, the highest specific activity was detected against the tetrapeptide substrates GlcNAc-anhydroMurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala

53 (tracheal cytotoxin; TCT; 21.5 nmol/min/mg of protein) and GlcNAc-MurNAc-L-Ala-γ-

D-Glu-meso-A2pm-D-Ala (PG monomer; 15.6 nmol/min/mg of protein), confirming that

FTL1678 exhibits L,D-carboxypeptidase activity (Table 2.1). Interestingly, FTL1678 activity against free tetrapeptide, L-Ala-γ-D-Glu-meso-A2pm-D-Ala, was approx. 6-fold lower (3.4 nmol/min/mg of protein) than TCT and 5-fold lower than the PG monomer

(Table 2.1), indicating that GlcNAc and MurNAc may be important for tetrapeptide recognition or FTL1678 binding. Importantly, FTL1678 exhibited specific activity against pentapeptide substrates MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala-D-Ala (9.8 nmol/min/mg of protein; Table 2.1) and UDP-MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-

Ala-D-Ala (5.9 nmol/min/mg of protein; Table 2.1), indicating that FTL1678 also functions as an L,D-endopeptidase (cleavage of the pentapeptide between meso-A2pm and

D-Ala) by FTL1678.

54 Table 2.1. Substrate specificity and specific activity of F. tularensis FTL1678 and

FTT0101

aThe standard enzyme assay conditions that were described in Materials and Methods were used. bNo activity on the peptidoglycan polymer means that no release of alanine could be detected. Depending on the substrate used, the amount of partially purified protein varied from 0.9 to 5 µg per assay and incubation times varied from 30 min to 4 h. To ensure linearity, substrate consumption was < 20% in all cases. Values represent the means for three independent experiments; the standard deviation was < 10% in all cases. Specific activities were calculated from the amounts of D-Ala (tetrapeptide substrates) or D-Ala-D-Ala (pentapeptide substrates) released during the reaction.

FTL1678 had negligible activity against L-lysine-containing substrates (0.7 to 1.3 nmol/min/mg of protein; Table 2.1), where L-lysine replaced meso-A2pm at the third amino acid position, indicating the importance of meso-A2pm. Assays were repeated with recombinant FTT0101 (SchuS4 homolog), where FTT0101 demonstrated similar tetrapeptide cleavage activity (i.e., LdcA activity) as FTL1678 (Table 2.1). FTT0101 also was not active on a peptidoglycan polymer and had either no or negligible activity on PG

55 monomers that were amidated at the meso-A2pm or D-Glu residues (Table 2.1).

Additionally, FTT0101 was not inhibited by 5 mM EDTA and did not require the presence of cations (Mg2+) for tetrapeptide cleavage (data not shown).

2.3.3 FTL1678 Controls Bacterial Morphology

LdcA has been shown to be important for maintenance of bacterial morphology and structural integrity (238, 247). In addition, it is well-documented that mutation or deletion of other PG-modifying proteins results in abnormal bacterial morphology, emphasizing the importance of PG modification and recycling (205, 236, 253-256). To assess if FTL1678 plays a similar role in F. tularensis, we generated an isogenic deletion of FTL1678, referred to hereafter as ΔFTL1678, in F. tularensis LVS. When examined by transmission electron microscopy (TEM), WT bacterial width ranged from 350 to 800 nm (Fig. 2-3A and D), whereas ΔFTL1678 were more uniform in cell width, averaging approx. 350 nm (Fig. 2-

3B and D). WT bacteria were observed to be coccobacilli with loosely-associated OM

(Fig. 2-3A), while ΔFTL1678 were found to be coccoid and appeared to have a more tightly-associated OM (Fig. 2-3B). Additionally, ΔFTL1678 bacteria appeared more electron dense and had a significantly-thicker OM than WT (Fig. 2-3A, B, and C).

56

Figure 2-3. F. tularensis ΔFTL1678 has altered cell morphology. Electron micrograph images of: (A) wild-type LVS or (B) ΔFTL1678 grown in sMHB to OD600 of 0.4. Scale bars represent 100 nm; (C) Outer membrane thickness measurements (257) were measured for WT and ΔFTL1678, n=50; (D) Cell width measurements (257) for WT and ΔFTL1678, n=175. ****P<0.0001

Previous studies have shown that deletion of PG-modifying proteins (e.g., murein hydrolases) can result in abnormal growth characteristics, including lysis during stationary phase (233) and an inability to separate daughter cells at the septa during cell division, resulting in abnormal bacterial chains (205, 236, 245, 255, 258-260). Although N- acetylmuramyl-L-alanine amidases have been shown to be predominantly involved in the

57 cleavage of bacterial septa, deletion of lytic transglycosylases and some endopeptidases, in combination with amidase deletions, also have resulted in abnormal bacterial chains (255,

258). To examine if ΔFTL1678 exhibited any of these septation defects or resulted in bacterial chains, we grew ΔFTL1678 in supplemented Mueller Hinton Broth (sMHB; standard for F. tularensis; (108)), finding that it did not have an inherent growth defect (Fig. 2-4A). When examined by TEM, approximately 10% of ΔFTL1678 bacteria grew in chains of three to four bacteria (Fig. 2-S3), whereas no WT bacteria exhibited this septation defect. Considering that ΔFTL1678 is 1.5- to 2-times smaller than

WT (Fig. 2-3D), is more coccoid in shape (Fig. 2-3B), and has a partial septum defect (Fig.

2-S3), these results further support the role of FTL1678 as a PG-modifying enzyme that is important for bacterial elongation and division.

58

Figure 2-4. Susceptibility of F. tularensis WT LVS and ΔFTL1678 to stressors. WT and ΔFTL1678 were grown in 75 ml sMHB at: (A) 37˚C, (B) 40˚C, (C) with 60 µM CuCl2, (D) with 5 mM H2O2, (E) with 5% NaCl, or (F) at pH 5.5. Bacteria were grown for 24 hours and OD600 measurements were recorded every 4 hours.

2.3.4 Deletion of FTL1678 Affects Membrane Integrity and

Permeability

To determine if there were differences in membrane integrity between WT and

ΔFTL1678, we grew both bacteria in the presence of various antibiotics, detergents, and dyes, and measured zones of inhibition after 48 h of growth (Table 2.2). ΔFTL1678 was more susceptible than WT to , , lysozyme, and SDS, indicating potential changes to PG, OM integrity, or efflux pumps. By comparison, ΔFTL1678 was more resistant than WT to gentamicin, , , ciprofloxacin, ethidium bromide, and Triton X-100. Given that gentamicin, tetracycline, and chloramphenicol normally inhibit protein synthesis and ciprofloxacin and ethidium

59 bromide normally interfere with DNA replication, these results indicate that ΔFTL1678 has a less permeable envelope that prevents these inhibitory molecules from entering the cytoplasm. To better understand potential differences in ΔFTL1678 PG, OM integrity, efflux pump activity, or envelope permeability, WT and ΔFTL1678 were either grown in sMHB at 37°C or in various stress conditions. In sMHB, ΔFTL1678 does not have an inherent growth defect but, instead, appeared to grow to a higher optical density (OD600) than WT (Fig. 2-4A). However, bacterial enumeration at several time points revealed that there was no significant difference between WT and ΔFTL1678 bacterial numbers at each time point (data not shown). Although speculative, differences in optical densities between

WT and ΔFTL1678 may be due to the morphological differences observed by TEM (Fig.

2-3 and 2-S3). Compared with growth in sMHB, no substantial differences in the growth rates of WT and ΔFTL1678 were observed at either 40°C or in the presence of 60 μM

CuCl2 (Fig. 2-4B and C). However, ΔFTL1678 grew considerably better than WT in the presence of 5 mM H2O2, 5% NaCl, and pH 5.5 (Fig. 2-4D, E, and F), providing further evidence that ΔFTL1678 has a less permeable envelope. These results also corroborate the increased OM thickness and more tightly associated OM in ΔFTL1678, compared to WT

(Fig. 2-3).

60 Table 2.2. Sensitivity of F. tularensis WT LVS and ΔFTL1678 to antibiotics, detergents, and dyes

R indicates ΔFTL1678 is significantly more resistant than WT by one-way ANOVA (P<0.05) S indicates ΔFTL1678 is significantly more sensitive than WT by one-way ANOVA (P<0.05)

2.3.5 FTL1678 and Ldc Activity are Required for F. tularensis

Virulence

To examine if FTL1678 plays a role in F. tularensis virulence, mice were intranasally infected with 104 CFU of either WT or ΔFTL1678 and monitored daily for signs of disease.

Whereas all WT-infected mice died by day 9 post-infection (median time-to-death day 6),

ΔFTL1678 was completely attenuated (100% survival through day 21 post-infection), demonstrating that FTL1678 is required for F. tularensis virulence (Fig. 2-5A). To confirm that the observed attenuation was solely due to the deletion of FTL1678, and not to polar

61 effects, we complemented ΔFTL1678 with a 6× His-tagged FTL1678 in trans, which fully- restored virulence to WT levels (all mice died by day 7; median time-to-death day 6; Fig.

2-5A).

To more carefully assess ΔFTL1678 attenuation in vivo, we intranasally-infected mice with 104 CFU of either WT LVS or ΔFTL1678, and enumerated bacterial CFUs from lungs, livers, spleens, and blood on days 2 and 5 post-infection to examine bacterial replication and dissemination to these organs/tissues over time. On day 2 post-infection, WT LVS replicated to >107 CFU/mg lung and had disseminated to livers, spleens (approx. 103

CFU/mg), and blood (103 CFU/ml; Fig. 2-5B). In contrast, ΔFTL1678 had an initial (day

2) colonization defect in the lungs (4-logs less than WT) and was unable to disseminate to livers, spleens, or blood (Fig. 2-5B). By day 5 post-infection, the attenuation of ΔFTL1678 was even more apparent, with WT LVS replicating to extremely high numbers (approx.

108 CFU/mg) in lungs, livers, and spleens, compared with ΔFTL1678, which replicated approx. 1-log in lungs (between day 2 and 5), but was 4-log attenuated in lungs and was not detectable in livers or spleens. Although ΔFTL1678 was detected in the blood on day

5, it was 2-logs less than WT LVS.

62

Fig 2-5. FTL1678 and Ldc activity are required for F. tularensis virulence. (A) Groups of 5 C3H/HeN mice were intranasally-infected with 105 CFU of either wild-type WT, ΔFTL1678, or the FTL1678 trans-complement [FTL1678 compl]. Animal health was monitored daily through day 21 post-infection. **** P<0.0001; (B) Lungs, livers, spleens, and blood were aseptically harvested from mice infected with 104 CFU of either WT or ΔFTL1678 on days 2 and 5 post-infection and plated to enumerate bacterial numbers. *P<0.01; (C) Groups of 5 C3H/HeN mice were intranasally-infected with 105 CFU of either LVS, ΔFTL1678, FTL1678 trans-complement [FTL1678 compl], or C. jejuni Pgp2 trans-complement [Pgp2 compl]. Animal health was monitored through day 21 post- infection. ** P<0.01

The Type A F. tularensis strain SchuS4 originally was isolated from a human tularemia patient and requires BSL3 containment. Given its relevance to human disease, we next generated an isogenic deletion mutant of the FTL1678 homolog, FTT0101, in SchuS4.

When mice were intranasally-infected with either WT SchuS4 or ΔFTT0101, all mice died by day 7 post-infection, indicating that FTT0101 is not required for SchuS4 virulence (Fig.

2-S4). Whereas ΔFTT0101-infected mice exhibited a slightly delayed time-to-death

63 (median time-to-death day 6; Fig. 2-S4), compared with WT SchuS4-infected mice

(median time-to-death day 5; Fig. 2-S4), this may be due to differences in the infectious dose administered to mice in this experiment (80 CFU SchuS4; 12 CFU ΔFTT0101).

However, it also remains possible that the extreme virulence of SchuS4 (intranasal LD50

<10 CFU in our hands) complicates assessments of mutant attenuation.

2.3.6 FTL1678 and Ldc Activity, but not the Ldc Catalytic Triad, are

Required for F. tularensis Virulence

As noted above and highlighted in Fig. 2-1, P. aeruginosa LdcA contains a Ser-Glu-

His catalytic triad which is essential for function and is characteristic of Ldc in the

Peptidase_S66 family (235). The Ser-Glu-His catalytic triad also has been confirmed in

Ldc from E. coli (261), Novosphingobium aromaticivorans (262), and N. gonorrhoeae

(239). Given the relatively conserved spacing of Ser-Glu-His residues in FTL1678 (Fig.

2-1), we tested if these residues were required for F. tularensis virulence (similar to Fig.

2-5A virulence assessments for ΔFTL1678 and the FTL1678 complemented strain). Site- directed mutagenesis was performed to independently generate FTL1678 complementation constructs containing either S134A, E239A, or H308A mutations. Next, ΔFTL1678 was complemented in-trans with each of these FTL1678 catalytic triad point mutants, and mice were intranasally infected with either WT, ΔFTL1678, FTL1678 trans-complement, or one of the FTL1678 catalytic triad point mutant trans-complements (referred to hereafter as

S134A, E239A, and H308A). Whereas ΔFTL1678 was completely attenuated (100% survival through day 21), complementation of ΔFTL1678 with either FTL1678 (all mice

64 dead by day 8), S134A (all mice dead by day 7), E239A (all mice dead by day 8), or H308A

(all mice dead by day 8), fully-restored virulence to WT LVS levels (all mice dead by day

10; Fig. 2-S5), indicating that although a putative Ser-Glu-His catalytic triad is present in

FTL1678, these residues are not individually important for F. tularensis virulence.

The Ldc ortholog alignment (Fig. 2-1) also highlighted that, of the six Ldc orthologs examined here, only C. jejuni Pgp2 lacked the Ser-Glu-His catalytic triad. However, C. jejuni Pgp2 previously was shown to exhibit LdcA activity (238), indicating that a Ser-

Glu-His catalytic triad is not required for LdcA function. To further assess if LdcA activity, independent of a Ser-Glu-His catalytic triad, was required for F. tularensis virulence, we complemented ΔFTL1678 with a 6× His-tagged Pgp2 from C. jejuni and infected mice with either WT LVS, ΔFTL1678, ΔFTL1678 trans-complemented with FTL1678, or

ΔFTL1678 trans-complemented with C. jejuni Pgp2. While ΔFTL1678 was fully attenuated (100% survival through day 21), the Pgp2 trans-complement was fully-virulent

(median time-to-death 6 days; all mice dead by day 7), nearly identical to WT LVS (median time-to-death 7 days; all mice dead by day 7) and the FTL1678 trans-complement (median time-to-death 6 days; all mice dead by day 7; Fig. 2-5C), providing further evidence that

FTL1678 is an Ldc and Ldc activity is required for F. tularensis virulence.

To provide additional evidence that FTL1678 exhibits Ldc activity and that C. jejuni

Pgp2 functionally complements ΔFTL1678, the FTL1678 trans-complement and Pgp2 trans-complement were examined by TEM, revealing that both complemented strains had similar morphology as WT LVS (Fig. 2-S6A and B; Fig. 2-3A). Additionally, OM thickness and cell width were measured for WT, ΔFTL1678, and both complemented strains, demonstrating that both complemented strains had OM thicknesses and cell widths

65 similar to WT, and both complemented strains were significantly different from ΔFTL1678

(Fig. 2-S6C and D). Finally, when grown in the presence of various stressors (e.g., 5 mM

H2O2, 5% NaCl, pH 5.5, antibiotics, SDS, lysozyme, or ethidium bromide), both complemented strains exhibited similar phenotypes as WT LVS (Fig. 2-S6E, F, G, and H and Table 2.S2).

2.3.7 FTL1678 is Required for F. tularensis Replication in

Macrophages

F. tularensis is an intracellular pathogen and macrophages appear to be one of the major targets for F. tularensis infection and replication (36, 190, 263). To investigate potential replication defects of ΔFTL1678 in macrophages, J774A.1 macrophages or murine bone marrow-derived macrophages (mBMDM) were infected with either WT

LVS or ΔFTL1678 (MOI 100:1) and bacterial numbers were enumerated at 0 h (entry), 6 h, and 24 h post-infection. At entry (0 h), approx. 2.5-logs more ΔFTL1678 were present in both macrophage lines, compared with WT LVS (Fig. 2-6A). This likely was due to the above noted gentamicin resistance of ΔFTL1678 (Table 2.2). Attempts to normalize entry numbers for both WT LVS and ΔFTL1678, using different antibiotics or combinations of antibiotics, were not successful. Despite higher numbers of ΔFTL1678 in both macrophages at entry and 6 h, ΔFTL1678 was unable to replicate in either macrophage, and decreased approx. 1-log from 6 h to 24 h (Fig. 2-6A). By comparison,

WT LVS numbers increased 1.5- to 2-logs from 6 h to 24 h (Fig. 2-6A). To normalize

WT LVS and ΔFTL1678 bacterial numbers or replication rates in both macrophages, fold change in bacterial numbers was calculated from 6 h to 24 h: WT was found to increase

66 2.5- and 1.5-logs in J774A.1 cells and mBMDMs, respectively, whereas ΔFTL1678 was found to decrease 0.5- to 1-log, respectively (Fig. 2-6B). Taken together, these in vitro results (Fig. 2-6A and B) confirm the observed in vivo attenuation of ΔFTL1678 (Fig. 2-

5A and B).

Figure 2-6. FTL1678 is required for F. tularensis replication in macrophages. (A) J774A.1 macrophages or mouse bone marrow-derived macrophages (mBMDMs) were infected with WT or ΔFTL1678 at an MOI of 100:1 and bacterial numbers were enumerated at entry (0 h), 6 h, and 24 h post-infection. (B) Fold change in bacterial numbers from 6 to 24 h post-infection was calculated. * P< 0.01

2.3.8 ΔFTL1678 Protects Mice Against Type A F. tularensis Infection

No FDA-approved vaccine currently is available to prevent tularemia. In addition,

F. tularensis is designated as an NIH category A priority pathogen and CDC Tier 1 Select

Agent, highlighting the extreme virulence of this bacterium and the need for a safe and effective vaccine to prevent tularemia. Given our above findings that 105 CFU of

ΔFTL1678 did not cause disease or death in mice (Fig. 2-5A), we next examined whether high doses (107 or 109 CFU) of ΔFTL1678 were attenuated or if ΔFTL1678 immunization could protect mice from fully-virulent Type A F. tularensis SchuS4 challenge. First, all mice intranasally immunized with either 105, 107, or 109 CFU of ΔFTL1678 survived

67 through day 28 post-infection, with no signs of clinical disease (Fig. 2-7A). Second, on day 29, all mice were boosted with 109 CFU of ΔFTL1678 and no mice demonstrated any signs of disease through day 50 (Fig. 2-7A). Third, on day 51, mice were intranasally- challenged with 120 CFU (6× the LD50) of SchuS4 and the health status of each immunization group was monitored for an additional 26 days. In a dose-dependent manner, the 109 prime-109 boost regimen conferred 80% protection, the 107 prime-109 boost regimen conferred 40% protection, and the 105 prime-109 boost regimen conferred 20% protection (Fig. 2-7B). These data demonstrate that ΔFTL1678 is highly attenuated (up to

109 CFU) and that ΔFTL1678 may be able to be used as a live, attenuated vaccine.

Figure 2-7. ΔFTL1678 protects against fully-virulent Type A F. tularensis SchuS4. (A) Groups of 5 C3H/HeN mice were intranasally infected with either 105 CFU WT or 105, 107, or 109 CFU ΔFTL1678. On day 29 post-infection, mice were boosted with 109 CFU ΔFTL1678 and animal health was monitored daily through day 50 post-infection. *** P<0.001; (B) Mice from A were intranasally-challenged with 120 CFU of wild-type SchuS4 [BSL3; 6× LD50]. Animal health was monitored daily through day 26 post- infection. *P<0.001

68

2.3.9 ΔFTL1678 Does not Cause Tissue Damage

The in vitro (Fig. 2-6) and in vivo (Fig. 2-5A, 2-5B, 2-7A) attenuation of

ΔFTL1678, as well as protection against SchuS4 pulmonary challenge (Fig. 2-7), indicated that ΔFTL1678 could be used as a live, attenuated vaccine. While live, attenuated vaccines have been extremely effective at preventing a number of diseases, they can pose safety challenges (264, 265). To assess whether ΔFTL1678 immunization induced any pathology in immunized mice, lungs, livers, and spleens from uninfected, WT LVS-, or ΔFTL1678- infected mice were assessed for pathologic changes on day 5 post-infection/immunization.

WT LVS-infected lungs demonstrated alveolar wall thickening, large areas of inflammation, and severe neutrophil infiltration (Fig. 2-8A). By comparison, little inflammation was observed in ΔFTL1678-infected lungs, although some red blood cell congestion was present, indicating a limited, acute immune response that was quickly resolved (Fig. 2-8A). Whereas WT LVS-infected livers were characterized by diffuse inflammation with focal areas of necrosis, ΔFTL1678-infected livers were virtually indistinguishable from uninfected livers, with no observable pathology (Fig. 2-8A).

Finally, although the architecture of WT LVS-infected spleens lacked distinct areas of white pulp or red pulp, indicative of a severe infection, ΔFTL1678-infected spleens were observed to contain distinct areas of red pulp and white pulp, with some red blood cell congestion – indicating a limited, acute immune response that was quickly resolved (Fig.

2-8A). All tissues were blindly scored using a pathology severity index (scale from 0 to 4, with 4 indicating severe pathology), confirming that ΔFTL1678-infected tissues were

69 virtually indistinguishable from uninfected tissues (pathology scores of 1 for lungs, 0 for liver, and 1.5 for spleens) and WT LVS-infected tissues had significantly higher pathology scores (pathology scores >3.5 for all tissues; Fig. 2-8B).

70

Fig 2-8. ΔFTL1678 does not induce pathology. (A) Hematoxylin and eosin (H&E)- stained tissues were examined from either uninfected, F. tularensis WT LVS-, or ΔFTL1678-infected mice at 10× objective. (B) Tissues were graded on a scale of 0 to 4 with 4 being the most severe. *P<0.05

71 2.4 Discussion

Bacterial PG is a complex, mesh-like structure, composed of a glycan backbone, crosslinked to varying degrees, by peptide chains (266). It is well known that this structure plays an important role in maintaining Gram-negative bacterial cell morphology, membrane integrity, regulating changes in osmotic pressure, and providing a platform for attachment of the OM (267, 268). Although a majority of PG studies have focused on how the thick layer of PG in Gram-positive bacteria contributes to virulence and antibiotic resistance, more recent studies have highlighted that Gram-negative PG also is intimately linked with pathogenicity (205).

PG recycling is an essential function of Gram-negative bacteria during cell growth and division to produce new cell wall components. In fact, Gram-negative bacteria recycle up to 60% of their PG with every generation, suggesting that both PG synthesis and PG recycling are dynamic (215, 242). A number of proteins are involved in these processes and, while they are well-characterized in E. coli, very little is known about these pathways in intracellular pathogens such as Burkholderia pseudomallei, , or

F. tularensis (146, 147, 269, 270).

E. coli LdcA, a cytoplasmic protein, was the first L,D-carboxypeptidase to be identified and was shown to be important for PG recycling and survival during stationary phase (233, 246). More recently, Ldcs have been identified in P. aeruginosa (235), C. jejuni (238), N. gonorrhoeae (271), and N. aromaticivorans (262). In this study, we identified an F. tularensis PG recycling enzyme, FTL1678, which we propose naming

LdcA. Unlike well-characterized cytoplasmic LdcAs from E. coli and P. aeruginosa, we demonstrated that F. tularensis LdcA was localized to OM fractions and, given co-

72 localization with PG-associated proteins Pal and TolB, is most likely located on the inner leaflet of the OM or in the periplasm. At this time, we can only speculate on why most

Gram-negative bacteria encode cytosolic LdcAs but F. tularensis encodes a periplasmic

LdcA, but in the context of PG repair and recycling, periplasmic LdcA certainly offers a fitness advantage. This is not the first report of a periplasmic LdcA, as C. jejuni Pgp2 is predicted to be periplasmic and N. gonorrhoeae LdcA is periplasmic (238, 239).

Our results demonstrated, for the first time, that F. tularensis LdcA directly acts on the tetrapeptides TCT and PG monomer. More importantly, we also found that F. tularensis

LdcA directly cleaves the pentapeptide to a tripeptide, highlighting that F. tularensis LdcA is a multi-functional enzyme that performs both L,D-carboxypeptidase and L,D- endopeptidase activities (Table 2.1). In addition, we demonstrated that F. tularensis LdcA cleaves PG pentapeptides without a prior cleavage event by a D,D- carboxypeptidase/Penicillin Binding Protein (PBP), such as DacD. It should be noted that although N. gonorrhoeae LdcA has been reported to have L,D-endopeptidase activity, this activity is specific to tetrapeptide cross links (4-3 and 3-3 links) (239). Interestingly, only two previous studies have examined putative F. tularensis PG modifying enzymes. In both studies, the authors primarily focused on the role of a F. tularensis DacD ortholog in virulence, with no PG activity assays to confirm function (146, 147). In our PG cleavage analysis, F. tularensis LdcA demonstrated the highest specific activity on disaccharide- tetrapeptide PG substrates (GlcNAc-anhydroMurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala

[TCT] and GlcNAc-MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala [PG monomer]), followed by cleavage of pentapeptide PG substrates (MurNAc-L-Ala-γ-D-Glu-meso-

A2pm-D-Ala-D-Ala and UDP-MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala-D-Ala).

73 Despite high specific activity of F. tularensis LdcA on tetrapeptide attached to the disaccharide, F. tularensis LdcA demonstrated approximately 6-times lower specific activity on free tetrapeptide (no sugars) (Table 2.1). In contrast, E. coli LdcA has been shown to have the highest specific activity on free tetrapeptide, monosaccharide- tetrapeptide (MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-Ala), and monosaccharide tetrapeptide linked to a glycan lipid carrier (UDP-MurNAc-tetrapeptide) (233), but is unable to cleave dimeric muropeptides. Additionally, F. tularensis LdcA was active against the TCT dimer (two crosslinked TCT monomers), but not PG polymers. Although purely speculative, this TCT dimer cleavage potentially indicates that F. tularensis LdcA may possess endopeptidase activity on peptide cross-links.

Previous studies have shown that Ldcs are important for cell morphology and membrane integrity. Deletion of the Ldc homologs Csd6 from H. pylori (247) and Pgp2 from C. jejuni (238) resulted in loss of helical morphology. Here, we demonstrated that

FTL1678 is essential for maintaining the coccobacillus morphology of F. tularensis, as

ΔFTL1678 were significantly-smaller cocci that were more electron dense than WT.

ΔFTL1678 had significantly thicker envelopes than WT bacteria and the ΔFTL1678 OM was more tightly-associated with the bacteria, compared to WT OMs that appeared to be loosely-associated. It is possible that deletion of FTL1678 may have prevented efficient recycling or breakdown of existing PG, resulting in a buildup of pentapeptides or tetrapeptides that are highly-crosslinked. However, repeated attempts to isolate and analyze F. tularensis PG have been unsuccessful and, therefore, we can only speculate on the true nature of the thick envelope in ΔFTL1678.

74 Due to the increased thickness of the envelope and altered cell morphology of

ΔFTL1678 observed by TEM, we investigated differences in membrane integrity between

WT and ΔFTL1678, finding that ΔFTL1678 was more susceptible to several compounds including ampicillin, vancomycin, lysozyme, and SDS. Vancomycin, usually not effective against Gram-negatives due to its inability to penetrate the OM, inhibited ΔFTL1678 growth, indicating increased permeability of the ΔFTL1678 OM. Vancomycin may have been especially effective on ΔFTL1678 because its mechanism of action includes binding to the two terminal D-Ala-D-Ala residues of PG pentapeptide chains and preventing cross- linking of monomers. ΔFTL1678 also may have increased the amount of pentapeptides present in PG, providing more targets for vancomycin action.

We also found that ΔFTL1678 was more resistant to several stressors, including gentamicin, tetracycline, chloramphenicol, ciprofloxacin, ethidium bromide, Triton-X 100,

H2O2, pH 5.5, and high NaCl. Although the specific mechanism for increased resistance of ΔFTL1678 to these compounds is unclear, we hypothesize that the thick PG of

ΔFTL1678 may be more highly crosslinked, preventing large compounds like gentamicin, tetracycline, chloramphenicol, ciprofloxacin, ethidium bromide, and Triton-X 100 from getting through and/or may have increased the expression of efflux pumps to efficiently remove toxic compounds from the bacteria.

Due to the extreme virulence of Type A F. tularensis and its designation as a Tier 1

Select Agent, identification of virulence factors and development of new vaccines is important. In this study, we identified the role of a previously unstudied protein, FTL1678, in PG recycling and envelope integrity maintenance, we demonstrated that FTL1678 was required for F. tularensis LVS virulence, and we found that ΔFTL1678 conferred 80%

75 protection against fully-virulent, Type A F. tularensis SchuS4 pulmonary challenge.

Further studies are needed to determine specific immune responses induced by ΔFTL1678 immunization, as well as identify the most effective immunization regimen (e.g., number of immunizations and time between immunizations).

Finally, given the importance of PG in F. tularensis virulence, we used bioinformatic approaches to predict the proteins involved in F. tularensis PG synthesis and recycling

(Fig. 2-9). With the exception of DacD, none of the predicted PG-associated proteins previously have been studied and homologues could not be found for at least seven genes/proteins (Fig. 2-9). Given our observed attenuation of ΔFTL1678, the PG synthesis and recycling pathways may offer more opportunities to better understand the virulence of

F. tularensis and other intracellular pathogens. Characterization of other proteins involved in PG pathways may provide clues as to why F. tularensis LdcA is periplasmic and encodes both L,D-carboxypeptidase and L,D-endopeptidase activities.

76

Figure 2-9. Peptidoglycan synthesis and recycling pathways remain to be defined in F. tularensis. Bioinformatic analyses were used to identify proteins that may be involved in peptidoglycan synthesis and recycling in F. tularensis LVS. Locus tags are indicated. Red text indicates prototypic E. coli proteins with no apparent F. tularensis orthologs. Blue text indicates predicted localization of F. tularensis proteins that are different from the E. coli orthologs.

2.5 Materials and Methods

2.5.1 Bacterial Strains and Culture Conditions

F. tularensis Type A strain SchuS4 and F. tularensis Type B strain LVS were obtained from BEI Resources and cultured as previously described (118, 170). All experiments with SchuS4 were performed under BSL3 containment conditions at the

University of Toledo Health Science Campus BSL3 laboratory. Routine F. tularensis

77 cultures were grown overnight at 37°C with 5% CO2 on supplemented Mueller-Hinton agar

(sMHA): Mueller-Hinton broth powder (Becton Dickinson) was mixed with 1.6% (wt/vol)

Bacto Agar (Becton Dickinson), autoclaved, and further supplemented with 2.5% (vol/vol) bovine calf serum (Hyclone), 2% (vol/vol) IsoVitaleX (Becton Dickinson), 0.1% (wt/vol) glucose, and 0.025% (wt/ vol) iron pyrophosphate. For mouse infections, F. tularensis was first grown on sMHA then transferred to Brain Heart Infusion agar (BHI; Becton

Dickinson). for mutant strain generation was prepared by mixing Mueller

Hinton broth powder with 1.6% (wt/vol) agar, 1% (wt/vol) tryptone, and 0.5% (wt/vol) sodium chloride, autoclaved, and further supplemented with 1% (wt/vol) hemoglobin and

1% (vol/vol) IsoVitaleX. For macrophage infections, F. tularensis was first grown on sMHA then transferred to modified chocolate agar: Mueller-Hinton broth powder was mixed with 1.6% (wt/vol) Bacto Agar, 1% hemoglobin (wt/vol), and 1% (vol/vol)

IsoVitaleX. All growth curves were performed in sMHB: Mueller-Hinton broth powder was mixed with 182 mg/L calcium chloride dihydrate, and 210 mg/L magnesium chloride hexahydrate, 0.1% (wt/vol) glucose, 0.025% (wt/vol) iron pyrophosphate, and 2% (vol/vol)

IsoVitaleX. All bacterial strains and plasmids are listed in Table 2.S3.

S17-1 and E. coli NEB10-β were grown in Luria Bertani (LB) broth or on LB agar at 37°C, supplemented as needed with antibiotics.

2.5.2 Sequence Alignments and Bioinformatic Predictions

Amino acid alignments of F. tularensis subsp. holarctica FTL_1678, F. tularensis subsp. tularensis FTT_0101, E. coli LdcA (BAA36050.1), P. aeruginosa LdcA

(Q9HTZ1), N. gonorrhoeae (YP_208343.1), and C. jejuni Pgp2 (WP_002856863) were

78 performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) and

MView (https://www.ebi.ac.uk/Tools/msa/mview/). Pairwise sequence alignments were performed and amino acid identities among Ldc homologues were calculated by EMBOSS

Needle (https://www.ebi.ac.uk/Tools/psa/emboss_needle/). The Prokaryotic Genome

Analysis Tool (PGAT) (http://tools.uwgenomics.org/pgat/), BlastP, and BlastX analyses

(http://blast.ncbi.nlm.nih.gov) were used to identify F. tularensis homologues. Bacterial protein sub-localization was predicted by PSORTb version 3.0.2

(https://www.psort.org/psortb/). Protein signal sequence prediction was performed by

LipoP version 1.0 (http://www.cbs.dtu.dk/services/LipoP/) and SignalP version 4.1

(http://www.cbs.dtu.dk/services/SignalP-4.1/).

2.5.3 Generation of F. tularensis Gene Deletion Mutants

F. tularensis isogenic deletion mutants were generated by homologous recombination as previously described (272). Briefly, 500 bp regions upstream and downstream from the gene of interest (FTL1678 or FTT0101) were PCR-amplified from

F. tularensis genomic DNA using the following primers: FTL1678_A and FTL1678_B;

FTL1678_C and FTL1678_D; FTT0101_A and FTT0101_B; FTT0101_C and

FTT0101_D (Table 2.S4). A FLP recombination target (FRT)-flanked Pfn-kanamycin resistance cassette, FRT-Pfn-kan-FRT, was PCR amplified from pLG66a (273) and splicing overlap extension PCR (SOE PCR) was used to join the upstream (A-B) and downstream (C-D) regions with FRT-Pfn-kan-FRT, which replaced the gene of interest.

The resulting insert and a suicide plasmid, pTP163 (103), were digested with ApaI (New

England Biolabs), and ligated using T4 DNA ligase (New England Biolabs). Gene

79 deletion constructs were transformed into NEB10-β E. coli (New England Biolabs), sequence-verified, transformed into E. coli S17-1, and conjugation was performed with

F. tularensis LVS on sMHA plates. Conjugants were recovered on chocolate agar supplemented with 200 mg/L hygromycin and 100 mg/L polymyxin B. Individual mutants were selected by sequential plating on sMHA supplemented with 10 mg/L kanamycin (sMHA-kan10), sMHA-kan10 with 8% (wt/vol) sucrose, and final replica plating onto sMHA containing either 200 mg/L hygromycin (sMHA-hyg200) or sMHA- kan10. Hyg-sensitive and kan-resistant colonies were sequence verified (referred to hereafter as either ΔFTL1678 or ΔFTT0101).

2.5.4 FTL1678 Complementation in trans

Complementation in trans was performed as previously described, with some modifications (118). FTL1678 was PCR-amplified from F. tularensis LVS using primers

5’FTL1678_NEBuilder and 3’FTL1678_NEBuilder (Table 2.S4), pQE-60 (Qiagen) was double-digested with NcoI and BglII (New England Biolabs), and the NEBuilder HiFi

DNA Assembly Cloning kit was used to ligate the FTL1678 amplicon and digested pQE-

60. The construct was transformed into NEB 10-β E. coli and transformants were selected on LB agar supplemented with 100 mg/L ampicillin (LB-amp). Plasmids were purified from individual clones using the Qiagen QIAprep Spin Miniprep kit (Qiagen), diagnostic PCR was performed to confirm insert presence and correct size, and DNA sequencing was performed to verify insert integrity. The resulting construct, FTL1678 with a C-terminal 6x histidine tag, was PCR-amplified using primers

5’FTL1678_pFNLTP6 and 3’FTL1678_pFNLTP6 (Table 2.S4), the amplicon and

80 pFNLTP6-gro-GFP (274) were double-digested with XhoI and BamHI (New England

Biolabs), and ligated using T4 DNA Ligase. The construct, pFNLTP6-gro-FTL1678-

6xHis, was transformed into NEB10-β E. coli, transformants were selected on LB plates supplemented with 50 mg/L kanamycin (LB-kan), and DNA sequencing was performed to verify FTL1678-6xHis integrity. Next, the kan resistance gene was removed from

ΔFTL1678 by suspending the strain in 0.5 M sucrose (in 1 mM EDTA, pH 7.5), washing three times, and electroporating the shuttle plasmid pTP405 (103), which encodes the Flp recombinase to remove FRT-Pfn-kan-FRT from the genome. Bacteria were grown overnight on sMHA-hyg200, hyg-resistant transformants were passaged three times on sMHA, then transformants were replica plated onto sMHA-hyg200 and sMHA-kan10 to confirm sensitivity to both antibiotics (kan-cured ΔFTL1678). pFNLTP6-gro-FTL1678-

6xHis was transformed into kan-cured ΔFTL1678 by electroporation, transformants were selected on sMHA-kan10, and expression of FTL1678 was confirmed by immunoblot analysis (referred to hereafter as ΔFTL1678 trans-complement).

2.5.5 C. jejuni Pgp2 Complementation in trans

Complementation of C. jejuni pgp2 (CJJ81176_0915) into ΔFTL1678 was performed as described above, with several modifications. The pgp2 gene, with the

FTL1678 signal sequence (amino acid residues 1-29) in place of the native Pgp2 signal sequence (amino acid residues 1-18), was synthesized and inserted in pQE-60 by GenScript

USA. pQE-60-pgp2 was transformed into NEB10-β E. coli and selection was performed on LB-amp. Pgp2-6xHis was amplified from pQE-60 using primers

5’FTL1678_pFNLTP6 and 3’FTL1678_pFNLTP6 (Table 2.S4), the amplicon was ligated

81 into similarly digested pFNLTP6, pFNLTP6-gro-pgp2-6xHis was transformed into

NEB10-β E. coli, and transformants were selected on LB-kan. Plasmids were purified from kan-resistant transformants, sequence verified, then electroporated into kan-cured

ΔFTL1678. Pgp2 expression was confirmed by immunoblot analysis.

2.5.6 Mouse Infections

All animal studies were approved by the University of Toledo Institutional Animal

Care and Use Committee (IACUC). Mouse infections were performed as previously described (78), with some modifications. Briefly, F. tularensis strains were grown on sMHA overnight, transferred to BHI agar for an additional 20-24 h, suspended in sterile

PBS, and diluted to the desired concentration (20 to 109 CFU/20 µl) based on previous

OD600 measurements and bacterial enumeration studies. Groups of 4-8 female C3H/HeN mice (6-8 weeks old; Charles River Laboratories) were anesthetized with a ketamine- xylazine sedative and intranasally (i.n.) infected with 20 µl of prepared bacterial suspensions. Bacterial innocula were serially-diluted and plated in quadruplet on sMHA to confirm CFUs. For survival studies, mice were monitored daily, for signs of disease, with health status scores (scale of 1-5, with 1 indicating healthy and 5 indicating mouse found dead) being recorded for each mouse. Moribund mice were humanely euthanized to minimize suffering. To quantitate bacterial tissue burdens, groups of 4 mice were euthanized on days 2 and 5 post-infection, blood was collected by cardiac puncture and plated onto sMHA, lungs, livers, and spleens were aseptically harvested, homogenized, 25

µl of PBS/mg of tissue was added to each tissue, serially-diluted, and dilutions were plated onto sMHA. Following 72 h of incubation, the number of colonies per plate were counted

82 and CFU/mg (tissues) or CFU/ml (blood) were calculated based on tissue weight and dilution factor. For immunization and challenge studies, groups of 4-10 mice were i.n. immunized with either 100-300 CFU LVS or 104-109 CFU ΔFTL1678, boosted 3-4 weeks later with either 103 CFU LVS or 109 CFU ΔFTL1678, transported to the ABSL3 facility

3-weeks later, and i.n. challenged with 20-120 CFU of F. tularensis SchuS4. Mice were monitored daily for signs of disease with health status scores being recorded for each mouse.

2.5.7 Membrane Integrity Testing

Sensitivity of LVS, ΔFTL1678, FTL1678 trans-complement, and the Pgp2 trans- complement to various antibiotics, detergents, dyes, and cell wall stressors was determined by disk diffusion assays or in liquid cultures, as previously described (118), with some modifications. Bacterial strains were grown on either sMHA or sMHA-kan10 (ΔFTL1678 and complement strains), scraped and resuspended in sterile PBS, adjusted to an OD600 of

0.2 (approx. 9x107 CFU/ml), diluted 1:1 in PBS, and 100 µl was plated onto sMHA plates using cotton tipped applicators (Puritan). Sterile paper disks (Whatman; 0.8 mm thick, 6.5 mm in diameter) were placed in the center of each plate and antibiotics, detergents, or dyes were added to the disks at the concentrations listed in Table 2.2. Antibiotics tested were: gentamicin (Gibco), tetracycline (Fisher Scientific), chloramphenicol (Acros Organics), ciprofloxacin (Oxoid), ampicillin (Fisher Scientific), vancomycin (Acros Organics), (Oxoid), bacitracin (Oxoid), ciprofloxacin (Oxoid), and polymyxin b (MP

Biomedicals). Detergents tested were: sodium dodecyl sulfate (SDS; anionic; Fisher

83 Scientific), Triton X-100 (nonionic; Acros Organics), cetyltrimethyl ammonium bromide

(CTAB; cationic; MP Biomedicals), 3-cholamidopropyl dimethylammonio 1- propanesulfonate (CHAPS; zwitterionic; Thermo Scientific). In addition, sensitivity to ethidium bromide (Thermo Scientific) and lysozyme (Thermo Fisher) also was tested.

After 48 h, diameters of zones of inhibition around the disks were measured. Experiments were performed in triplicate. For liquid cultures, bacteria were suspended in sMHB, adjusted to OD600 0.4, and 5 ml of each bacterial suspensions was inoculated into 100 ml of either sMHB or sMHB with 5 mM hydrogen peroxide (H2O2), 5% sodium chloride

(NaCl), or pH 5.5 (pH of sMHB is 6.5). Cultures were grown at 37˚C with rotation at 180 rpm for 24 h with OD600 readings recorded every 4 hours.

2.5.8 Electron Microscopy

Electron microscopy was used to visualize differences in bacterial envelope structure and cell shape, as previously described (118), with some modifications. LVS,

ΔFTL1678, FTL1678 trans-complement, and the Pgp2 trans-complement were grown overnight in sMHB, approx. 1x109 CFU of each bacterial strain was pelleted by centrifugation at 7000 × g at 4˚C, washed three times in PBS, fixed in 3% (vol/vol) glutaraldehyde (Electron Microscopy Sciences [EMS]) for approx. 24 hours, washed twice in sodium cacodylate buffer (pH 7.4; EMS) for 10 min, suspended in 1% (wt/vol) osmium tetroxide (EMS) in s-collidine buffer (pH 7.4; EMS) for 45 min at room temperature (r/t) to stain and fix the samples, washed two times with sodium cacodylate buffer for 10 min each, and tertiary fixation was performed using an aqueous saturated solution of uranyl acetate (pH 3.3; EMS) for 45 min at r/t. Samples were then dehydrated

84 at room temperature using a series of ethanol washes: two washes with 30% ethanol for

10 min each; two washes with 50% ethanol for 10 min each; two washes with 95% ethanol for 10 min each; two washes with 100% ethanol for 10 min each; and two washes with 100% acetone for 10 min each. Samples were then infiltrated with 50% acetone and

50% embedding media (Hard Plus Resin 812, EMS) for 8 h to overnight at r/t. Samples were embedded in 100% embedding media (EMS) and allowed to polymerize for 8 h to overnight at 85°C, then sectioned at 85–90 nm, and visualized using a Tecnai G2 Spirit transmission electron microscope (FEI) at 80 kV and Radius 1.3 (Olympus) camera software at the University of Toledo Electron Microscopy Facility.

2.5.9 Spheroplasting and Sucrose Density Gradient Centrifugation

Spheroplasting, osmotic lysis, and sucrose density gradient centrifugation was performed as previously described (108) to determine subcellular localization of FTL1678.

Briefly, the histidine-tagged FTL1678 trans-complement was grown in sMHB to an OD600 of 0.3-0.4, pelleted at 7500 × g for 30 min at 10˚C, supernatants were removed, pellets were resuspended in 0.75 M sucrose (in 5 mM Tris, pH 7.5) with gentle mixing, 10 mM

EDTA (in 5 mM Tris, pH 7.8) was slowly added over 10 min, and the suspension was incubated for 30 min at r/t. After incubation, lysozyme was slowly added to a final concentration of 200 μg/ml, incubated for 30 min at r/t, bacteria were osmotically lysed by dilution into 4.5 × volume of molecular-grade water (Corning) over 11 min with gentle mixing, and incubated for 30 min at r/t. Lysates were centrifuged at 7,500 × g for 30 min at 10°C to remove intact cells and cellular debris. Supernatants were collected and centrifuged at 182,500 × g for 2 h at 4°C in a F37L 8 × 100 Fiberlite Ultracentrifuge rotor.

85 Following centrifugation, supernatants were removed, membrane pellets were gently resuspended in 6 ml of resuspension buffer (25% [wt/wt] sucrose, 5 mM Tris, 30 mM

MgCl2, 1 tablet of Pierce Protease Inhibitor Mini Tablets, EDTA-Free [Thermo Scientific],

5 U Benzonase [Novagen]), suspensions were incubated with gentle mixing for 30 min at room temperature to degrade DNA, and a DC protein assay (Bio-Rad) was performed to determine total protein yield. Linear sucrose gradients were prepared by layering 1.8 ml each of sucrose solutions (wt/wt; prepared in 5 mM EDTA, pH 7.5) into 14- by 95-mm ultracentrifuge tubes (Beckman) in the following order: 55%, 50%, 45%, 40%, 35%, and

30%. Membrane suspensions were layered on top of each sucrose gradient, with less than

1.5 mg of protein per gradient. Sucrose gradients were centrifuged in an SW40 swinging bucket rotor (Beckman) at 256,000 × g for 17 h at 4°C. After centrifugation, 500 μl fractions were collected from each gradient by puncturing the bottom of each tube and allowing fractions to drip into microcentrifuge tubes. The refractive index of each fraction was determined using a refractometer (Fisher Scientific) and correlated with a specific density in g/ml (275) to identify outer membrane (OM; 1.17-1.20 g/ml) and inner membrane (IM; 1.13-1.14 g/ml) fractions. Sucrose gradient fractions were examined by immunoblotting as described below.

2.5.10 Immunoblotting

Whole cell lysates of FTL1678 trans-complement were prepared by suspending bacteria (pelleted at 7000 × g) in molecular biology grade water, diluting with SDS-

PAGE loading buffer, and boiling for 10 min. Whole cell lysates, OM fractions, IM fractions, and molecular mass standards (Precision Plus protein all blue prestained protein

86 standards; BioRad Laboratories) were separated on a 12.5% polyacrylamide gel, transferred to nitrocellulose, and blots were incubated overnight in blot block (0.1%

(vol/vol) Tween 20 and 2% (wt/vol) bovine serum albumin in PBS) at 4°C.

Immunoblotting was performed using rat polyclonal antiserum specific for either F. tularensis OM protein FopA, F. tularensis IM protein SecY (108) or the Penta-His HRP conjugate antibody (Qiagen).

2.5.11 Infections of Mouse Bone Marrow Derived-Macrophages

(mBMDMs) and J774A.1 Cells

Macrophage culture (37˚C with 5% CO2 unless otherwise indicated) and infections were performed as previously described (118), with some modifications. Bone marrow macrophages were harvested from female C3H/HeN mice. Mice were euthanized by CO2 asphyxiation and cervical dislocation. Femurs and tibias of both hind legs were aseptically-harvested, marrow was flushed from each bone with RPMI-1640

(Hyclone) containing 10% heat-inactivated fetal bovine serum ([HI-FBS], Atlanta

Biologicals) and 30% supernatants from day 7 L929 cultures (ATCC). Bone marrow was disrupted by repeated passage through a 23-gauge needle and cultured for 4 days. Next, cell media was removed and replaced with RPMI containing 10% HI-FBS and 30% supernatant from day 14 L929 cultures, and cells were cultured for 2 days. Approx. 24 h before infection, media was removed, cells were harvested by scraping and centrifugation at 400 × g for 10 min at 10˚C, cells were enumerated using a hemocytometer, and diluted to 1x105 cells in RPMI containing 10% HI-FBS. J774A.1 cells (ATCC) were cultured in

Dulbecco’s Modified Eagle Medium ([DMEM], Gibco) containing 10% HI-FBS.

87 Approx. 24 h before infection, cells were harvested as described above, seeded into individual wells of 24-well plates (Corning) at a concentration of 1x105 cells/well, and incubated overnight. mBMDMs and J774A.1 cells were infected with a multiplicity of infection (MOI) of 100 bacteria to 1 cell (100:1). Following infection, cells were centrifuged at 1,000 × g for 10 min at 4˚C, incubated at 37˚C with 5% CO2 for 1 h, washed 1 × with RPMI (or DMEM), media containing 100 µg/ml gentamicin was added to kill extracellular bacteria, cells were incubated at 37˚C with 5% CO2 for 1 h, washed

1x with RPMI (or DMEM), lysed with 1% saponin for 4 min, serially diluted in PBS, plated onto sMHA plates, and bacteria were enumerated (entry) after 48h. Alternatively, after gentamicin treatment and washing, RPMI (or DMEM) containing 10% HI-FBS was added to cells and they were incubated for 6 or 24 h, lysed, serially-diluted, and plated to determine bacterial numbers.

2.5.12 Expression and Purification of Recombinant FTL1678 and

FTT0101

F. tularensis LVS and SchuS4 genomic DNA were extracted using phenol/chloroform/isoamyl alcohol (Fisher Bioreagents). FTL1678 and FTT0101, without signal sequences (amino acid residues 1-29), were PCR-amplified from LVS and SchuS4 genomic DNA, respectively, using High Fidelity Platinum Taq Polymerase (Life

Technologies), and primers 5’FTL1678_BamHI and 3’FTL1678_XhoI and

5’FTT0101_BamHI and 3’FTT0101_XhoI, respectively (Table 2.S4). Amplicons and pPROEX HTb were double-digested with BamHI and XhoI, ligated using T4 DNA ligase,

88 and transformed into NEB 10-β E. coli. Plasmids were purified using the Qiagen QIAprep

Spin Miniprep kit and diagnostic PCR was performed to confirm presence and correct size of the insert. DNA sequencing was performed to confirm insert integrity and plasmid constructs were transformed into Rosetta DE3 E. coli (Millipore) for protein expression.

Recombinant proteins were expressed and purified as previously described (170) with some modifications. Bacteria were grown in LB-amp to an OD600 of 0.4, protein expression was induced for 2 h by the addition of isopropyl β-D-thiogalactopyranoside

(IPTG) to a final concentration of 100 mM, bacteria were pelleted by centrifugation, and frozen overnight at −80°C to aid in lysis. Cell pellets were suspended in 10 mM Tris, 500 mM NaCl, and 10 mM imidazole, pH 8.0, sonicated on ice for 10 min with 30 sec intervals, insoluble material was removed by centrifugation at 8,000 × g, and supernatants were collected for affinity purification over pre-equilibrated Nickel-nitrilotriacetic acid (Ni-

NTA) agarose (Qiagen) columns. Eluted recombinant proteins were concentrated in

Amicon Ultra-4 centrifugal filter units with 30-kDa cutoff (Millipore), concentrations were determined using the DC BCA protein assay (BioRad), and purity was assessed by SDS-

PAGE and Imperial protein staining (Thermo Scientific). An empty vector construct also was expressed and purified as a control in enzymatic assays.

2.5.13 Enzymatic Assays for FTL1678/FTT0101 Activity

FTL1678 and FTT0101 recombinant protein activity toward various PG-related compounds were tested in 50 µl reaction mixtures containing 50 mM Tris-HCl, pH 8.0, 0.1 mM substrate, and partially purified enzyme stock (10 µl in 1 M NaCl, 10 mM Tris, pH

8.0). Mixtures were incubated for 30 min to 2 h at 37°C and reactions were stopped by

89 freezing. Substrate and reaction products were separated by HPLC on an ODS-Hypersil 3

µm particle-size C18 column (250 by 4.6 mm; Thermo Scientific). Elutions were performed with 50 mM sodium phosphate buffer, pH 4.5, with or without application of a linear gradient of methanol (from 0 to 50% in 50 min), at a flow rate of 0.5 ml/min. Peaks were detected by measuring the absorbance at 207 nm or at 262 nm for UDP-containing nucleotide precursors. Identification of compounds was based on their retention times, compared to authentic standards, as well as on their amino acid and amino sugar composition, determined with a Hitachi model L8800 analyzer (Sciencetec) after hydrolysis of samples in 6 M HCl for 16 h at 95°C. Enzyme activity was calculated by integration of peaks corresponding to substrate and product. Amounts of alanine released by the L,D-carboxypeptidase activity also were determined using an amino acid analyzer.

Depending on the substrate used, the amount of partially purified protein varied from 0.9 to 5 µg per assay and incubation times varied from 30 min to 4 h. To ensure linearity, substrate consumption was < 20% in all cases. Values represent the means for three independent experiments; the standard deviation was < 10% in all cases. Specific activities were calculated from the amounts of D-Ala (tetrapeptide substrates) or D-Ala-D-Ala

(pentapeptide substrates) released during the reaction.

2.5.14 Peptidoglycan Precursors and Muropeptides

UDP-MurNAc-pentapeptide precursors containing either meso-diaminopimelic acid (A2pm) or L-lysine were prepared by enzymatic synthesis using purified Mur ligases, and UDP-MurNAc-tetrapeptides were generated by treatment of the UDP-MurNAc- pentapeptide precursors with purified E. coli PBP5 DD-carboxypeptidase as previously

90 described (276). MurNAc-peptides were obtained by mild acid hydrolysis of UDP-

MurNAc-peptides (0.1 M HCl, 100°C, 15 min) and were not reduced and thus purified as a mixture of the two α and β anomers (277). Free peptides were prepared by cleavage of

MurNAc-peptides with E. coli AmiD N-acetylmuramoyl-L-alanine amidase (278). The E. coli peptidoglycan polymer was purified from a Δlpp mutant strain that does not express the Lpp lipoprotein (279). GlcNAc-1,6-anhydro-MurNAc-L-Ala-γ-D-Glu-meso-A2pm-D-

Ala (TCT) and its dimer (two cross-linked TCT monomers) were produced by digestion of peptidoglycan with E. coli SltY lytic transglycosylase and the non-anhydro forms of these monomer and dimer were generated by digestion of the polymer with mutanolysin (280).

All these compounds were HPLC-purified and their composition was controlled by amino acid and amino sugar content analysis and/or by MALDI-TOF mass spectrometry.

2.5.15 Statistics

GraphPad Prism6 was used in various statistical analyses, including: differences in antibiotic, detergent, dye, or lysozyme susceptibility were calculated by one-way

ANOVA with multiple comparisons and the Holm-Sidak post-hoc test; differences in

EM measurements were determined by unpaired t-tests; differences in median time-to- death and percent survival following F. tularensis infection of mice were calculated using the log-rank Mantel-Cox test; differences in pathology scores of F. tularensis-infected tissues were calculated by two-way ANOVA with multiple comparisons and a Tukey post-hoc test. Differences in lung, liver, spleen, and blood bacterial burdens from infected mice were calculated by one-way ANOVA with multiple comparisons using R software.

91

2.6 Supplemental Material

Table 2.S1: Bioinformatic analyses of FTL1678

a PSORTb version 3.0.2 bacterial protein subcellular localization prediction program (http://www.psort.org/psortb/). Protein localization to the outer membrane, cytoplasm, cytoplasmic membrane, periplasm, or extracellular space based on scores > 7.5 (considered significant). Unknown (Unk) localization indicated by scores < 7.5. b BOMP beta-barrel integral outer membrane protein prediction program (http://services.cbu.uib.no/tools/bomp). Scores > 2 considered significant. c LipoP 1.0 lipoprotein signal peptide prediction program (http://www.cbs.dtu.dk/services/LipoP/). SpI indicates predicted signal peptide for signal peptidase I cleavage, with log-odds score, and predicted cleavage site. Lipoprotein signal peptidase II cleavage prediction was not significant. d SignalP 4.1 signal peptide prediction program (http://www.cbs.dtu.dk/services/SignalP- 4.1/). D score >0.450 is considered significant. Position indicates signal peptide residues.

92 Table 2.S2: Sensitivity of F. tularensis WT LVS, ΔFTL1678, FTL1678 trans- complement, and Pgp2 trans-complement to antibiotics, detergents, and dyes

R indicates ΔFTL1678 is significantly more resistant than WT by one-way ANOVA (P<0.05) S indicates ΔFTL1678 is significantly more sensitive than WT by one-way ANOVA (P<0.05)

93 Table 2.S3: Bacterial strains and plasmids used in this study

94 Table 2.S4: Primers used in this study

Figure 2-S1. FTL1678 contains a putative L,D-carboxypeptidase domain. NCBI Conserved Domain search results for F. tularensis FTL1678.

95

Figure 2-S2. TolB is OM-localized. Spheroplasting, osmotic lysis, and sucrose density gradient centrifugation were performed to separate inner membranes (IM) and outer membranes (OM) from F. tularensis ΔFTL1678 trans-complemented with a 6× histidine- tagged FTL1678. Whole-cell lysates (WCL), OM fractions, and IM fractions were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotting was performed using antisera specific for the periplasmic protein TolB (αTolB).

96

Figure 2-S3. FTL1678 is required for correct septation. Electron micrograph image of ΔFTL1678 showing aberrant septal formation and reduced ability to separate cells. Red arrows point to formed septa that have not separated and white arrows point to new septa that are forming. Scale bars represent 200 nm.

97

Figure 2-S4. FTT0101 is not required for virulence. C3H/HeN mice were intranasally infected with either 80 CFU SchuS4 (n=3 mice) or 12 CFU ΔFTT0101 (n=5 mice).

98

Figure 2-S5. Catalytic triad is not essential for F. tularensis virulence. Groups of 5 C3H/HeN mice were intranasally-infected with 105 CFU of either F. tularensis WT LVS, ΔFTL1678, FTL1678 trans-complement [FTL1678 compl], S134A trans-complement [S134A], E239A trans-complement [E239A], or H308A trans-complement [H308A]. Animal health was monitored daily through day 21 post-infection. ** P<0.01

99

100 Figure 2-S6. FTL1678 trans-complement and C. jejuni Pgp2 trans-complement restore WT F. tularensis LVS phenotype. Electron micrograph images of: (A) FTL1678 trans-complement [FTL1678 compl] or (B) C. jejuni Pgp2 trans-complement [Pgp2 compl] grown in sMHB to OD600 of 0.4. Scale bars represent 100 nm. (C) Outer membrane thickness and (D) cell width of FTL1678 trans-complement and C. jejuni Pgp2 trans-complement were compared to LVS and ΔFTL1678, ****P<0.0001. LVS, ΔFTL1678, FTL1678 trans-complement, or C. jejeuni Pgp2 trans-complement were grown in sMHB at (E) 37°C, (F) with 5 mM H2O2, (G) with 5% NaCl, or (H) at pH 5.5 for 24 hours and OD600 measurements were recorded every 4 hours.

101

Chapter 3

Ticks and Tularemia: Do We Know What We Don’t Know?

Published in final edited form as: Front Cell Infect Microbiol. 2019 May 8; 9:146. doi:

10.3389/fcimb.2019.00146. eCollection 2019

Ticks and Tularemia: Do We Know What We Don’t Know?

Briana Zellner and Jason F Huntley*

Department of and Immunology, University of Toledo College of

Medicine and Life Sciences, Toledo, OH, United States

*Correspondence: [email protected]

102

3.1 Abstract

Francisella tularensis, the causative agent of the zoonotic disease tularemia, is characterized by high morbidity and mortality rates in over 190 different mammalian species, including humans. Based on its low infectious dose, multiple routes of infection, and ability to induce rapid and lethal disease, F. tularensis has been recognized as a severe public health threat—being designated as a NIH Category A Priority Pathogen and a CDC Tier 1 Select Agent. Despite concerns over its use as a bioweapon, most U.S. tularemia cases are tick-mediated, and ticks are believed to be the major environmental reservoir for F. tularensis in the U.S. The American dog tick () has been reported to be the primary tick vector for F. tularensis, but the lone star tick

() and other tick species also have been shown to harbor F. tularensis. This review highlights what is known, not known, and is debated, about the roles of different tick species as environmental reservoirs and transmission vectors for a variety of F. tularensis genotypes/strains.

3.2 Introduction

Francisella tularensis (Ft), the causative agent of the zoonotic disease tularemia, can infect and cause lethal disease in over 300 species, including humans (13, 29). This

Gram-negative coccobacillus is divided into three subspecies: subsp. tularensis (Type A), subsp. holarctica (Type B), and subsp. mediasiatica. However, only subsp. tularensis and subsp. holarctica are virulent for humans. A separate species, F. novicida, is

103 associated with rare disease in immunocompromised humans and is sometimes used as a surrogate to study Ft pathogenesis (281, 282). Type A strains, found solely in North

America, are the most virulent for humans with a low infectious dose (less than 10 organisms) and high mortality rates (up to 60% mortality if untreated) (283). Type B strains, although less virulent, still cause debilitating illness and are distributed throughout the northern hemisphere (281, 283). Type A strains can be further divided into three subpopulations: A1a, A1b, and A2, with A1b causing the most serious infections (284). Interest in tularemia research has increased over the past two decades due to the classification of this organism as a Tier 1 select agent by the U.S. Centers for

Disease Control, highlighting the high morbidity and mortality, ease of aerosolization, and low infectious dose of this pathogen (285). Aside from aerosolization, Ft can be transmitted to humans via the handling of infected animal carcasses, ingestion of contaminated food or water, or by bites by infected arthropods (286).

In the U.S. alone, tick-borne disease (TBD) cases have nearly doubled between 2004 and 2016, with nearly 50,000 TBD reported in 2016. TBD include , /, , , Powassan virus, and tularemia (287).

Ticks initially were discovered as a vector of tularemia in 1923 (7). In the 1960’s, 85% of all tularemia cases in the south-central U.S. were reported to be associated with tick exposure (288). More recently, approximately half of U.S. tularemia infections are tick- associated (33, 287). Ulceroglandular tularemia, the most common presentation of the disease in the U.S., typically is attributed to bites by infected arthropods (283). In the

U.S., the most commonly reported tularemia tick vectors include Amblyomma americanum, , D. occidentalis, and Dermacentor variabilis

104 (Figure 1 and Table 1). In Europe, D. reticulatus and ricinus are most frequently associated with Ft (Table 1). These ticks are members of the family (hard ticks) but variations in their host preference, geographic distribution, and habitat likely influence their ability to transmit Ft (Table 1). Despite evidence that ticks are important for both the environmental persistence and transmission of Ft (289), major questions remain about which tick species allow Ft replication and persistence, transmit Ft to naïve hosts, or prime Ft for mammalian infection. A cursory review of published literature indicates that despite over 1300 reports of Ft infections in humans and animals, less than

10% (n=141) of those examined the role of ticks – highlighting that Ft-tick studies are understudied. This review will highlight what is known, and not known, about Ft prevalence in different ticks, Ft transmission by infected ticks, Ft-tick interactions, and areas for future research.

3.3 Tularemia-Associated Tick Species, Tick Infection Rates,

and Geographic Locations

From 2004 – 2016, 2,102 tick-borne tularemia cases were reported to the U.S.

National Notifiable Disease Surveillance System (287), with the majority of infections occurring in Missouri and Arkansas (33). D. variabilis (American dog tick) and A. americanum (lone star tick), arguably the two most important tick vectors of U.S. human tularemia, both are found in Missouri and Arkansas (Figure 1) (286). Seasonal peaks of tularemia, April – August, correlate with the active period for both tick species (33). D. variabilis has the widest geographic range, being found in nearly every state east of the

105 Rocky Mountains and most of California (Figure 1). By comparison, A. americanum is confined mainly to the south-east U.S. (Figure 1). Although both D. variabilis and A. americanum are naturally infected with Ft (290, 291), percentages of ticks infected by

Type A or Type B Ft are unknown. At least three studies have demonstrated that A. americanum and D. variabilis can maintain Ft infections over winter (or for >4 months), supporting their role as environmental reservoirs (292-294).

Ticks are responsible for the majority of U.S. tularemia cases, yet Ft-tick prevalence studies indicate wide variations of infected ticks in the environment: less than 0.1% of

Minnesota D. variabilis ticks (n=2,000) were Ft-infected (295); 17% of South Dakota D. variabilis ticks were Ft-infected (296); no Mississippi A. americanum ticks (n=191) were Ft-infected (297); finally, 2% of Arkansas A. americanum ticks (n=12,845) were

Ft-infected but no Arkansas D. variabilis (n=2,201) leporispalustris

(rabbit tick; n=1,494) or (deer/blacklegged tick; vector for Lyme disease and many other pathogens; n=142) were Ft-infected (290).

Martha’s Vineyard, Massachusetts is an important site in the of U.S. tularemia, as two major outbreaks have been reported: one in 1978 affecting 15 people and a second in 2000 affecting 15 people (291). Although the cause of each outbreak remains unknown, four of the cases were linked to bites from D. variabilis ticks (291).

Analysis of >4200 Martha’s Vineyard D. variabilis ticks following the 2000 outbreak revealed that 0.7% were infected with Type A Ft but no other ticks (Ixodes dammini deer ticks; >600 tested) were infected (291). Although sequence analyses of fopA (outer membrane protein) and PPI-helicase from these Ft strains indicated that they were nearly identical to the Type A reference strain SchuS4, multiple tandem-repeat analysis of two

106 loci identified 10 unique genotypes, indicating that the degree of Ft genetic diversity on

Martha’s Vineyard is as great as the diversity found in Ft strains across and that Martha’s Vineyard has a long history of enzootic Ft transmission (291).

Between 2004 and 2007, Ft DNA was detected in 2.7% – 4.3% of Martha’s Vineyard D. variabilis ticks (>7000 ticks tested), with 13 different Ft genotypes being identified by multiple tandem-repeat analysis (298). Importantly, Ft numbers in Martha’s Vineyard infected ticks were found to range from 0 – 1011 Ft genome equivalents (ge)/tick, with half of ticks harboring 108 – 109 Ft ge/tick (299).

Dogs have been implicated to bring infected ticks into contact with humans. Early studies reported that Ft was detected in 0.4% of A. americanum ticks collected from

Arkansas dogs (290). From 2006 – 2016, 1,814 U.S. human tularemia cases were reported, 735 (40%) of which had records indicating how exposure might have occurred

(300). Of those, 24 (3.3%) were dog-related and four (0.5%) were due to tick exposure from dogs (300). In 1984, a tick-borne tularemia outbreak in twenty people from South

Dakota Indian reservations was linked to dog exposures, with 17% of D. variabilis ticks from dogs found to harbor either Type A (12.5%) or Type B (87.5%) Ft (296).

Unfortunately, clinical isolates were not collected from those patients so correlations between transmission of Type A and Type B Ft from infected ticks could not be determined.

Rabbit and lagomorph infections likely have contributed to the perpetuation of tularemia in the environment and to humans. The rabbit tick, H. leporispalustris, which is distributed across North America, likely is important for transmitting Ft to rabbits

(289, 293) and has been found to be naturally infected with Ft. Those findings are in

107 contrast to the previously referenced study that did not detect Ft in Arkansas H. leporispalustris ticks (290). Although H. leporispalustris was reported to transovarially transmit Ft to its offspring and serve as a reservoir for F. tularensis (301), H. leporispalustris has not been associated with human tularemia, questioning the relevance of H. leporispalustris to human disease. Ft also has been reported to naturally infect other ticks, including Dermacentor andersoni (Rocky Mountain wood tick; Figure 3-1)

(7), Dermacentor occidentalis (Pacific coast tick; Figure 3-1) (302), and Haemaphysalis cinnabarina ( tick) (303), but transmission of Ft from these ticks to humans needs further study.

108

Figure 3-1. U.S. geographic distribution of ticks associated with human tularemia. Data adapted from the Centers for Disease Control and Prevention, https://www.cdc.gov/ ticks/geographic_distribution.html

Ft-infected ticks are not unique to the U.S., as Ft Type B has been found in several

European tick vectors. Between 0 – 2.3% of (ornate cow tick; n=5131; Table 3.1) in Austria, Czech Republic, Germany, Poland, and Slovakia were found to be infected with Ft Type B (304). Ft was not detected in (castor bean tick; n=8994) in France, Denmark, Italy, the Netherlands, Norway, or Poland (305-

309). However, other studies noted that 0.02 – 3.8% (n=123,761) of I. ricinus were Ft

Type B infected in France, Germany, Poland, Serbia, Slovakia, and Switzerland (Table

3.1) (304, 309-314). Finally, in Slovakia, 2.8% of (bush tick; n=35) were infected with Ft Type B (304). In summary, more information is needed about tick infection rates and infected tick species in the U.S., primarily in states with high tularemia rates (e.g. Arkansas, Colorado, Kansas, Missouri, Oklahoma, South

Dakota). In addition, although more Ft-tick prevalence studies have been performed in

109 Europe and more tularemia cases occur yearly in Europe (relative to the U.S.) (315), it still is unclear what tick species transmits Ft Type B in Europe or if differences in tick species and Ft genotypes between Europe and the U.S. correlate with differences in tularemia disease severity.

Table 3.1: Ticks Associated with Human Tularemia.

aL=larvae, N=nymph, A=adult bExp= experimental, Nat=natural cUnk=unknown

110

3.4 Francisella-like Endosymbionts

As noted above, ticks harbor and transmit several human pathogens but they also are colonized with endosymbionts that are closely related to pathogenic bacteria, offer fitness advantaged to host ticks, and appear to promote pathogen acquisition/transmission

(316). Francisella-like endosymbionts (FLEs) share 16s rDNA similarity to Ft, are widely distributed in many different ticks, replicate intracellularly, can be transmitted transovarially, and appear to have evolved from pathogenic Ft strains (317-319).

However, unlike virulent Ft, FLEs do not grow in cell-free media and their transmission to and virulence in humans is unknown (309, 320). FLEs have been found in various

Dermacentor sp., as well as Hyalomma marginatum, Hyalomma aegyptium (tortoise tick), and (brown dog tick), among others (309, 320).

Importantly, one U.S. study reported that up to 60% of ticks colonized with FLEs were falsely identified as Ft-positive when using 16S rRNA PCR only (321). However, additional testing of the same ticks using a Ft multitarget TaqMan assay, specifically amplifying the insertion sequence ISFtu2, outer membrane lipoprotein tul4, and intracellular growth locus iglC, revealed that the ticks actually were not Ft-infected (321).

The wide distribution of FLEs in different tick species is further highlighted by studies finding that >94% of D. andersoni, D. variabilis, and D. occidentalis ticks from the western U.S. were positive for FLEs (322, 323). A Canadian study reported that 86% –

93% of D. variabilis and D. andersoni ticks were colonized with FLEs (324). Further afield, 50% (n=530) of Polish D. reticulatus ticks (309), 84% – 100% (n=257) of Israeli

Haemaphysalis sp. ticks (325, 326), and 3% (n=361) of Hungarian D. reticulatus ticks

111 have been found to contain FLEs (325). FLEs are not the only microbe in ticks and, interestingly, FLEs were found to comprise up to 41% of the microbiome of California D. occidentalis ticks (no ticks were positive for Ft) (327). Another study reported that Ft and FLEs accounted for ~80% (20% Ft, 60% FLE) of the midgut microbiome of D. andersoni ticks collected in Oregon and Montana (328). In summary, because of genetic similarity to virulent Ft, FLEs may have artificially inflated Ft infection rates in some of the above referenced Ft-tick prevalence studies. In addition, although it is clear that

FLEs are present in many ticks that transmit Ft, much more work is needed to determine if FLEs interact with Ft, determine if FLEs aid in Ft infection of ticks, and examine if

FLEs play important roles in Ft transmission to naïve hosts.

3.5 Transstadial Transmission of F. tularensis in Ticks

The tick lifecycle is complex, spanning up to three years, requiring a blood meal to transition from one life stage to the next (larva-nymph-adult), and requiring a final blood meal before mating and/or egg laying (286). The frequency and length of tick blood meals depends on the type of tick (soft vs. hard) and on the tick species. Important for

Ft, hard ticks (e.g. A. americanum and D. variabilis) feed for up to eleven days, taking two-thirds of the total blood volume in the last 24 – 48 hours (329). Female hard ticks feed once per life stage and die several days after oviposition. Because of this complex life cycle, there are questions about whether Ft can be transstadially-transmitted from one life stage to the next, if all tick life stages can transmit Ft to naïve hosts, or if infected female adult ticks can transovarially transmit Ft to their eggs.

112 Ft-infected D. andersoni and D. variabilis ticks have been shown to molt from larvae to nymphs and from nymphs to adults, demonstrating that transstadial transmission of Ft can occur at all life stages. Importantly, all tick life stages also were shown to transmit Ft to naive guinea pigs, hares, or rabbits (7, 330). More recent studies demonstrated the Ft

Type B attenuated live vaccine strain (LVS) was transstadially-transmitted in D. variabilis larvae to nymphs and nymphs to adults, noting that bacterial numbers decreased before each molt, then increased 3-4 logs each molt (331). However, only

22% of nymphs maintained LVS infection through day 28 post-infection (close to molting), 25% of those infected nymphs survived molting, and only 25% of LVS-infected adult ticks maintained LVS through day 165 post-infection. Because D. variabilis in those studies were artificially fed using capillary tubes, it is difficult to determine if most

Ft infections are cleared in naturally-infected ticks or if natural Ft infections negatively impact molting (331). The authors of that study also capillary-fed A. americanum with

LVS, observing transstadial transmission between all life stages, LVS decreases before molting, LVS increases after molting, and low maintenance of LVS over time (294). By comparison, one study noted very high transstadial transmission rates of virulent Ft strains from D. variabilis larvae to nymph (fed on infected mice): Type A1b (93.3%),

Type A2 (96.7%), and Type B (100%) (332).

Although older studies detected Ft in the eggs of infected adult female D. variabilis ticks and noted that oviposition was unimpaired by infected ticks (333), neither study examined if Ft was present in hatched larvae. More recently, transovarial transmission from capillary-infected A. americanum or D. variabilis ticks was not observed (294, 331).

Additionally, while Ft Type B was detected in oocytes of infected adult female D.

113 reticulatus and I. ricinus ticks fed on infected guinea pigs, transovarial transmission was not observed (334). Taken together, it appears that Ft can be transstadially-transmitted between all tick life stages and all tick life stages have been reported to transmit Ft to naïve hosts. However, transovarial transmission of virulent Ft should be examined, more studies are needed to understand if naturally-infected ticks clear Ft over time, and additional studies are needed to examine transmission of virulent Ft by infected ticks to naive hosts.

3.6 F. tularensis-Tick Interactions

Questions about potential negative impacts of Ft infections on ticks and whether ticks restrict Ft replication/persistence have been examined in a number of studies in D. variabilis. Whereas environmentally-collected ticks have a number of limitations (e.g., low infection rate), Ft-tick infection experiments in the laboratory have their own limitations, including targeting biologically-relevant Ft numbers in ticks, selecting the appropriate tick life stage to infect, and selecting the tick infection model (e.g. infected mouse vs. capillary feeding). These limitations are further confounded by ticks requiring

3 to 7 days to feed to repletion and mice succumbing to virulent Ft infection within 4 to 5 days (335). Although blood meal feeding mimics natural infection cues, bacterial numbers can be highly variable (292, 335) and some ticks die while feeding on an infected host, suggesting that Ft has negative impacts on ticks. Uninfected D. variabilis nymphs have been reported to survive significantly better (58.5% survival) than nymphs infected with Ft Type A2 (11.6% survival) or Type B (29.8% survival). Interestingly, no significant difference in survival of uninfected and Type A1b-infected D. variabilis was

114 observed (332). In contrast, another study noted that A1b-infected adult D. variabilis ticks had significantly lower survival rates (82% survival) than uninfected (92% survival), A2-infected (95% survival), or Type B-infected ticks (90% survival) (336). Ft

Type A1a also appears to negatively impact tick survival, as only 11% of A1a-infected D. variabilis collected from Martha’s Vineyard survived six months, compared with 52% survival for uninfected ticks (337). By comparison, an older study found no significant difference in mortality rates between uninfected and Ft-infected D. variabilis (333).

Some evidence indicates that high bacterial numbers or rapid bacterial replication (2- to

5-log increases in Ft Type A2 over 65 days) in ticks correlated with tick mortality (332).

In contrast, another study found virtually no difference in survival rates for D. variabilis that were either uninfected (65% survival) of capillary-infected with Ft LVS (63% survival) (331). Considering the wide variations in reported survival rates for both uninfected (52% to 92%) and Ft-infected ticks (11% to 95% survival), it is difficult to conclude if Ft infections negatively impact ticks or if these results hold true for other tick species, including A. americanum.

With respect to Ft numbers and replication in ticks, two studies reported that Ft LVS numbers decline in capillary-tube fed D. variabilis or A. americanum ticks (294, 331).

For naturally-infected ticks, it has been speculated that anti-Ft antibodies from the mammalian host may limit bacterial replication/survival in ticks, as D. variabilis ticks fed on an immune host cleared Ft infections (333). Conversely, it also has been reported that

Ft-infected A. americanum nymphs fed on hyperimmune dogs, rabbits, or rats retained Ft infections (292, 293). A fairly recent study reported reproducible tick infections by placing D. variabilis nymphs onto uninfected mice for approx. 77 h, retro-orbitally

115 infecting those mice with 106 to 108 CFU of Ft LVS, and harvesting ticks 24 h later. In that study, mouse blood CFU/ml directly correlated with CFU/tick, Ft numbers increased over time in D. variabilis (after an initial decrease), and Ft doses <106 CFU resulted in less efficient infection of and maintenance in ticks through molting to adult, indicating that a threshold of F. tularensis is needed to infect D. variabilis (335). Similarly, another study noted that ticks must feed on an infected host during peak bacteremia to become infected (333). Finally, another study concluded that, as compared to direct injections of

Ft, natural infections of ticks (feeding on an infected host) are necessary for proper colonization and bacterial dissemination (334).

In theory, capillary tube feeding or direct injection of Ft into ticks can produce more consistent, standardized infections, but these methods lack natural infection cues (294,

331). In one capillary feeding study, D. variabilis nymphs were capillary fed 107

CFU/ml Ft LVS. One day later, only 30% of nymphs were infected and, of those, bacterial numbers were 4-logs less than the infectious dose (331). In direct injection studies, <2 CFU Ft LVS delivered into the hemocoel of D. variabilis adults resulted in

~40% infection rate, whereas similar A. americanum adult injections did not establish infections (294, 331).

Compared to burgdorferi, which is found exclusively in I. scapularis midguts (338), Ft has been reported to quickly (<24 h) disseminate from the gut to hemolymph and salivary glands of capillary-fed A. americanum ticks (294). Ft dissemination is further supported by one study noting that Ft migrated to the salivary glands of D. reticulatus and I. ricinus six days after ticks were removed from infected guinea pigs (334) and another study noting that capillary-fed D. variabilis maintained Ft

116 in their guts for up to 21 days before the bacteria spread to hemolymph and salivary glands (331). Conversely, a separate study noted that Ft did not disseminate to D. variabilis salivary glands (335).

Transmission efficiency of Ft from infected ticks to hosts appears to be dependent on many factors, including the Ft strain, tick species, tick attachment efficiency, and feeding time. Results from one study suggested that Ft infection decreases tick attachment rates to naïve mice, with 96% attachment for uninfected D. variabilis adults, 86% attachment for A1b-infected, 58% attachment for A2-infected, and 52% attachment for Type B- infected ticks. In addition, Ft infection appeared to limit tick feeding, with 46% of uninfected ticks feeding to repletion, and only 23% of A1b-infected ticks feeding to repletion (336). In another study, 55% of D. variabilis ticks on uninfected mice fed to repletion, compared with only 3.7% of ticks feeding to repletion on A2-infected mice, and most of the ticks dying while feeding on A1b- and A2-infected mice (332). Although those results indicate that Ft infections alter tick feeding behaviors, other variables could account for these findings, including the reported preference of adult D. variabilis ticks for larger hosts (339). Interestingly, one study noted that different Ft genotypes may be transmitted to naïve hosts at different frequencies (using infected D. variabilis): Type

A1b transmitted to 67% of mice; Type A2 transmitted to 89% of mice; and Type B transmitted to 58% of mice (336). Differences in transmission could not be correlated to differences in bacterial numbers in ticks, as bacterial burdens in A1b-infected ticks (>109

CFU) were significantly higher than bacterial burdens in A2- or Type B-infected ticks

(~108 CFU) (336). Given these conflicting findings, more studies are needed to better understand if Ft infections negatively impact different tick species, if Ft infections alter

117 tick feeding behaviors, and if Ft genotypes are transmitted to naïve hosts at different frequencies.

3.7 Conclusions

A large number of complex studies have been performed to understand which tick vectors are infected with Ft, which ticks are most likely to transmit Ft, which Ft genotypes are most likely to be tick-transmitted, what tick life stage is the most infectious, or if Ft infections have impacts on ticks. The majority of Ft-tick studies have focused on D. variabilis which, in the U.S., has the widest geographic range (Figure 1) and is most often associated with human tularemia. The second major tick vector for

U.S. tularemia appears to be A. americanum. However, a number of other ticks, including those that feed primarily on small , likely play important roles in Ft environmental persistence (Table 1). FLEs are a relatively new research field and much remains to be learned about how they interact with virulent Ft, if they provide metabolites/nutrients that support Ft persistence/replication in ticks, or if they contribute to transmission and disease. All tick life stages appear to support Ft and Ft can be transstadially transmitted from larva-nymph-adult. However, more studies are needed to understand if naturally-infected ticks can control or restrict Ft persistence/replication or if

Ft infections have negative consequences on infected ticks. Finally, although it is clear that Ft is transmitted from infected ticks to naïve hosts, detailed studies are needed to understand if Ft genotypes are transmitted at different efficiencies.

In many cases, it is difficult to directly compare the highlighted studies because of differences in tick infection techniques (e.g. feeding on infected animals, capillary tube

118 feeding, intrahemocoelic injection), environmental vs. laboratory infections, animals that transmitted Ft to ticks (e.g. mice, guinea pigs, rabbits, dogs), tick life stage used (larvae, nymph, adult), Ft infectious dose, and Ft genotypes/strains used (A1a, A1b, A2, B, LVS).

Given these differences, future studies should directly compare bacterial replication in different ticks over time, transstadial transmission efficiency in different ticks, survival rates of different infected ticks, and Ft transmission to naïve hosts for D. variabilis and A. americanum, as well as other relevant ticks.

Finally, very little is known about Ft genes/proteins required for tick infection, persistence/replication in ticks, and transmission to naïve hosts. To our knowledge, only one study investigated the ability of a Ft mutant, a ΔpurMCD strain, to infect and replicate in ticks (335). Although ΔpurMCD is avirulent in mice, it successfully colonized D. variabilis but was unable to persist in these ticks through the molt to the adult stage (335). This finding indicated that, similar to biosynthetic pathways required for mammalian infections, the ability of Ft to synthesize purines is essential for replication in ticks. Studies to identify Ft genes/proteins required for persistence/replication in ticks, or the development of small molecule inhibitors that block Ft persistence/replication in ticks, could be important for reducing bacterial numbers in the environment, limiting enzootic episodes, and reducing human tularemia infections.

119

Chapter 4

Type IV Pili and Virulence in Francisella tularensis

4.1 Introduction to Type IV Pili

Type IV pili (TFP) are long, flexible, surface-exposed filaments that are expressed by a wide range of pathogens and have been demonstrated to play important roles in host cell adhesion, biofilm formation, twitching motility, and uptake of DNA (340, 341). TFP are composed of a major pilin protein and several minor pilin proteins that function as peptidases, ATPases, and secretins, and are involved in biosynthesis, assembly, secretion, and retraction (342). Most of what we know regarding TFP in Gram-negative pathogens has been studied in organisms such as P. aeruginosa, V. cholerae, and N. gonorrhoeae, and in contrast, very little is known regarding TFP in intracellular pathogens. Disruption of TFP formation leads to reduced virulence in several Gram- negative human pathogens (343-345), suggesting a similar role for TFP in F. tularensis.

Indeed, TFP orthologs have been identified in F. tularensis and, similar to other

Gram-negative bacteria, are found in clusters on the genome, pilNOPQ, pilFG, and pilE1-E3, and as individual genes scattered throughout the genome, pilT, pilD, pilE4, pilE5, and pilE6 (346, 347). However, genomic sequencing has revealed that F.

120 tularensis TFP composition is strain-dependent – while both F. tularensis subsp. tularensis and F. novicida have pilE1-E6, F. tularensis subsp. holarctica has truncations in pilE1-E3 and LVS completely lacks pilE1, pilT is truncated in F. tularensis subsp. holarctica strains, and pilE4 is truncated in Type A strains (104, 342, 348).

Pili were first detected on the surface of F. tularensis LVS in 2004 (346), and were later found to be present in F. tularensis SchuS4 and F. novicida (348, 349). Gil et al. reported that when LVS was grown on solid medium, the bacteria produced thick, horn- like protrusions that were not seen in liquid culture. In fact, long, thin fibers, characteristic of traditional TFP were reported only when LVS was grown in MH broth or in liquid Chamberlain’s Defined Medium (CDM), indicating a correlation between growth conditions and pilin formation (346). Since their discovery, several studies have demonstrated the importance of TFP components to F. tularensis virulence. Deletion of surface components pilE5 or pilE6, or the inner membrane ATPase pilT, resulted in LVS attenuation in subcutaneous or intradermal mouse models of infection, although none of the three contributed to SchuS4 virulence (348). Another surface component, PilE1, was demonstrated to be essential for full virulence of Type A and B F. tularensis strains in vivo (350, 351). Several pilus assembly proteins are also required for virulence, as deletion of either pilC or pilQ resulted in SchuS4 attenuation in subcutaneous mouse infections (351). Interestingly, pilE4, the major pilin component of F. tularensis and F. novicida does not appear to be required for virulence, though it is essential for surface fiber expression (348).

The function of F. tularensis pili is not clearly understood although several studies have investigated roles in attachment. While one study found that pil mutants exhibited

121 increased attachment to J774A.1 cells (348), a second study reported a defect in attachment of pilT (predicted involved in pilus retraction) and pilF (predicted involved in pilus extension) mutants (352). Surprisingly, LVS pilT transposon mutants expressed fewer surface fibers indicating that the truncated PilT in LVS may have a distinct role in comparison to prototypic PilT which is involved in pilus retraction. In addition, several

F. tularensis TFP genes are homologous to genes involved in Type II Secretion (T2S).

However, T2S only has been observed in the nonpathogenic F. novicida, which was shown to secrete seven different proteins (353). Interestingly, secretion mutants of F. novicida exhibited increased virulence in mouse models, suggesting a possible explanation for the absence of T2S in virulent F. tularensis strains, although the mechanism is not fully-understood (354).

Taken together, although TFP are known virulence factors in many different bacteria, much remains to be learned about their roles in F. tularensis pathogenicity. Previous studies have demonstrated a link between F. tularensis TFP and virulence, but data interpretation is complicated as a result of the differences in TFP gene composition between subspecies. It is important to note that any studies involving TFP should include

Type A and Type B strains, as there are clear differences in contributions to virulence and function.

122

4.2 Explanation of Second Research Project

When the Francisella genome first became available, it was noted that few genes shared homology with known virulence factors of other organisms, making it a very interesting pathogen to study (347). One of the goals of our laboratory is to identify and characterize novel F. tularensis virulence factors. As mentioned above, we previously identified over 50 different F. tularensis DsbA substrates using a molecular trapping approach and hypothesized that a majority of these DsbA substrates were required for F. tularensis virulence. One of those DsbA substrates, FTL1695, contains a Type IV pili conserved domain and shares homology with the minor pilus assembly protein PilW (Fig.

4-1).

Figure 4-1. FTL1695 contains a putative type IV pili domain. NCBI conserved domain search results for F. tularensis FTL1695.

123

The bioinformatic prediction program PilFind provided further evidence that

FTL1695 is a TFP, based on the identification of a single transmembrane domain (TMD) located downstream of a prepilin peptidase cleavage site (Table 4-1). We used additional bioinformatic programs to predict where FTL1695 may localize within the bacteria and, while bacterial subcellular localization predictions were inconclusive, the results suggested that FTL1695 may be IM-bound due to the presence of a short TMD and the absence of β-barrels (Table 4.1). Additional support for the IM-localization of FTL1695 comes from the P. aeruginosa homolog, PilW, that previously was predicted to be IM- bound with the majority of the protein present in the cytoplasm (355). However, the exact role of PilW in pathogenic bacteria is somewhat elusive. While P. aeruginosa

PilW has been demonstrated to be important for pilus assembly and virulence (355, 356),

N. gonorrhoeae PilW only appears to be involved in the pilus fiber stability after assembly is complete (343).

Given that FTL1695 had not been previously studied and that many DsbA substrates appear to play important roles in F. tularensis virulence, the goal of this second research project was to determine if FTL1695 is an F. tularensis virulence factor and what functional role it plays.

124 Table 4.1: Bioinformatic analyses of FTL1695

a PSORTb version 3.0.2 bacterial protein subcellular localization prediction program (http://www.psort.org/psortb/). Protein localization to the outer membrane, cytoplasm, cytoplasmic membrane, periplasm, or extracellular space based on scores > 7.5 (considered significant). Unknown (Unk) localization indicated by scores < 7.5. b BOMP beta-barrel integral outer membrane protein prediction program (http://services.cbu.uib.no/tools/bomp). Scores > 2 considered significant. c LipoP 1.0 lipoprotein signal peptide prediction program (http://www.cbs.dtu.dk/services/LipoP/). SpI indicates predicted signal peptide for signal peptidase I cleavage, with log-odds score, and predicted cleavage site. Lipoprotein signal peptidase II cleavage prediction was not significant. d SignalP 4.1 signal peptide prediction program (http://www.cbs.dtu.dk/services/SignalP- 4.1/). D score >0.450 is considered significant. Position indicates signal peptide residues. e PilFind version 1.0 Type IV pilin prediction program (http://signalfind.org/pilfind.html). Type IV pilin identification based on the presence of a single transmembrane domain within 50 amino acids of the N-terminus and just downstream of a prepilin peptidase cleavage site.

125

4.3 Results and Future Directions

To determine if FTL1695 was required for F. tularensis virulence, we first generated an isogenic deletion mutant in F. tularensis LVS, hereafter known as

ΔFTL1695, using the same procedure described in Chapter 2. After confirming that

ΔFTL1695 did not have an inherent growth defect in sMHA (Fig. 4-2), we assessed attenuation in a mouse pulmonary infection model, described in Chapter 2.

Figure 4-2. ΔFTL1695 does not have an inherent growth defect. WT LVS and

ΔFTL1695 were grown in 100 ml sMHA for 28 hours at 37°C. OD600 was recorded every 4 hours through 28 hours.

126 Following intranasal administration of approx. 1×105 CFU of either WT LVS or

ΔFTL1695, ΔFTL1695-infected mice were observed to have an increased median time- to-death (10.3 days) and higher survival rate (20%), compared to WT LVS-infected mice

(median time-to-death 8.6 days; 0% survival; Fig. 4-3A). Admittedly, 1×105 CFU is approx. 100× the LD50 of LVS, so to further characterize the attenuation of ΔFTL1695, we performed a second infection experiment, where mice were infected with approx.

1×103 CFU of either WT LVS or ΔFTL1695. ΔFTL1695-infected mice demonstrated a two-day delayed time-to-death and 80% survival, while only 20% of LVS-infected mice survived (Fig. 4-3B). Finally, to determine whether ΔFTL1695 could be used as a live attenuated vaccine, we boosted the surviving mice from Fig. 4-3B with 5×105 CFU

ΔFTL1695 after 28 days (Fig. 4-3C). Approx. three weeks after the boost, mice were challenged with 29 CFU of the fully-virulent F. tularensis strain SchuS4 (in the BSL3).

While immunization with ΔFTL1695 appeared to be protective against SchuS4 (100% survival), the results must be interpreted with caution because 50% of non-immunized mice survived challenge (Fig. 4-3D). Although these protection results are not significant, they indicate that FTL1695 plays a minor role in F. tularensis virulence.

127

Figure 4-3. FTL1695 plays a role in F. tularensis virulence. Groups of 5 C3H/HeN mice were intranasally-infected with: (A) 105 CFU of either WT LVS or ΔFTL1695; or (B) 103 CFU of either WT LVS or ΔFTL1695. Animal health was monitored through day 21 post-infection. (C) Animals from (B) received a 105 CFU boost of ΔFTL1695 at day 28 post-infection. Animal health was monitored through day 23 post-boost. (D) Immunized mice from (C) and 3 naïve mice were intranasally-challenged with 29 CFU of fully-virulent SchuS4 (in the BSL3). Animal health was monitored through day 14 post- challenge.

Future studies for this project include potentially characterizing the function of

FTL1695, the role of FTL1695 in TFP formation, and studying how FTL1695 contributes to F. tularensis pathogenesis. As noted above, PilW, a minor pilin protein, has been demonstrated to be important for TFP assembly in P. aeruginosa (356). To determine if

FTL1695 is involved in pilus assembly, we could examine WT LVS and ΔFTL1695 by

EM and note any differences in the presence of TFP on the surface of the bacteria.

128 However, based on the Gil et al. studies highlighted above, we would need to examine bacteria from both solid and liquid media. Our lab has well-established protocols for fixing bacterial samples and imaging by EM to detect changes in surface structures (118).

PilW is responsible for repressing twitching motility in P. aeruginosa to promote biofilm formation over a planktonic lifestyle, as ΔpilW mutants are hypermotile (355).

Because F. tularensis is non-motile, and Type A and B strains form poor biofilms, it is likely that PilW has a unique function in this pathogen. Out of all the known functions of

TFP – twitching motility, uptake of DNA, attachment to host cells, protein secretion, and biofilm formation – the most likely candidate for F. tularensis is attachment to host cells.

Future experiments should investigate the effects of ΔFTL1695 on attachment to macrophages or other host cells in vitro. To specifically determine attachment efficiency, rather than bacterial internalization, macrophages would first be treated with cytochalasin

D to prevent internalization of bacteria (352). Following infection of the macrophage monolayer with WT LVS or ΔFTL1695 and incubation of the macrophages for 2 hours, cultures would be extensively washed, then lysed with saponin, and bacterial CFUs enumerated.

Finally, it would be interesting to determine if pilW is required for virulence in

SchuS4 (FTT1627c). Future students could generate the SchuS4 ΔFTT1627c mutant, assess attenuation in mice, and examine ΔFTT1627c attachment to host cells to determine if FTL1695 and FTT1627c have similar roles.

129

Chapter 5

Summary and Future Directions of FTL1678 Project

My dissertation research began with the goal of characterizing a previously unstudied hypothetical protein, FTL1678, and determine its role in F. tularensis virulence. A conserved domain search indicated that FTL1678 may function as an L,D- carboxypeptidase, suggested PG recycling activity. PG, a mesh-like structure surrounding the cytoplasmic membrane, is well-known to function as a protective barrier for bacteria against osmotic stress and external insults. Thus, PG homeostasis is extremely critical for bacteria to remain viable and PG homeostasis becomes even more critical for intracellular pathogens that are exposed to ROS, RNS, low pH, etc.

Consequently, PG is the target of several antibiotics and host defense mechanisms (e.g.,

β-lactams and lysozyme). Many successful pathogens modify their PG structures to block antibacterial activity and/or encode robust synthesis and recycling pathways to repair damaged cell wall material. PG synthesis and recycling pathways are extremely important, as mutation or deletion of any of these enzymes can result in altered bacterial cell morphology, reduced virulence, and inefficient cell division.

130 Most of what is known regarding PG synthesis, recycling, and damage repair comes from studies on nonpathogenic E. coli and very little is known about Gram- negative intracellular pathogens such as F. tularensis. Over 50 different proteins, mostly cytoplasmic, are involved in E. coli PG synthesis and recycling (215) and, previous to this study, only one putative PG-related protein had been identified in F. tularensis (146,

147). However, those studies did not directly examine PG-related activity of that protein.

To our knowledge, this is the first study to directly confirm that an F. tularensis protein is involved in PG modification.

The majority of my efforts were focused on studying FTL1678, which I hypothesized encoded a periplasmic L,D-carboxypeptidase (LdcA) that is required for F. tularensis virulence. I tested this hypothesis by using the following techniques: (1)

Bioinformatic analyses indicated that FTL1678 contains a putative LdcA domain; (2)

Enzymatic assays with purified PG and recombinant FTL1678 demonstrated that

FTL1678 functions as both an L,D-carboxypeptidase and L,D-endopeptidase, resulting in cleavage of PG pentapeptides and tetrapeptides to tripeptides; (3) Subcellular fractionation of F. tularensis, indicated that FTL1678 is outer membrane-associated.

This finding distinguishes F. tularensis from most other Gram-negative LdcAs orthologs that are located in the cytoplasm; (4) Membrane integrity and electron microscopy visualizations revealed that ΔFTL1678 has a thicker OM and PG layer (compared to WT

F. tularensis), further supporting the role of FTL1678 in PG recycling; (5) ΔFTL1678 was severely attenuated in macrophages and in a mouse pulmonary infection model (up to 109 CFU); (6) ΔFTL1678 complementation with either FTL1678 or a C. jejuni LdcA ortholog (proposed to be periplasmic) restored virulence to WT levels; (7) Immunization

131 with ΔFTL1678 protected mice against pulmonary challenge with fully-virulent F. tularensis strain SchuS4 (in the BSL3).

Although the studies in my dissertation project are mostly complete, there are several experiments we are currently working on in anticipation of reviewer comments to the submitted manuscript:

1. We have shown that immunization of mice with ΔFTL1678 protects against

challenge with fully-virulent F. tularensis SchuS4. However, we have not yet

characterized the immune responses that confer protection against challenge.

Current studies involve immunizing mice with either WT LVS or ΔFTL1678

and collecting spleens and sera two to three weeks after immunization to

compare IFN-γ levels and antibody isotypes and titers induced by each

bacterial strain. From spleens, we prepare splenocytes and either leave

unstimulated (as a control), or stimulate with ethanol-inactivated LVS or

SchuS4, collect splenocyte culture supernatants 24- and 48-hours after

stimulation, and then quantitate IFN-γ production by ELISA. Similarly, we

analyze antibody isotype and titers from mice sera using ELISA. We follow a

prime-boost immunization regimen, with the immunization boost occurring 4

weeks after the initial prime, and SchuS4 challenge occurring 3 weeks after

the boost. In these experiments, we have collected spleens and sera from

immunized mice on week 4 (after prime) and at week 7 (after prime and

boost) to assess immune response differences (e.g., IFN-γ and antibody

isotype and titer) between our ΔFTL1678 live attenuated strain and LVS. We

132 will compare our immune response results to recently-published data on F.

tularensis immune correlates of protection (190).

2. We demonstrated above that the C. jejuni LdcA homolog Pgp2 (predicted to

be periplasmic) complements ΔFTL1678 and fully restores virulence to WT

LVS levels (in pulmonary infection studies). We hypothesized that

periplasmic localization of FTL1678 is important for F. tularensis virulence

and that the periplasmic localization of F. tularensis LdcA provides an

advantage to quickly and efficiently repair damaged PG. However, to further

confirm this, we currently are developing an ΔFTL1678 strain complemented

with the E. coli LdcA gene (cytoplasmic localization). To examine all

potential possibilities, I will be complementing ΔFTL1678 with both the

native E. coli LdcA (cytoplasmic) and with an E. coli LdcA harboring the F.

tularensis signal sequence (for periplasmic localization). These studies will

allow us to determine if periplasmic localization of LdcA is required for F.

tularensis virulence.

3. Prototypic LdcA possess a catalytic triad, Ser-Glu-His, which is required for

function of the enzyme. We demonstrated that these residues were not

required for F. tularensis virulence in vivo. However, we would like to

confirm the putative catalytic triad in vitro by generating individual point

mutants, S134A, E239A, and H308A, in recombinant FTL1678. We will use

site-directed mutagenesis to individually mutate each of the three amino acids

in the catalytic triad to alanine. We will then produce and purify each mutated

recombinant protein and repeat the functional assays using different PG

133 substrates (in collaboration with Dominique Mengin-Lecreulx at the

University of Paris).

134 References

1. McCoy G. 1910. Plague among Ground Squirrels in America. J Hyg (Lond)

10:589-601.

2. McCoy G and Chapin CW. 1912. Further Observations on a Plague-Like Disease

of Rodents with a Preliminary Note on the Causative Agent, Bacterium tularense

The Journal of Infectious Diseases 10:61-72.

3. Wherry W and Lamb BH. 1914. Infection of man with Bacterium tularense. J

Infect Dis 15:331-340.

4. Sjostedt A. 2007. Tularemia: history, epidemiology, pathogen physiology, and

clinical manifestations. Ann N Y Acad Sci 1105:1-29.

5. Francis E. 1919. Deer-Fly Fever, or Pahvant Valley Plague: A disease of man

hitherto unknown etiology Public Health Reports 34:2061-2062.

6. Francis E. 1925. Tularemia. JAMA 84:1243-1250.

7. Parker R, Spencer R, and Francis E. 1924. Tularemia Infection in Ticks of the

Species Dermacentor andersoni Stiles in the Bitterroot Valley, Montana. Sage

Publications I,

8. Francis E. 1928. Symptoms, diagnosis, and pathology of tularemia. JAMA

91:1151-1161.

9. Ellis J, Oyston PC, Green M, and Titball RW. 2002. Tularemia. Clin Microbiol

Rev 15:631-46.

135 10. Kaufmann A, Meltzer MI, and Schmid GP. 1997. The economic impact of a

bioterrorist attack: are prevention and postattack intervention programs

justifiable? Emerg Infect Dis 3:83-94.

11. Harris S. 1992. Japanese biological warfare research on humans: a case study of

microbiology and ethics. Ann N Y Acad Sci 666:21-52.

12. Alibek K. 1999. Biohazard. Random House, New York, New York.

13. Dennis D, Inglesby TV, Henderson DA, Bartlett JG, Ascher MS, Eitzen E, Fine

AD, Friedlander AM, Hauer J, Layton M, Lillibridge SR, McDade JE, Osterholm

MT, O'Toole T, Parker G, Perl TM, Russell PK, and Tonat K. 2001. Tularemia as

a biological weapon: medical and public health management. JAMA 285:2763-73.

14. Petersen J and Schriefer ME. 2005. Tularemia: emergence/re-emergence. Vet Res

36:455-67.

15. McLendon M, Apicella MA, and Allen LA. 2006. Francisella tularensis:

, genetics, and Immunopathogenesis of a potential agent of biowarfare.

Annu Rev Microbiol 60:167-85.

16. Forsman M, Sandstrom G, and Sjostedt A. 1994. Analysis of 16S ribosomal DNA

sequences of Francisella strains and utilization for determination of the

phylogeny of the genus and for identification of strains by PCR. Int J Syst

Bacteriol 44:38-46.

17. Challacombe J, Petersen JM, Gallegos-Graves V, Hodge D, Pillai S, and Kuske

CR. 2017. Whole-Genome Relationships among Francisella Bacteria of Diverse

Origins Define New Species and Provide Specific Regions for Detection. Appl

Environ Microbiol 83.

136 18. Johansson A and Petersen JM. 2010. Genotyping of Francisella tularensis, the

causative agent of tularemia. J AOAC Int 93:1930-43.

19. Kingry L and Petersen JM. 2014. Comparative review of Francisella tularensis

and . Front Cell Infect Microbiol 4:35.

20. Larson C, Wicht W, and Jellison WL. 1955. A new organism resembling P.

tularensis isolated from water. Public Health Rep 70:253-8.

21. Huber B, Escudero R, Busse HJ, Seibold E, Scholz HC, Anda P, Kampfer P, and

Splettstoesser WD. 2010. Description of Francisella hispaniensis sp. nov.,

isolated from human blood, reclassification of Francisella novicida (Larson et al.

1955) Olsufiev et al. 1959 as Francisella tularensis subsp. novicida comb. nov.

and emended description of the genus Francisella. Int J Syst Evol Microbiol

60:1887-96.

22. Johansson A, Celli J, Conlan W, Elkins KL, Forsman M, Keim PS, Larsson P,

Manoil C, Nano FE, Petersen JM, and Sjostedt A. 2010. Objections to the transfer

of Francisella novicida to the subspecies rank of Francisella tularensis. Int J Syst

Evol Microbiol 60:1717-8; author reply 1718-20.

23. Johansson A and Petersen JM. 2010. Genotyping of Francisella tularensis, the

causative agent of tularemia. J AOAC Int 93:1930-43.

24. Kugeler K, Mead PS, Janusz AM, Staples JE, Kubota KA, Chalcraft LG, and

Petersen JM. 2009. Molecular Epidemiology of Francisella tularensis in the

United States. Clin Infect Dis 48:863-70.

25. Vogler A, Birdsell D, Price LB, Bowers JR, Beckstrom-Sternberg SM, Auerbach

RK, Beckstrom-Sternberg JS, Johansson A, Clare A, Buchhagen JL, Petersen JM,

137 Pearson T, Vaissaire J, Dempsey MP, Foxall P, Engelthaler DM, Wagner DM,

and Keim P. 2009. Phylogeography of Francisella tularensis: global expansion of

a highly fit clone. J Bacteriol 191:2474-84.

26. Eigelsbach HT and Downs CM. 1961. Prophylactic effectiveness of live and

killed tularemia vaccines. I. Production of vaccine and evaluation in the white

mouse and guinea pig. J Immunol 87:415-25.

27. Hornick RB and Eigelsbach HT. 1966. Aerogenic immunization of man with live

tularemia vaccine. Bacteriol Rev 30:532-538.

28. Rohmer L, Brittnacher M, Svensson K, Buckley D, Haugen E, Zhou Y, Chang J,

Levy R, Hayden H, Forsman M, Olson M, Johansson A, Kaul R, and Miller SI.

2006. Potential source of Francisella tularensis live vaccine strain attenuation

determined by genome comparison. Infect Immun 74:6895-906.

29. Keim P, Johansson A, and Wagner DM. 2007. Molecular epidemiology,

evolution, and ecology of Francisella. Ann N Y Acad Sci 1105:30-66.

30. Petersen J, Mead PS, and Schriefer ME. 2009. Francisella tularensis: an

arthropod-borne pathogen. Vet Res 40:7.

31. Francis E. 1937. Sources of infection and seasonal incidence of tularaemia in

man. Public Health Rep 52:103-113.

32. Hennebique A, Boisset S, and Maurin M. 2019. Tularemia as a waterborne

disease: a review. Emerg Microbes Infect 8:1027-1042.

33. Eisen L. 2007. A call for renewed research on tick-borne Francisella tularensis in

the Arkansas-Missouri primary national focus of tularemia in humans. J Med

Entomol 44:389-97.

138 34. Staples J, Kubota KA, Chalcraft LG, Mead PS, and Petersen JM. 2006.

Epidemiologic and molecular analysis of human tularemia, United States, 1964-

2004. Emerg Infect Dis 12:1113-8.

35. Forestal C, Malik M, Catlett SV, Savitt AG, Benach JL, Sellati TJ, and Furie MB.

2007. Francisella tularensis has a significant extracellular phase in infected mice.

J Infect Dis 196:134-7.

36. Hall J, Woolard MD, Gunn BM, Craven RR, Taft-Benz S, Frelinger JA, and

Kawula TH. 2008. Infected-host-cell repertoire and cellular response in the lung

following inhalation of Francisella tularensis Schu S4, LVS, or U112. Infect

Immun 76:5843-52.

37. Clemens D, Lee BY, and Horwitz MA. 2005. Francisella tularensis enters

macrophages via a novel process involving pseudopod loops. Infect Immun

73:5892-902.

38. Celli J and Zahrt TC. 2013. Mechanisms of Francisella tularensis intracellular

pathogenesis. Cold Spring Harb Perspect Med 3:a010314.

39. Balagopal A, MacFarlane AS, Mohapatra N, Soni S, Gunn JS, and Schlesinger

LS. 2006. Characterization of the receptor-ligand pathways important for entry

and survival of Francisella tularensis in human macrophages. Infect Immun

74:5114-25.

40. Schulert G and Allen LA. 2006. Differential infection of mononuclear phagocytes

by Francisella tularensis: role of the macrophage mannose receptor. J Leukoc

Biol 80:563-71.

139 41. Pierini L. 2006. Uptake of serum-opsonized Francisella tularensis by

macrophages can be mediated by class A scavenger receptors. Cell Microbiol

8:1361-70.

42. Geier H and Celli J. 2011. Phagocytic receptors dictate phagosomal escape and

intracellular proliferation of Francisella tularensis. Infect Immun 79:2204-14.

43. Clemens D, Lee BY, and Horwitz MA. 2004. Virulent and avirulent strains of

Francisella tularensis prevent acidification and maturation of their

and escape into the cytoplasm in human macrophages. Infect Immun 72:3204-17.

44. Checroun C, Wehrly TD, Fischer ER, Hayes SF, and Celli J. 2006. Autophagy-

mediated reentry of Francisella tularensis into the endocytic compartment after

cytoplasmic replication. Proc Natl Acad Sci U S A 103:14578-83.

45. Chong A, Wehrly TD, Nair V, Fischer ER, Barker JR, Klose KE, and Celli J.

2008. The early phagosomal stage of Francisella tularensis determines optimal

phagosomal escape and Francisella pathogenicity island protein expression. Infect

Immun 76:5488-99.

46. Golovliov I, Baranov V, Krocova Z, Kovarova H, and Sjostedt A. 2003. An

attenuated strain of the facultative intracellular bacterium Francisella tularensis

can escape the phagosome of monocytic cells. Infect Immun 71:5940-50.

47. Lindgren H, Golovliov I, Baranov V, Ernst RK, Telepnev M, and Sjostedt A.

2004. Factors affecting the escape of Francisella tularensis from the

phagolysosome. J Med Microbiol 53:953-8.

48. Wehrly T, Chong A, Virtaneva K, Sturdevant DE, Child R, Edwards JA, Brouwer

D, Nair V, Fischer ER, Wicke L, Curda AJ, Kupko JJ 3rd, Martens C, Crane DD,

140 Bosio CM, Porcella SF, and Celli J. 2009. Intracellular biology and virulence

determinants of Francisella tularensis revealed by transcriptional profiling inside

macrophages. Cell Microbiol 11:1128-50.

49. Barker J, Chong A, Wehrly TD, Yu JJ, Rodriguez SA, Liu J, Celli J,

Arulanandam BP, and Klose KE. 2009. The Francisella tularensis pathogenicity

island encodes a secretion system that is required for phagosome escape and

virulence. Mol Microbiol 74:1459-70.

50. Broms J, Meyer L, Lavander M, Larsson P, and Sjostedt A. 2012. DotU and

VgrG, core components of type VI secretion systems, are essential for Francisella

LVS pathogenicity. PLoS One 7:e34639.

51. Lai X, Golovliov I, and Sjostedt A. 2001. Francisella tularensis induces

cytopathogenicity and apoptosis in murine macrophages via a mechanism that

requires intracellular bacterial multiplication. Infect Immun 69:4691-4.

52. Horzempa J, O'Dee DM, Stolz DB, Franks JM, Clay D, and Nau GJ. 2011.

Invasion of erythrocytes by Francisella tularensis. J Infect Dis 204:51-9.

53. Lamps L, Havens JM, Sjostedt A, Page DL, and Scott MA. 2004. Histologic and

molecular diagnosis of tularemia: a potential agent endemic to North

America. Mod Pathol 17:489-95.

54. Baskerville A and Hambleton P. 1976. Pathogenesis and pathology of respiratory

tularaemia in the rabbit. Br J Exp Pathol 57:339-47.

55. Baskerville A, Hambleton P, and Dowsett AB. 1978. The pathology of untreated

and antibiotic-treated experimental tularaemia in monkeys. Br J Exp Pathol

59:615-23.

141 56. Telepnev M, Golovliov I, Grundstrom T, Tarnvik A, and Sjostedt A. 2003.

Francisella tularensis inhibits Toll-like receptor-mediated activation of

intracellular signalling and secretion of TNF-alpha and IL-1 from murine

macrophages. Cell Microbiol 5:41-51.

57. Bosio C, Bielefeldt-Ohmann H, and Belisle JT. 2007. Active suppression of the

pulmonary immune response by Francisella tularensis Schu4. J Immunol

178:4538-47.

58. Andersson H, Hartmanova B, Kuolee R, Ryden P, Conlan W, Chen W, and

Sjostedt A. 2006. Transcriptional profiling of host responses in mouse lungs

following aerosol infection with type A Francisella tularensis. J Med Microbiol

55:263-71.

59. Conlan J, Zhao X, Harris G, Shen H, Bolanowski M, Rietz C, Sjostedt A, and

Chen W. 2008. Molecular immunology of experimental primary tularemia in mice

infected by respiratory or intradermal routes with type A Francisella tularensis.

Mol Immunol 45:2962-9.

60. Cowley S and Elkins KL. 2011. Immunity to Francisella. Front Microbiol 2:26.

61. Elkins K, Rhinehart-Jones TR, Culkin SJ, Yee D, and Winegar RK. 1996.

Minimal requirements for murine resistance to infection with Francisella

tularensis LVS. Infect Immun 64:3288-93.

62. Leiby DA FA, Crawford RM, Schreiber RD, and Nacy CA. 1992. In vivo

modulation of the murine immune response to Francisella tularensis LVS by

administration of anticytokine antibodies. Infect Immun 60:84-9.

142 63. Elkins K, Cowley SC, and Bosio CM. 2007. Innate and adaptive immunity to

Francisella. Ann N Y Acad Sci 1105:284-324.

64. Bosio C and Dow SW. 2005. Francisella tularensis induces aberrant activation of

pulmonary dendritic cells. J Immunol 175:6792-801.

65. Shirey K, Cole LE, Keegan AD, and Vogel SN. 2008. Francisella tularensis live

vaccine strain induces macrophage alternative activation as a survival mechanism.

J Immunol 181:4159-67.

66. Lopez M, Duckett NS, Baron SD, and Metzger DW. 2004. Early activation of NK

cells after lung infection with the intracellular bacterium, Francisella tularensis

LVS. Cell Immunol 232:75-85.

67. McCaffrey R and Allen LA. 2006. Francisella tularensis LVS evades killing by

human neutrophils via inhibition of the respiratory burst and phagosome escape. J

Leukoc Biol 80:1224-30.

68. Cowley S, Meierovics AI, Frelinger JA, Iwakura Y, and Elkins KL. 2010. Lung

CD4-CD8- double-negative T cells are prominent producers of IL-17A and IFN-

gamma during primary respiratory murine infection with Francisella tularensis

live vaccine strain. J Immunol 184:5791-801.

69. Sjostedt A, Conlan JW, and North RJ. 1994. Neutrophils are critical for host

defense against primary infection with the facultative intracellular bacterium

Francisella tularensis in mice and participate in defense against reinfection. Infect

Immun 62:2779-83.

70. Conlan J, KuoLee R, Shen H, and Webb A. 2002. Different host defences are

required to protect mice from primary systemic vs pulmonary infection with the

143 facultative intracellular bacterial pathogen, Francisella tularensis LVS. Microb

Pathog 32:127-34.

71. Koskela P and Salminen A. 1985. Humoral immunity against Francisella

tularensis after natural infection. J Clin Microbiol 22:973-9.

72. Krocova Z, Hartlova A, Souckova D, Zivna L, Kroca M, Rudolf E, Macela A, and

Stulik J. 2008. Interaction of B cells with intracellular pathogen Francisella

tularensis. Microb Pathog 45:79-85.

73. Fulton K, Zhao X, Petit MD, Kilmury SL, Wolfraim LA, House RV, Sjostedt A,

and Twine SM. 2011. Immunoproteomic analysis of the human antibody response

to natural tularemia infection with Type A or Type B strains or LVS vaccination.

Int J Med Microbiol 301:591-601.

74. Waag D, McKee, KT Jr, Sandstrom G, Pratt LL, Bolt CR, England MJ, Nelson

GO, and Williams JC. 1995. Cell-mediated and humoral immune responses after

vaccination of human volunteers with the live vaccine strain of Francisella

tularensis. Clin Diagn Lab Immunol 2:143-8.

75. El Sahly H, Atmar RL, Patel SM, Wells JM, Cate T, Ho M, Guo K, Pasetti MF,

Lewis DE, Sztein MB, and Keitel WA. 2009. Safety, reactogenicity and

immunogenicity of Francisella tularensis live vaccine strain in humans. Vaccine

27:4905-11.

76. Fulop M, Mastroeni P, Green M, and Titball RW. 2001. Role of antibody to

lipopolysaccharide in protection against low- and high-virulence strains of

Francisella tularensis. Vaccine 19:4465-72.

144 77. Kirimanjeswara G, Golden JM, Bakshi CS, and Metzger DW. 2007. Prophylactic

and therapeutic use of antibodies for protection against respiratory infection with

Francisella tularensis. J Immunol 179:532-9.

78. Huntley J, Conley PG, Rasko DA, Hagman KE, Apicella MA, and Norgard MV.

2008. Native outer membrane proteins protect mice against pulmonary challenge

with virulent type A Francisella tularensis. Infect Immun 76:3664-71.

79. Crane D, Warner SL, and Bosio CM. 2009. A novel role for plasmin-mediated

degradation of opsonizing antibody in the evasion of host immunity by virulent,

but not attenuated, Francisella tularensis. J Immunol 183:4593-600.

80. Elkins K, Rhinehart-Jones T, Nacy CA, Winegar RK, and Fortier AH. 1993. T-

cell-independent resistance to infection and generation of immunity to Francisella

tularensis. Infect Immun 61:823-9.

81. Conlan J, Sjostedt A, and North RJ. 1994. CD4+ and CD8+ T-cell-dependent and

-independent host defense mechanisms can operate to control and resolve primary

and secondary Francisella tularensis LVS infection in mice. Infect Immun

62:5603-7.

82. Cowley S and Elkins KL. 2003. Multiple T cell subsets control Francisella

tularensis LVS intracellular growth without stimulation through macrophage

interferon gamma receptors. J Exp Med 198:379-89.

83. Ito M, Nishiyama K, Hyodo S, Shigeta S and Ito T. 1985. Weight reduction of

thymus and depletion of lymphocytes of T-dependent areas in peripheral

lymphoid tissues of mice infected with Francisella tularensis. Infect Immun

49:812-8.

145 84. Chen W, Kuolee R, Austin JW, Shen H, Che Y, and Conlan JW. 2005. Low dose

aerosol infection of mice with virulent type A Francisella tularensis induces

severe thymus atrophy and CD4+CD8+ thymocyte depletion. Microb Pathog

39:189-96.

85. Conlan J, Shen H, KuoLee R, Zhao X, and Chen W. 2005. Aerosol-, but not

intradermal-immunization with the live vaccine strain of Francisella tularensis

protects mice against subsequent aerosol challenge with a highly virulent type A

strain of the pathogen by an alpha beta T cell- and interferon gamma- dependent

mechanism. Vaccine 23:2477-85.

86. Salerno-Goncalves R, Hepburn MJ, Bavari S, and Sztein MB. 2009. Generation

of heterogeneous memory T cells by live attenuated tularemia vaccine in humans.

Vaccine 28:195-206.

87. Shen H, Harris G, Chen W, Sjostedt A, Ryden P, and Conlan W. 2010. Molecular

immune responses to aerosol challenge with Francisella tularensis in mice

inoculated with live vaccine candidates of varying efficacy. PLoS One 5:e13349.

88. Gal-Mor O and Finlay BB. 2006. Pathogenicity islands: a molecular toolbox for

bacterial virulence. Cell Microbiol 8:1707-19.

89. Nano F, Zhang N, Cowley SC, Klose KE, Cheung KK, Roberts MJ, Ludu JS,

Letendre GW, Meierovics AI, Stephens G, and Elkins KL. 2004. A Francisella

tularensis pathogenicity island required for intramacrophage growth. J Bacteriol

186:6430-6.

90. Santic M, Molmeret M, Klose KE, Jones S, and Kwaik YA. 2005. The

Francisella tularensis pathogenicity island protein IglC and its regulator MglA

146 are essential for modulating phagosome biogenesis and subsequent bacterial

escape into the cytoplasm. Cell Microbiol 7:969-79.

91. de Bruin O, Duplantis BN, Ludu JS, Hare RF, Nix EB, Schmerk CL, Robb CS,

Boraston AB, Hueffer K, and Nano FE. 2011. The biochemical properties of the

Francisella pathogenicity island (FPI)-encoded proteins IglA, IglB, IglC, PdpB

and DotU suggest roles in type VI secretion. Microbiology 157:3483-91.

92. de Bruin O, Ludu JS, and Nano FE. 2007. The Francisella pathogenicity island

protein IglA localizes to the bacterial cytoplasm and is needed for intracellular

growth. BMC Microbiol 7:1.

93. Clemens D, Lee BY, and Horwitz MA. 2018. The Francisella Type VI Secretion

System. Front Cell Infect Microbiol 8:121.

94. Cianfanelli F, Monlezun L, and Coulthurst SJ. 2016. Aim, Load, Fire: The Type

VI Secretion System, a Bacterial Nanoweapon. Trends Microbiol 24:51-62.

95. Clemens D, Ge P, Lee BY, Horwitz MA, and Zhou ZH. 2015. Atomic structure of

T6SS reveals interlaced array essential to function. Cell 160:940-951.

96. Eshraghi A, Kim J, Walls AC, Ledvina HE, Miller CN, Ramsey KM, Whitney JC,

Radey MC, Peterson SB, Ruhland BR, Tran BQ, Goo YA, Goodlett DR, Dove

SL, Celli J, Veesler D, and Mougous JD. 2016. Secreted Effectors Encoded

within and outside of the Francisella Pathogenicity Island Promote

Intramacrophage Growth. Cell Host Microbe 20:573-583.

97. Golovliov I, Sjostedt A, Mokrievich A, and Pavlov V. 2006. A method for allelic

replacement in Francisella tularensis FEMS Microbiol Lett 222:273-280.

147 98. Lai X and Golovliov, I. 2004. Expression of IglC is necessary for intracellular

growth and induction of apoptosis in murine macrophages by Francisella

tularensis. Microb Pathog 37:225-30.

99. Bonquist L, Lindgren H, Golovliov I, Guina T, and Sjostedt A. 2008. MglA and

Igl proteins contribute to the modulation of Francisella tularensis live vaccine

strain-containing phagosomes in murine macrophages. Infect Immun 76:3502-10.

100. Schmerk C, Duplantis BN, Howard PL, and Nano FE. 2009. A Francisella

novicida pdpA mutant exhibits limited intracellular replication and remains

associated with the lysosomal marker LAMP-1. Microbiology 155:1498-504.

101. Brodmann M, Dreier RF, Broz P, and Basler M. 2017. Francisella requires

dynamic type VI secretion system and ClpB to deliver effectors for phagosomal

escape. Nat Commun 8:15853.

102. Santic M, Molmeret M, Barker JR, Klose KE, Dekanic A, Doric M, and Kwaik

YA. 2007. A Francisella tularensis pathogenicity island protein essential for

bacterial proliferation within the host cell cytosol. Cell Microbiol 9:2391-403.

103. Robertson G, Child R, Ingle C, Celli J, and Norgard MV. 2013. IglE is an outer

membrane-associated lipoprotein essential for intracellular survival and murine

virulence of type A Francisella tularensis. Infect Immun 81:4026-40.

104. Rowe H and Huntley JF. 2015. From the Outside-In: The Francisella tularensis

Envelope and Virulence. Front Cell Infect Microbiol 5:94.

105. Delepelaire P. 2004. Type I secretion in gram-negative bacteria. Biochim Biophys

Acta 1694:149-61.

148 106. Ma D, Cook DN, Alberti M, Pon NG, Nikaido H, and Hearst JE. 1993. Molecular

cloning and characterization of acrA and acrE genes of Escherichia coli. J

Bacteriol 175:6299-313.

107. Thomas S, Holland IB, and Schmitt L. 2014. The Type 1 secretion pathway - the

hemolysin system and beyond. Biochim Biophys Acta 1843:1629-41.

108. Huntley J, Conley PG, Hagman KE, and Norgard MV. 2007. Characterization of

Francisella tularensis outer membrane proteins. J Bacteriol 189:561-74.

109. Gil H, Platz GJ, Forestal CA, Monfett M, Bakshi CS, Sellati TJ, Furie MB,

Benach JL, and Thanassi DG. 2006. Deletion of TolC orthologs in Francisella

tularensis identifies roles in multidrug resistance and virulence. Proc Natl Acad

Sci U S A 103:12897-902.

110. Kadzhaev K, Zingmark C, Golovliov I, Bolanowski M, Shen H, Conlan W, and

Sjostedt A. 2009. Identification of genes contributing to the virulence of

Francisella tularensis SCHU S4 in a mouse intradermal infection model. PLoS

One 4:e5463.

111. Doyle C, Pan JA, Mena P, Zong WX, and Thanassi DG. 2014. TolC-dependent

modulation of host cell death by the Francisella tularensis live vaccine strain.

Infect Immun 82:2068-78.

112. Platz G, Bublitz DC, Mena P, Benach JL, Furie MB, and Thanassi DG. 2010. A

tolC mutant of Francisella tularensis is hypercytotoxic compared to the wild type

and elicits increased proinflammatory responses from host cells. Infect Immun

78:1022-31.

149 113. Kopping E, Doyle CR, Sampath V, and Thanassi DG. 2019. Contributions of

TolC Orthologs to Francisella tularensis Schu S4 Multidrug Resistance,

Modulation of Host Cell Responses, and Virulence. Infect Immun 87.

114. Alqahtani M, Ma Z, Ketkar H, Suresh RV, Malik M, and Bakshi CS. 2018.

Characterization of a Unique Outer Membrane Protein Required for Oxidative

Stress Resistance and Virulence of Francisella tularensis. J Bacteriol 200.

115. Jones B, Faron M, Rasmussen JA, and Fletcher JR. 2014. Uncovering the

components of the Francisella tularensis virulence stealth strategy. Front Cell

Infect Microbiol 4:32.

116. Hood A. 1977. Virulence factors of Francisella tularensis. J Hyg (Lond) 79:47-

60.

117. Apicella M, Post DM, Fowler AC, Jones BD, Rasmussen JA, Hunt JR, Imagawa

S, Choudhury B, Inzana, TJ, Maier TM, Frank DW, Zahrt TC, Chaloner K,

Jennings MP, McLendon MK, and Gibson BW. 2010. Identification,

characterization and immunogenicity of an O-antigen capsular polysaccharide of

Francisella tularensis. PLoS One 5:e11060.

118. Wu X, Ren G, Gunning WT 3rd, Weaver DA, Kalinoski AL, Khuder SA, and

Huntley JF. 2016. FmvB: A Francisella tularensis Magnesium-Responsive Outer

Membrane Protein that Plays a Role in Virulence. PLoS One 11:e0160977.

119. Sandstrom G, Lofgren S, and Tarnvik A. 1988. A capsule-deficient mutant of

Francisella tularensis LVS exhibits enhanced sensitivity to killing by serum but

diminished sensitivity to killing by polymorphonuclear leukocytes. Infect Immun

56:1194-202.

150 120. Clay C, Soni S, Gunn JS, and Schlesinger LS. 2008. Evasion of complement-

mediated lysis and complement C3 deposition are regulated by Francisella

tularensis lipopolysaccharide O antigen. J Immunol 181:5568-78.

121. Lindemann S, Peng K, Long ME, Hunt JR, Apicella MA, Monack DM, Allen LA,

and Jones BD. 2011. Francisella tularensis Schu S4 O-antigen and capsule

biosynthesis gene mutants induce early cell death in human macrophages. Infect

Immun 79:581-94.

122. Miller S, Ernst RK, and Bader MW. 2005. LPS, TLR4 and infectious disease

diversity. Nat Rev Microbiol 3:36-46.

123. Matsuura M. 2013. Structural Modifications of Bacterial Lipopolysaccharide that

Facilitate Gram-Negative Bacteria Evasion of Host Innate Immunity. Front

Immunol 4:109.

124. Vinogradov E, Perry MB, and Conlan JW. 2002. Structural analysis of

Francisella tularensis lipopolysaccharide. Eur J Biochem 269:6112-8.

125. Wang X, Ribeiro AA, Guan Z, McGrath SC, Cotter RJ, and Raetz CR. 2006.

Structure and biosynthesis of free lipid A molecules that replace

lipopolysaccharide in Francisella tularensis subsp. novicida. Biochemistry

45:14427-40.

126. Hajjar A, Harvey MD, Shaffer SA, Goodlett DR, Sjostedt A, Edebro H, Forsman

M, Bystrom M, Pelletier M, Wilson CB, Miller SI, Skerrett SJ, and Ernst RK.

2006. Lack of in vitro and in vivo recognition of Francisella tularensis subspecies

lipopolysaccharide by Toll-like receptors. Infect Immun 74:6730-8.

151 127. Llewellyn A, Zhao J, Song F, Parvathareddy J, Xu Q, Napier BA, Laroui H,

Merlin D, Bina JE, Cotter PA, Miller MA, Raetz CR, and Weiss DS. 2012. NaxD

is a deacetylase required for lipid A modification and Francisella pathogenesis.

Mol Microbiol 86:611-27.

128. Kanistanon D, Hajjar AM, Pelletier MR, Gallagher LA, Kalhorn T, Shaffer SA,

Goodlett DR, Rohmer L, Brittnacher MJ, Skerrett SJ, and Ernst RK. 2008. A

Francisella mutant in lipid A carbohydrate modification elicits protective

immunity. PLoS Pathog 4:e24.

129. Li Y, Powell, DA, Shaffer SA, Rasko DA, Pelletier MR, Leszyk JD, Scott AJ,

Masoudi A, Goodlett DR, Wang X, Raetz CR, and Ernst RK. 2012. LPS

remodeling is an evolved survival strategy for bacteria. Proc Natl Acad Sci U S A

109:8716-21.

130. Demeure C, Dussurget O, Fiol GM, Le Guern AS, Savin C, and Pizarro-Cerda J.

2019. and plague: an updated view on evolution, virulence

determinants, immune subversion, vaccination and diagnostics. Microbes Infect

doi:10.1016/j.micinf.2019.06.007.

131. Confer A and Ayalew S. 2013. The OmpA family of proteins: roles in bacterial

pathogenesis and immunity. Vet Microbiol 163:207-22.

132. Javed S, Skoog EC, and Solnick JV. 2019. Impact of Helicobacter pylori

Virulence Factors on the Host Immune Response and Gastric Pathology. Curr

Top Microbiol Immunol 421:21-52.

133. Smith S, Mahon V, Lambert MA, and Fagan RP. 2007. A molecular Swiss army

knife: OmpA structure, function and expression. FEMS Microbiol Lett 273:1-11.

152 134. Casey W, Spink N, Cia F, Collins C, Romano M, Berisio R, Bancroft GJ, and

McClean S. 2016. Identification of an OmpW homologue in Burkholderia

pseudomallei, a protective vaccine antigen against . Vaccine 34:2616-

21.

135. Ramakrishnan G, Meeker A, and Dragulev B. 2008. fslE is necessary for

siderophore-mediated iron acquisition in Francisella tularensis Schu S4. J

Bacteriol 190:5353-61.

136. Twine S, Bystrom M, Chen W, Forsman M, Golovliov I, Johansson A, Kelly J,

Lindgren H, Svensson K, Zingmark C, Conlan W, and Sjostedt A. 2005. A mutant

of Francisella tularensis strain SCHU S4 lacking the ability to express a 58-

kilodalton protein is attenuated for virulence and is an effective live vaccine.

Infect Immun 73:8345-52.

137. Lindgren H, Honn M, Golovlev I Kadzhaev K, Conlan W, and Sjostedt A. 2009.

The 58-kilodalton major virulence factor of Francisella tularensis is required for

efficient utilization of iron. Infect Immun 77:4429-36.

138. Ramakrishnan G, Sen B, and Johnson R. 2012. Paralogous outer membrane

proteins mediate uptake of different forms of iron and synergistically govern

virulence in Francisella tularensis tularensis. J Biol Chem 287:25191-202.

139. Chong A, Wehrly TD, Child R, Hansen B, Hwang S, Virgin HW, and Celli J.

2012. Cytosolic clearance of replication-deficient mutants reveals Francisella

tularensis interactions with the autophagic pathway. Autophagy 8:1342-56.

140. Mahawar M, Atianand MK, Dotson RJ, Mora V, Rabadi SM, Metzger DW,

Huntley JF, Harton JA, Malik M, and Bakshi CS. 2012. Identification of a novel

153 Francisella tularensis factor required for intramacrophage survival and

subversion of innate immune response. J Biol Chem 287:25216-29.

141. Mahawar M, Rabadi SM, Banik S, Catlett SV, Metzger DW, Malik M, and

Bakshi CS. 2013. Identification of a live attenuated vaccine candidate for

tularemia prophylaxis. PLoS One 8:e61539.

142. Melillo A, Sledjeski DD, Lipski S, Wooten RM, Basrur V, and Lafontaine ER.

2006. Identification of a Francisella tularensis LVS outer membrane protein that

confers adherence to A549 human lung cells. FEMS Microbiol Lett 263:102-8.

143. Su J, Yang J, Zhao D, Kawula TH, Banas JA, and Zhang JR. 2007. Genome-wide

identification of Francisella tularensis virulence determinants. Infect Immun

75:3089-101.

144. Miller S and Salama NR. 2018. The gram-negative bacterial periplasm: Size

matters. PLoS Biol 16:e2004935.

145. Melillo A, Mahawar M, Sellati TJ, Malik M, Metzger DW, Melendez JA, and

Bakshi CS. 2009. Identification of Francisella tularensis live vaccine strain CuZn

superoxide dismutase as critical for resistance to extracellularly generated reactive

oxygen species. J Bacteriol 191:6447-56.

146. Kijek T, Mou S, Bachert BA, Kuehl KA, Williams JA, Daye SP, Worsham PL,

and Bozue JA. 2019. The D-alanyl-d-alanine carboxypeptidase enzyme is

essential for virulence in the Schu S4 strain of Francisella tularensis and a dacD

mutant is able to provide protection against a pneumonic challenge. Microb

Pathog 137:103742.

154 147. Spidlova P, Stojkova P, Dankova V, Senitkova I, Santic M, Pinkas D,

Philimonenko V, and Stulik J. 2018. Francisella tularensis D-Ala D-Ala

Carboxypeptidase DacD Is Involved in Intracellular Replication and It Is

Necessary for Bacterial Cell Wall Integrity. Front Cell Infect Microbiol 8:111.

148. Silhavy T, Kahne D, and Walker S. 2010. The bacterial cell envelope. Cold

Spring Harb Perspect Biol 2:a000414.

149. Clarke M, Hughes DT, Zhu C, Boedeker EC, and Sperandio V. 2006. The QseC

sensor kinase: a bacterial adrenergic receptor. Proc Natl Acad Sci U S A

103:10420-10425.

150. Mokrievich A, Kondakova AN, Valade E, Platonov ME, Vakhrameeva GM,

Shaikhutdinova RZ, Mironova RI, Blaha D, Bakhteeva IV, Titareva GM,

Kravchenko TB, Kombarova TI, Vidal D, Pavlov VM, Lindner B, Dyatlov IA,

and Knirel YA. 2010. Biological properties and structure of the

lipopolysaccharide of a vaccine strain of Francisella tularensis generated by

inactivation of a quorum sensing system gene qseC. Biochemistry (Mosc) 75:443-

51.

151. Rasko D, Moreira CG, Li de R, Reading NC, Ritchie JM, Waldor MK, Williams

N, Taussig R, Wei S, Roth M, Hughes DT, Huntley JF, Fina MW, Falck JR, and

Sperandio V. 2008. Targeting QseC signaling and virulence for antibiotic

development. Science 321:1078-80.

152. van Hoek M, Hoang KV, and Gunn JS. 2019. Two-Component Systems in

Francisella Species. Front Cell Infect Microbiol 9:198.

155 153. Weiss D, Brotcke A, Henry T, Margolis JJ, Chan K, and Monack DM. 2007. In

vivo negative selection screen identifies genes required for Francisella virulence.

Proc Natl Acad Sci U S A 104:6037-42.

154. Fuller J, Craven RR, Hall JD, Kijek TM, Taft-Benz S, and Kawula TH. 2008.

RipA, a cytoplasmic membrane protein conserved among Francisella species, is

required for intracellular survival. Infect Immun 76:4934-43.

155. Huang M, Mortensen BL, Taxman DJ, Craven RR, Taft-Benz S, Kijek TM, Fuller

JR, Davis BK, Allen IC, Brickey WJ, Gris D, Wen H, Kawula TH, and Ting JP.

2010. Deletion of ripA alleviates suppression of the inflammasome and MAPK by

Francisella tularensis. J Immunol 185:5476-85.

156. Brunton J, Steele S, Miller C, Lovullo E, Taft-Benz S, and aKawula T. 2015.

Identifying Francisella tularensis genes required for growth in host cells. Infect

Immun 83:3015-25.

157. Bardwell J, McGovern K, and Beckwith J. 1991. Identification of a protein

required for disulfide bond formation in vivo. Cell 67:581-9.

158. Collet J and Bardwell JC. 2002. Oxidative protein folding in bacteria. Mol

Microbiol 44:1-8.

159. Kadokura H and Beckwith J. 2009. Detecting folding intermediates of a protein as

it passes through the bacterial translocation channel. Cell 138:1164-73.

160. Martin J, Bardwell JC, and Kuriyan J. 1993. Crystal structure of the DsbA protein

required for disulphide bond formation in vivo. Nature 365:464-8.

156 161. Ren Q, Zhang RR, Zhao XF, and Wang JX. 2010. A thioredoxin response to the

WSSV challenge on the Chinese white shrimp, Fenneropenaeus chinensis. Comp

Biochem Physiol C Toxicol Pharmacol 151:92-8.

162. Zapun A, Bardwell JC, and Creighton TE. 1993. The reactive and destabilizing

disulfide bond of DsbA, a protein required for protein disulfide bond formation in

vivo. Biochemistry 32:5083-92.

163. Bardwell J, Lee JO, Jander G, Martin N, Belin D, and Beckwith J. 1993. A

pathway for disulfide bond formation in vivo. Proc Natl Acad Sci U S A 90:1038-

42.

164. Berkmen M, Boyd D, and Beckwith J. 2005. The nonconsecutive disulfide bond

of Escherichia coli phytase (AppA) renders it dependent on the protein-disulfide

isomerase, DsbC. J Biol Chem 280:11387-94.

165. Katzen F and Beckwith J. 2000. Transmembrane electron transfer by the

membrane protein DsbD occurs via a disulfide bond cascade. Cell 103:769-79.

166. Zapun A, Missiakas D, Raina S, and Creighton TE. 1995. Structural and

functional characterization of DsbC, a protein involved in disulfide bond

formation in Escherichia coli. Biochemistry 34:5075-89.

167. Qin A and Mann BJ. 2006. Identification of transposon insertion mutants of

Francisella tularensis tularensis strain Schu S4 deficient in intracellular

replication in the hepatic cell line HepG2. BMC Microbiol 6:69.

168. Qin A, Scott DW, Thompson JA, and Mann BJ. 2009. Identification of an

essential Francisella tularensis subsp. tularensis virulence factor. Infect Immun

77:152-61.

157 169. Qin A, Zhang Y, Clark ME, Rabideau MM, Millan Barea LR, and Mann BJ.

2014. FipB, an essential virulence factor of Francisella tularensis subsp.

tularensis, has dual roles in disulfide bond formation. J Bacteriol 196:3571-81.

170. Ren G, Champion MM, and Huntley JF. 2014. Identification of disulfide bond

isomerase substrates reveals bacterial virulence factors. Mol Microbiol 94:926-44.

171. Qin A, Scott DW, Rabideau MM, Moore EA, and Mann BJ. 2011. Requirement

of the CXXC motif of novel Francisella infectivity potentiator protein B FipB,

and FipA in virulence of F. tularensis subsp. tularensis. PLoS One 6:e24611.

172. Schmidt M, Klimentova J, Rehulka P, Straskova A, Spidlova P, Szotakova B,

Stulik J, and Pavkova I. 2013. Francisella tularensis subsp. holarctica DsbA

homologue: a thioredoxin-like protein with chaperone function. Microbiology

159:2364-74.

173. Oyston P and Quarry JE. 2005. Tularemia vaccine: past, present and future.

Antonie Van Leeuwenhoek 87:277-81.

174. Tigertt W. 1962. Soviet viable Pasteurella tularensis vaccines. A review of

selected articles. Bacteriol Rev 26:354-73.

175. Eigelsbach H, Braun W, and Herring RD. 1951. Studies on the variation of

Bacterium tularense. J Bacteriol 61:557-69.

176. Cowley SC MS, and Nano FE. 1996. Phase variation in Francisella tularensis

affecting intracellular growth, lipopolysaccharide antigenicity and nitric oxide

production. Mol Microbiol 20:867-74.

177. Soni S, Ernst RK, Muszynski A, Mohapatra NP, Perry MB, Vinogradov E,

Carlson RW, and Gunn JS. 2010. Francisella tularensis blue-gray phase variation

158 involves structural modifications of lipopolysaccharide o-antigen, core and lipid a

and affects intramacrophage survival and vaccine efficacy. Front Microbiol

1:129.

178. Saslaw S, Eigelsbach HT, Prior JA, Wilson HE, and Carhart S. 1961. Tularemia

vaccine study. II. Respiratory challenge. Arch Intern Med 107:702-14.

179. Conlan J, Chen W, Bosio CM, Cowley SC, and Elkins KL. 2011. Infection of

mice with Francisella as an immunological model. Curr Protoc Immunol

Chapter 19:Unit 19 14.

180. Hansen B, Tilseth R, Finjord T, and Bruserud O. 2018. Francisella tularensis

bacteraemia causing multi-organ failure. Oxf Med Case Reports 2018:omy067.

181. Johansson A, Larkeryd A, Widerstrom M, Mortberg S, Myrtannas K, Ohrman C,

Birdsell D, Keim P, Wagner DM, Forsman M, and Larsson P. 2014. An outbreak

of respiratory tularemia caused by diverse clones of Francisella tularensis. Clin

Infect Dis 59:1546-53.

182. KuoLee R, Zhao X, Austin J, Harris G, Conlan JW, and Chen W. 2007. Mouse

model of oral infection with virulent type A Francisella tularensis. Infect Immun

75:1651-60.

183. KuoLee R, Harris G, Conlan JW, and Chen W. 2007. Oral immunization of mice

with the live vaccine strain (LVS) of Francisella tularensis protects mice against

respiratory challenge with virulent type A F. tularensis. Vaccine 25:3781-91.

184. Sunagar R, Kumar S, Franz BJ, and Gosselin EJ. 2016. Tularemia vaccine

development: or progress? Vaccine (Auckl) 6:9-23.

159 185. Wu T, Hutt JA, Garrison KA, Berliba LS, Zhou Y, and Lyons CR. 2005.

Intranasal vaccination induces protective immunity against intranasal infection

with virulent Francisella tularensis biovar A. Infect Immun 73:2644-54.

186. Twine S, Shen H, Harris G, Chen W, Sjostedt A, Ryden P, and Conlan W. 2012.

BALB/c mice, but not C57BL/6 mice immunized with a DeltaclpB mutant of

Francisella tularensis subspecies tularensis are protected against respiratory

challenge with wild-type bacteria: association of protection with post-vaccination

and post-challenge immune responses. Vaccine 30:3634-45.

187. Shen H, Chen W, and Conlan JW. 2004. Susceptibility of various mouse strains to

systemically- or aerosol-initiated tularemia by virulent type A Francisella

tularensis before and after immunization with the attenuated live vaccine strain of

the pathogen. Vaccine 22:2116-21.

188. Conlan J, Shen H, Golovliov I, Zingmark C, Oyston PC, Chen W, House RV, and

Sjostedt A. 2010. Differential ability of novel attenuated targeted deletion mutants

of Francisella tularensis subspecies tularensis strain SCHU S4 to protect mice

against aerosol challenge with virulent bacteria: effects of host background and

route of immunization. Vaccine 28:1824-31.

189. Ray H, Chu P, Wu TH, Lyons CR, Murthy AK, Guentzel MN, Klose KE, and

Arulanandam BP. 2010. The Fischer 344 rat reflects human susceptibility to

Francisella pulmonary challenge and provides a new platform for virulence and

protection studies. PLoS One 5:e9952.

190. De Pascalis R, Hahn A, Brook HM, Ryden P, Donart N, Mittereder L, Frey B, Wu

TH, and Elkins KL. 2018. A panel of correlates predicts vaccine-induced

160 protection of rats against respiratory challenge with virulent Francisella

tularensis. PLoS One 13:e0198140.

191. Snoy P. 2010. Establishing efficacy of human products using animals: the US

food and drug administration's "animal rule". Vet Pathol 47:774-8.

192. De Pascalis R, Chou AY, Bosio CM, Huang CY, Follmann DA, and Elkins KL.

2012. Development of functional and molecular correlates of vaccine-induced

protection for a model intracellular pathogen, F. tularensis LVS. PLoS Pathog

8:e1002494.

193. Barry E, Cole LE, and Santiago AE. 2009. Vaccines against tularemia. Hum

Vaccin 5:832-8.

194. Dreisbach V, Cowley S, and Elkins KL. 2000. Purified lipopolysaccharide from

Francisella tularensis live vaccine strain (LVS) induces protective immunity

against LVS infection that requires B cells and gamma interferon. Infect Immun

68:1988-96.

195. Fulop M, Manchee R, and Titball R. 1995. Role of lipopolysaccharide and a

major outer membrane protein from Francisella tularensis in the induction of

immunity against tularemia. Vaccine 13:1220-5.

196. Post D, Slutter B, Schilling B, Chande AT, Rasmussen JA, Jones BD, D'Souza

AK, Reinders LM, Harty JT, Gibson BW, and Apicella MA. 2017.

Characterization of Inner and Outer Membrane Proteins from Francisella

tularensis Strains LVS and Schu S4 and Identification of Potential Subunit

Vaccine Candidates. MBio 8.

161 197. Foshay L, Hesselbrock WH, Wittenberg HJ, and Rodenberg AH. 1942. Vaccine

Prophylaxis against Tularemia in Man. Am J Public Health Nations Health

32:1131-45.

198. Kadull P, Reames HR, Coriell LL, and Foshay L. 1950. Studies on tularemia. V.

Immunization of man. J Immunol 65:425-35.

199. Pasetti M, Cuberos L, Horn TL, Shearer JD, Matthews SJ, House RV, and Sztein

MB. 2008. An improved Francisella tularensis live vaccine strain (LVS) is well

tolerated and highly immunogenic when administered to rabbits in escalating

doses using various immunization routes. Vaccine 26:1773-85.

200. Jia Q and Horwitz MA. 2018. Live Attenuated Tularemia Vaccines for Protection

Against Respiratory Challenge With Virulent F. tularensis subsp. tularensis.

Front Cell Infect Microbiol 8:154.

201. Pechous R, McCarthy TR, Mohapatra NP, Soni S, Penoske RM, Salzman NH,

Frank DW, Gunn JS, and Zahrt TC. 2008. A Francisella tularensis Schu S4

purine auxotroph is highly attenuated in mice but offers limited protection against

homologous intranasal challenge. PLoS One 3:e2487.

202. Jia Q, Lee BY, Bowen R, Dillon BJ, Som SM, and Horwitz MA. 2010. A

Francisella tularensis live vaccine strain (LVS) mutant with a deletion in capB,

encoding a putative capsular biosynthesis protein, is significantly more attenuated

than LVS yet induces potent protective immunity in mice against F. tularensis

challenge. Infect Immun 78:4341-55.

203. Marohn M and Barry EM. 2013. Live attenuated tularemia vaccines: recent

developments and future goals. Vaccine 31:3485-91.

162 204. Vollmer W, Joris B, Charlier P, and Foster S. 2008. Bacterial peptidoglycan

(murein) hydrolases. FEMS Microbiol Rev 32:259-86.

205. Juan C, Torrens G, Barcelo IM, and Oliver A. 2018. Interplay between

Peptidoglycan Biology and Virulence in Gram-Negative Pathogens. Microbiol

Mol Biol Rev 82:e00033-18.

206. Braun V and Rehn K. 1969. Chemical characterization, spatial distribution and

function of a lipoprotein (murein-lipoprotein) of the E. coli cell wall. The specific

effect of trypsin on the membrane structure. Eur J Biochem 10:426-38.

207. Braun V. 1975. Covalent lipoprotein from the outer membrane of Escherichia

coli. Biochim Biophys Acta 415:335-77.

208. Bouveret E, Benedetti H, Rigal A, Loret E, and Lazdunski C. 1999. In vitro

characterization of peptidoglycan-associated lipoprotein (PAL)-peptidoglycan and

PAL-TolB interactions. J Bacteriol 181:6306-11.

209. Leduc M, Ishidate K, Shakibai N, and Rothfield L. 1992. Interactions of

Escherichia coli membrane lipoproteins with the murein sacculus. J Bacteriol

174:7982-8.

210. Pratt R. 2008. Substrate specificity of bacterial DD-peptidases (penicillin-binding

proteins). Cell Mol Life Sci 65:2138-55.

211. Glauner B, Holtje JV, and Schwarz U. 1988. The composition of the murein of

Escherichia coli. J Biol Chem 263:10088-95.

212. Cigana C, Curcuru L, Leone MR, Ierano T, Lore NI, Bianconi I, Silipo A,

Cozzolino F, Lanzetta R, Molinaro A, Bernardini ML, and Bragonzi A. 2009.

Pseudomonas aeruginosa exploits lipid A and muropeptides modification as a

163 strategy to lower innate immunity during lung infection. PLoS One

4:e8439.

213. Davis K and Weiser JN. 2011. Modifications to the peptidoglycan backbone help

bacteria to establish infection. Infect Immun 79:562-70.

214. Park J and Uehara T. 2008. How bacteria consume their own exoskeletons

(turnover and recycling of cell wall peptidoglycan). Microbiol Mol Biol Rev

72:211-27.

215. Dhar S, Kumari H, Balasubramanian D, and Mathee K. 2018. Cell-wall recycling

and synthesis in Escherichia coli and Pseudomonas aeruginosa - their role in the

development of resistance. J Med Microbiol 67:1-21.

216. Mengin-Lecreulx D, and van Heijenoort J. 1996. Characterization of the essential

gene glmM encoding phosphoglucosamine mutase in Escherichia coli. J Biol

Chem 271:32-9.

217. Benson T, Marquardt JL, Marquardt AC, Etzkorn FA, and Walsh CT. 1993.

Overexpression, purification, and mechanistic study of UDP-N-

acetylenolpyruvylglucosamine reductase. Biochemistry 32:2024-30.

218. Barreteau H, Kovac A, Boniface A, Sova M, Gobec S, and Blanot D. 2008.

Cytoplasmic steps of peptidoglycan biosynthesis. FEMS Microbiol Rev 32:168-

207.

219. Bouhss A, Crouvoisier M, Blanot D, and Mengin-Lecreulx D. 2004. Purification

and characterization of the bacterial MraY translocase catalyzing the first

membrane step of peptidoglycan biosynthesis. J Biol Chem 279:29974-80.

164 220. Anderson J, Matsuhashi M, Haskin MA, and Strominger JL. 1965. Lipid-

Phosphoacetylmuramyl-Pentapeptide and Lipid-Phosphodisaccharide-

Pentapeptide: Presumed Membrane Transport Intermediates in Cell Wall

Synthesis. Proc Natl Acad Sci U S A 53:881-9.

221. Mengin-Lecreulx D, Texier L, Rousseau M, and van Heijenoort J. 1991. The

murG gene of Escherichia coli codes for the UDP-N-acetylglucosamine: N-

acetylmuramyl-(pentapeptide) pyrophosphoryl-undecaprenol N-

acetylglucosamine transferase involved in the membrane steps of peptidoglycan

synthesis. J Bacteriol 173:4625-36.

222. Mohammadi T, van Dam V, Sijbrandi R, Vernet T, Zapun A, Bouhss A,

Diepeveen-de Bruin M, Nguyen-Disteche M, de Kruijff B, and Breukink E. 2011.

Identification of FtsW as a transporter of lipid-linked cell wall precursors across

the membrane. EMBO J 30:1425-32.

223. Sham L, Butler EK, Lebar MD, Kahne D, Bernhardt TG, and Ruiz N. 2014.

Bacterial cell wall. MurJ is the flippase of lipid-linked precursors for

peptidoglycan biogenesis. Science 345:220-2.

224. Pazos M and Peters K. 2019. Peptidoglycan. Subcell Biochem 92:127-168.

225. Johnson J, Fisher JF, and Mobashery S. 2013. Bacterial cell-wall recycling. Ann N

Y Acad Sci 1277:54-75.

226. Bugg T, Braddick D, Dowson CG, and Roper DI. 2011. Bacterial cell wall

assembly: still an attractive antibacterial target. Trends Biotechnol 29:167-73.

227. Scheurwater E, Reid CW, and Clarke AJ. 2008. Lytic transglycosylases: bacterial

space-making autolysins. Int J Biochem Cell Biol 40:586-91.

165 228. Sauvage E, Kerff F, Terrak M, Ayala JA, and Charlier P. 2008. The penicillin-

binding proteins: structure and role in peptidoglycan biosynthesis. FEMS

Microbiol Rev 32:234-58.

229. Van Heijenoort Y and Van Heijenoort J. 1971. Study of the N-acetylmuramyl-L-

alanine amidase activity in Escherichia coli. FEBS Lett 15:137-141.

230. Cheng Q and Park JT. 2002. Substrate specificity of the AmpG permease required

for recycling of cell wall anhydro-muropeptides. J Bacteriol 184:6434-6.

231. Votsch W and Templin MF. 2000. Characterization of a beta -N-

acetylglucosaminidase of Escherichia coli and elucidation of its role in

muropeptide recycling and beta -lactamase induction. J Biol Chem 275:39032-8.

232. Holtje JV KU, Ursinus A, and Wiedemann B. 1994. The negative regulator of

beta-lactamase induction AmpD is a N-acetyl-anhydromuramyl-L-alanine

amidase. FEMS Microbiol Lett 122:159-64.

233. Templin M, Ursinus A, and Holtje JV. 1999. A defect in cell wall recycling

triggers autolysis during the stationary growth phase of Escherichia coli. EMBO J

18:4108-17.

234. Frirdich E, Biboy J, Adams C, Lee J, Ellermeier J, Gielda LD, Dirita VJ, Girardin

SE, Vollmer W, and Gaynor EC. 2012. Peptidoglycan-modifying enzyme Pgp1 is

required for helical cell shape and pathogenicity traits in Campylobacter jejuni.

PLoS Pathog 8:e1002602.

235. Korza H and Bochtler M. 2005. Pseudomonas aeruginosa LD-carboxypeptidase,

a serine peptidase with a Ser-His-Glu triad and a nucleophilic elbow. J Biol Chem

280:40802-12.

166 236. Priyadarshini R, de Pedro MA, and Young KD. 2007. Role of peptidoglycan

amidases in the development and morphology of the division septum in

Escherichia coli. J Bacteriol 189:5334-47.

237. Metz R, Henning S, and Hammes WP. 1986. LD-carboxypeptidase activity in

Escherichia coli. II. Isolation, purification and characterization of the enzyme

from E. coli K 12. Arch Microbiol 144:181-6.

238. Frirdich E, Vermeulen J, Biboy J, Soares F, Taveirne ME, Johnson JG, DiRita VJ,

Girardin SE, Vollmer W, and Gaynor EC. 2014. Peptidoglycan LD-

carboxypeptidase Pgp2 influences Campylobacter jejuni helical cell shape and

pathogenic properties and provides the substrate for the DL-carboxypeptidase

Pgp1. J Biol Chem 289:8007-18.

239. Lenz J, Hackett KT, and Dillard JP. 2017. A Single Dual-Function Enzyme

Controls the Production of Inflammatory NOD Agonist Peptidoglycan Fragments

by Neisseria gonorrhoeae. MBio 8:e01464-17.

240. Mengin-Lecreulx D, and Lemaitre B. 2005. Structure and metabolism of

peptidoglycan and molecular requirements allowing its detection by the

Drosophila innate . J Endotoxin Res 11:105-11.

241. Pazos M and Peters K. 2019. Peptidoglycan. Subcell Biochem 92:127-168.

242. Typas A, Banzhaf M, Gross CA, and Vollmer W. 2011. From the regulation of

peptidoglycan synthesis to bacterial growth and morphology. Nat Rev Microbiol

10:123-36.

243. Holtje J. 1998. Growth of the stress-bearing and shape-maintaining murein

sacculus of Escherichia coli. Microbiol Mol Biol Rev 62:181-203.

167 244. Russo T, MacDonald U, Beanan JM, Olson R, MacDonald IJ, Sauberan SL, Luke

NR, Schultz LW, and Umland TC. 2009. Penicillin-binding protein 7/8

contributes to the survival of Acinetobacter baumannii in vitro and in vivo. J

Infect Dis 199:513-21.

245. Chaput C, Ecobichon C, Pouradier N, Rousselle JC, Namane A, and Boneca IG.

2016. Role of the N-Acetylmuramoyl-l-Alanyl Amidase, AmiA, of Helicobacter

pylori in Peptidoglycan Metabolism, Daughter Cell Separation, and Virulence.

Microb Drug Resist 22:477-86.

246. Ursinus A, Steinhaus H, and Holtje JV. 1992. Purification of a nocardicin A-

sensitive LD-carboxypeptidase from Escherichia coli by affinity chromatography.

J Bacteriol 174:441-6.

247. Sycuro L, Rule CS, Petersen TW, Wyckoff TJ, Sessler T, Nagarkar DB, Khalid F,

Pincus Z, Biboy J, Vollmer W, and Salama NR. 2013. Flow cytometry-based

enrichment for cell shape mutants identifies multiple genes that influence

Helicobacter pylori morphology. Mol Microbiol 90:869-83.

248. Ray K, Marteyn B, Sansonetti PJ, and Tang CM. 2009. Life on the inside: the

intracellular lifestyle of cytosolic bacteria. Nat Rev Microbiol 7:333-40.

249. Jones C, Napier BA, Sampson TR, Llewellyn AC, Schroeder MR, and Weiss DS.

2012. Subversion of host recognition and defense systems by Francisella spp.

Microbiol Mol Biol Rev 76:383-404.

250. Metz R, Henning S, and Hammes WP. 1986. LD-carboxypeptidase activity in

Escherichia coli. I. The LD-carboxypeptidase activity in ether treated cells. Arch

Microbiol 144:175-80.

168 251. Clavel T, Germon P, Vianney A, Portalier R, and Lazzaroni JC. 1998. TolB

protein of Escherichia coli K-12 interacts with the outer membrane

peptidoglycan-associated proteins Pal, Lpp and OmpA. Mol Microbiol 29:359-67.

252. Walburger A, Lazdunski C, and Corda Y. 2002. The Tol/Pal system function

requires an interaction between the C-terminal domain of TolA and the N-

terminal domain of TolB. Mol Microbiol 44:695-708.

253. Nelson D and Young KD. 2000. Penicillin binding protein 5 affects cell diameter,

contour, and morphology of Escherichia coli. J Bacteriol 182:1714-21.

254. Guinane C, Cotter PD, Ross RP, and Hill C. 2006. Contribution of penicillin-

binding protein homologs to antibiotic resistance, cell morphology, and virulence

of monocytogenes EGDe. Antimicrob Agents Chemother 50:2824-8.

255. Heidrich C, Templin MF, Ursinus A, Merdanovic M, Berger J, Schwarz H, de

Pedro MA, and Holtje JV. 2001. Involvement of N-acetylmuramyl-L-alanine

amidases in cell separation and antibiotic-induced autolysis of Escherichia coli.

Mol Microbiol 41:167-78.

256. Sycuro L, Pincus Z, Gutierrez KD, Biboy J, Stern CA, Vollmer W, and Salama

NR. 2010. Peptidoglycan crosslinking relaxation promotes Helicobacter pylori's

helical shape and stomach colonization. Cell 141:822-33.

257. Lindgren H, Stenman L, Tarnvik A, and Sjostedt A. 2005. The contribution of

reactive nitrogen and oxygen species to the killing of Francisella tularensis LVS

by murine macrophages. Microbes Infect 7:467-75.

169 258. Heidrich C, Ursinus A, Berger J, Schwarz H, and Holtje JV. 2002. Effects of

multiple deletions of murein hydrolases on viability, septum cleavage, and

sensitivity to large toxic molecules in Escherichia coli. J Bacteriol 184:6093-9.

259. Denome S, Elf PK, Henderson TA, Nelson DE, and Young KD. 1999.

Escherichia coli mutants lacking all possible combinations of eight penicillin

binding proteins: viability, characteristics, and implications for peptidoglycan

synthesis. J Bacteriol 181:3981-93.

260. Priyadarshini R, Popham DL, and Young KD. 2006. Daughter cell separation by

penicillin-binding proteins and peptidoglycan amidases in Escherichia coli. J

Bacteriol 188:5345-55.

261. Meyer K, Addy C, Akashi S, Roper DI, and Tame JRH. 2018. The crystal

structure and oligomeric form of Escherichia coli L,D-carboxypeptidase A.

Biochem Biophys Res Commun 499:594-599.

262. Das D, Herve M, Elsliger MA, Kadam RU, Grant JC, Chiu HJ, Knuth MW,

Klock HE, Miller MD, Godzik A, Lesley SA, Deacon AM, Mengin-Lecreulx D,

and Wilson IA. 2013. Structure and function of a novel LD-carboxypeptidase a

involved in peptidoglycan recycling. J Bacteriol 195:5555-66.

263. Steiner D, Furuya Y, Jordan MB, and Metzger DW. 2017. Protective Role for

Macrophages in Respiratory Francisella tularensis Infection. Infect Immun 85.

264. Minor P. 2015. Live attenuated vaccines: Historical successes and current

challenges. 479-480:379-92.

170 265. Roberts L, Powell DA, and Frelinger JA. 2018. Adaptive Immunity to Francisella

tularensis and Considerations for Vaccine Development. Front Cell Infect

Microbiol 8:115.

266. de Pedro M and Cava F. 2015. Structural constraints and dynamics of bacterial

cell wall architecture. Front Microbiol 6:449.

267. den Blaauwen T, de Pedro MA, Nguyen-Disteche M, and Ayala JA. 2008.

Morphogenesis of rod-shaped sacculi. FEMS Microbiol Rev 32:321-44.

268. Silhavy TJ, Kahne D, and Walker S. 2010. The bacterial cell envelope. Cold

Spring Harb Perspect Biol 2:a000414.

269. van Heijenoort J. 2011. Peptidoglycan hydrolases of Escherichia coli. Microbiol

Mol Biol Rev 75:636-63.

270. Jenkins C, Wallis R, Allcock N, Barnes KB, Richards MI, Auty JM, Galyov EE,

Harding SV, and Mukamolova GV. 2019. The lytic transglycosylase, LtgG,

controls cell morphology and virulence in Burkholderia pseudomallei. Sci Rep

9:11060.

271. Lenz JD, Hackett KT, and Dillard JP. 2017. A Single Dual-Function Enzyme

Controls the Production of Inflammatory NOD Agonist Peptidoglycan Fragments

by Neisseria gonorrhoeae. MBio 8.

272. Wu X, Ren G, and Huntley JF. 2015. Generating Isogenic Deletions (Knockouts)

in Francisella tularensis, a Highly-infectious and Fastidious Gram-negative

Bacterium. Bio Protoc 5:e1500.

273. Gallagher L, McKevitt M, Ramage ER, and Manoil C. 2008. Genetic dissection

of the Francisella novicida restriction barrier. J Bacteriol 190:7830-7.

171 274. Maier T, Havig A, Casey M, Nano FE, Frank DW, and Zahrt TC. 2004.

Construction and characterization of a highly efficient Francisella shuttle plasmid.

Appl Environ Microbiol 70:7511-9.

275. Price C. 1982. Centrifugation in density gradients. Academic Press:335-343.

276. Herve M, Boniface A, Gobec S, Blanot D, and Mengin-Lecreulx D. 2007.

Biochemical characterization and physiological properties of Escherichia coli

UDP-N-acetylmuramate:L-alanyl-gamma-D-glutamyl-meso-diaminopimelate

ligase. J Bacteriol 189:3987-95.

277. Blanot D, Kretsovali A, Abo-Ghalia M, Mengin-Lecreulx D, and van Heijenoort

J. 1983. Synthesis of analogues of precursors of bacterial peptidoglycan, p 311-

314. In Malon BaP (ed), Peptides. Walter de Gruyter, Berlin, Germany.

278. Pennartz A, Genereux C, Parquet C, Mengin-Lecreulx D, and Joris B. 2009.

Substrate-induced inactivation of the Escherichia coli AmiD N-acetylmuramoyl-

L-alanine amidase highlights a new strategy to inhibit this class of enzyme.

Antimicrob Agents Chemother 53:2991-7.

279. Leulier F, Parquet C, Pili-Floury S, Ryu JH, Caroff M, Lee WJ, Mengin-Lecreulx

D, and Lemaitre B. 2003. The Drosophila immune system detects bacteria

through specific peptidoglycan recognition. Nat Immunol 4:478-84.

280. Stenbak C, Ryu JH, Leulier F, Pili-Floury S, Parquet C, Herve M, Chaput C,

Boneca IG, Lee WJ, Lemaitre B, and Mengin-Lecreulx D. 2004. Peptidoglycan

molecular requirements allowing detection by the Drosophila immune deficiency

pathway. J Immunol 173:7339-48.

172 281. Oyston PC and Quarry JE. 2005. Tularemia vaccine: past, present and future.

Antonie Van Leeuwenhoek 87:277-81.

282. Kingry LC and Petersen JM. 2014. Comparative review of Francisella tularensis

and Francisella novicida. Front Cell Infect Microbiol 4:35.

283. Ellis J, Oyston PC, Green M, and Titball RW. 2002. Tularemia. Clin Microbiol

Rev 15:631-46.

284. Kugeler KJ, Mead PS, Janusz AM, Staples JE, Kubota KA, Chalcraft LG, and

Petersen JM. 2009. Molecular Epidemiology of Francisella tularensis in the

United States. Clin Infect Dis 48:863-70.

285. Petersen JM and Schriefer ME. 2005. Tularemia: emergence/re-emergence. Vet

Res 36:455-67.

286. Petersen JM, Mead PS, and Schriefer ME. 2009. Francisella tularensis: an

arthropod-borne pathogen. Vet Res 40:7.

287. Rosenberg R, Lindsey NP, Fischer M, Gregory CJ, Hinckley AF, Mead PS, Paz-

Bailey G, Waterman SH, Drexler NA, Kersh GJ, Hooks H, Partridge SK, Visser

SN, Beard CB, and Petersen LR. 2018. Vital Signs: Trends in Reported

Vectorborne Disease Cases - United States and Territories, 2004-2016. MMWR

Morb Mortal Wkly Rep 67:496-501.

288. Brooks GF and Buchanan TM. 1970. Tularemia in the United States:

epidemiologic aspects in the 1960s and follow-up of the outbreak of tularemia in

Vermont. J Infect Dis 121:357-9.

289. Telford SR 3rd and Goethert HK. 2010. Toward an understanding of the

perpetuation of the agent of tularemia. Front Microbiol 1:150.

173 290. Calhoun EL. 1954. Natural occurrence of tularemia in the lone star tick,

Amblyomma americanus (Linn.), and in dogs in Arkansas. Am J Trop Med Hyg

3:360-6.

291. Goethert HK, Shani I, and Telford SR 3rd. 2004. Genotypic diversity of

Francisella tularensis infecting Dermacentor variabilis ticks on Martha's

Vineyard, Massachusetts. J Clin Microbiol 42:4968-73.

292. Hopla CE. 1953. Experimental Studies on Tick Transmission of Tularemia

Organisms. American Journal of Hygeine 58:101-118.

293. Hopla CE. 1960. The transmission of tularemia organisms by ticks in the southern

states. South Med J 53:92-7.

294. Mani RJ, Metcalf JA, and Clinkenbeard KD. 2015. Amblyomma americanum as a

Bridging Vector for Human Infection with Francisella tularensis. PLoS One

10:e0130513.

295. Green R. 1931. The Occurence of Bact. Tularense in the Eastern Wood Tick,

Dermacentor variabilis. Am J Epidemiol 14:600-613.

296. Markowitz LE, Hynes NA, de la Cruz P, Campos E, Barbaree JM, Plikaytis BD,

Mosier D, and Kaufmann AF. 1985. Tick-borne tularemia. An outbreak of

lymphadenopathy in children. JAMA 254:2922-5.

297. Castellaw AH, Showers J, Goddard J, Chenney EF, and Varela-Stokes AS. 2010.

Detection of vector-borne agents in lone star ticks, Amblyomma americanum

(: Ixodidae), from Mississippi. J Med Entomol 47:473-6.

174 298. Goethert HK and Telford SR 3rd. 2009. Nonrandom distribution of vector ticks

(Dermacentor variabilis) infected by Francisella tularensis. PLoS Pathog

5:e1000319.

299. Goethert HK and Telford SR 3rd. 2010. Quantum of infection of Francisella

tularensis tularensis in host-seeking Dermacentor variabilis. Ticks Tick Borne Dis

1:66-8.

300. Kwit NA, Schwartz A, Kugeler KJ, Mead PS, and Nelson CA. 2018. Human

tularaemia associated with exposure to domestic dogs-United States, 2006-2016.

Zoonoses Public Health doi:10.1111/zph.12552.

301. Parker RR. 1934. Recent Studies of Tick Borne Diseases Made at the U.S. Public

Health Service Laboratories at Hamilton, MT. Proceedings of 5th Pacific Science

Conference 5:3367-3374.

302. Parker RR, Brooks CS, and Marsh H 1929. The occurrence of Bacterium

tularense in the wood tick (Dermacentor occidentalis) in California. Public

Health Rep 44:1299-1300.

303. Parker RR, Philip CB, and and Davis GE. 1932. Tularaemia: Occurrence in the

sage hen, Centrocercus urophasianus. Public Health Reports 47:479-487.

304. Gurycova D, Kocianova E, Vyrostekova V, and Rehacek J. 1995. Prevalence of

ticks infected with Francisella tularensis in natural foci of tularemia in western

Slovakia. Eur J Epidemiol 11:469-74.

305. Mancini F, Di Luca M, Toma L, Vescio F, Bianchi R, Khoury C, Marini L, Rezza

G, and Ciervo A. 2014. Prevalence of tick-borne pathogens in an urban park in

Rome, Italy. Ann Agric Environ Med 21:723-7.

175 306. Michelet L, Delannoy S, Devillers E, Umhang G, Aspan A, Juremalm M, Chirico

J, van der Wal FJ, Sprong H, Boye Pihl TP, Klitgaard K, Bodker R, Fach P, and

Moutailler S. 2014. High-throughput screening of tick-borne pathogens in Europe.

Front Cell Infect Microbiol 4:103.

307. Quarsten H, Skarpaas T, Fajs L, Noraas S, and Kjelland, V. 2015. Tick-borne

bacteria in Ixodes ricinus collected in southern Norway evaluated by a

commercial kit and established real-time PCR protocols. Ticks Tick Borne Dis

6:538-44.

308. Stensvold CR, Al Marai D, Andersen LO, Krogfelt KA, Jensen JS, Larsen KS,

and Nielsen HV. 2015. Babesia spp. and other pathogens in ticks recovered from

domestic dogs in Denmark. Parasit Vectors 8:262.

309. Wojcik-Fatla A, Zajac V, Sawczyn A, Cisak E, Sroka J, and Dutkiewicz J. 2015.

Occurrence of Francisella spp. in Dermacentor reticulatus and Ixodes ricinus ticks

collected in eastern Poland. Ticks Tick Borne Dis 6:253-7.

310. Gehringer H, Schacht E, Maylaender N, Zeman E, Kaysser P, Oehme R, Pluta S,

and Splettstoesser WD. 2013. Presence of an emerging subclone of Francisella

tularensis holarctica in Ixodes ricinus ticks from south-western Germany. Ticks

Tick Borne Dis 4:93-100.

311. Reis C, Cote M, Paul RE, and Bonnet S. 2011. Questing ticks in suburban forest

are infected by at least six tick-borne pathogens. Vector Borne Zoonotic Dis

11:907-16.

312. Wittwer M, Altpeter E, Pilo P, Gygli SM, Beuret C, Foucault F, Ackermann-

Gaumann R, Karrer U, Jacob D, Grunow R, and Schurch N. 2018. Population

176 Genomics of Francisella tularensis subsp. holarctica and its Implication on the

Eco-Epidemiology of Tularemia in Switzerland. Front Cell Infect Microbiol 8:89.

313. Milutinovic M, Masuzawa T, Tomanovic S, Radulovic Z, Fukui T, and Okamoto

Y. 2008. sensu lato, Anaplasma phagocytophilum,

Francisella tularensis and their co-infections in host-seeking Ixodes ricinus ticks

collected in Serbia. Exp Appl Acarol 45:171-83.

314. Tomaso H, Otto P, Peters M, Suss J, Karger A, Schamoni H, Zuchantke E, and

Hotzel H. 2018. Francisella tularensis and other bacteria in hares and ticks in

North Rhine-Westphalia (Germany). Ticks Tick Borne Dis 9:325-329.

315. Hestvik G, Warns-Petit E, Smith LA, Fox NJ, Uhlhorn H, Artois M, Hannant D,

Hutchings MR, Mattsson R, Yon L, and Gavier-Widen D. 2015. The status of

tularemia in Europe in a one-health context: a review. Epidemiol Infect 143:2137-

60.

316. Bonnet SI, Binetruy F, Hernandez-Jarguin AM, and Duron O. 2017. The Tick

Microbiome: Why Non-pathogenic Matter in Tick Biology and

Pathogen Transmission. Front Cell Infect Microbiol 7:236.

317. Gerhart JG, Moses AS, and Raghavan R. 2016. A Francisella-like endosymbiont

in the Gulf Coast tick evolved from a mammalian pathogen. Sci Rep 6:33670.

318. Liu JN, Yu ZJ, Liu LM, Li NX, Wang RR, Zhang CM, and Liu JZ. 2016.

Identification, Distribution and Population Dynamics of Francisella-like

Endosymbiont in Haemaphysalis doenitzi (Acari: Ixodidae). Sci Rep 6:35178.

177 319. Gerhart JG, Auguste Dutcher H, Brenner AE, Moses AS, Grubhoffer L, and

Raghavan R. 2018. Multiple Acquisitions of Pathogen-Derived Francisella

Endosymbionts in Soft Ticks. Genome Biol Evol 10:607-615.

320. Ivanov IN, Mitkova N, Reye AL, Hubschen JM, Vatcheva-Dobrevska RS,

Dobreva EG, Kantardjiev TV, and Muller CP. 2011. Detection of new

Francisella-like tick endosymbionts in Hyalomma spp. and Rhipicephalus spp.

(Acari: Ixodidae) from Bulgaria. Appl Environ Microbiol 77:5562-5.

321. Kugeler KJ, Gurfield N, Creek JG, Mahoney KS, Versage JL, and Petersen JM.

2005. Discrimination between Francisella tularensis and Francisella-like

endosymbionts when screening ticks by PCR. Appl Environ Microbiol 71:7594-7.

322. Rounds MA, Crowder CD, Matthews HE, Philipson CA, Scoles GA, Ecker DJ,

Schutzer SE, and Eshoo MW. 2012. Identification of endosymbionts in ticks by

broad-range polymerase chain reaction and electrospray ionization mass

spectrometry. J Med Entomol 49:843-50.

323. Niebylski ML, Peacock MG, Fischer ER, Porcella SF, and Schwan TG. 1997.

Characterization of an endosymbiont infecting wood ticks, Dermacentor

andersoni, as a member of the genus Francisella. Appl Environ Microbiol

63:3933-40.

324. Dergousoff SJ and Chilton NB. 2012. Association of different genetic types of

Francisella-like organisms with the rocky mountain wood tick (Dermacentor

andersoni) and the American dog tick (Dermacentor variabilis) in localities near

their northern distributional limits. Appl Environ Microbiol 78:965-71.

178 325. Kreizinger Z, Hornok S, Dan A, Hresko S, Makrai L, Magyar T, Bhide M,

Erdelyi K, Hofmann-Lehmann R, and Gyuranecz M. 2013. Prevalence of

Francisella tularensis and Francisella-like endosymbionts in the tick population of

Hungary and the genetic variability of Francisella-like agents. Vector Borne

Zoonotic Dis 13:160-3.

326. Azagi T, Klement E, Perlman G, Lustig Y, Mumcuoglu KY, Apanaskevich DA,

and Gottlieb Y. 2017. Francisella-Like Endosymbionts and Species in

Local and Imported Hyalomma Ticks. Appl Environ Microbiol 83.

327. Gurfield N, Grewal S, Cua LS, Torres PJ, and Kelley ST. 2017. Endosymbiont

interference and microbial diversity of the Pacific coast tick, Dermacentor

occidentalis, in San Diego County, California. PeerJ 5:e3202.

328. Gall CA, Reif KE, Scoles GA, Mason KL, Mousel M, Noh SM, and Brayton KA.

2016. The bacterial microbiome of Dermacentor andersoni ticks influences

pathogen susceptibility. ISME J 10:1846-55.

329. Sojka D, Franta Z, Horn M, Caffrey CR, Mares M, and Kopacek P. 2013. New

Insights into the Machinery of Blood Digestion by Ticks. Trends in

29:276-285.

330. Philip CB and Jellison, WL. 1934. The American Dog Tick, Dermacentor

variabilis, as a Host of Bacterium Tularense. Sage Publications I,

331. Mani RJ, Reichard MV, Morton RJ, Kocan KM, and Clinkenbeard KD. 2012.

Biology of Francisella tularensis subspecies holarctica live vaccine strain in the

tick vector Dermacentor variabilis. PLoS One 7:e35441.

179 332. Reese SM, Dietrich G, Dolan MC, Sheldon SW, Piesman J, Petersen JM, and

Eisen RJ. 2010. Transmission dynamics of Francisella tularensis subspecies and

clades by nymphal Dermacentor variabilis (Acari: Ixodidae). Am J Trop Med Hyg

83:645-52.

333. Bell FJ. 1945. The Infection of Ticks (Dermacentor variabilis) with Pasteurella

tularensis. The Journal of Infectious Diseases 76:83-95.

334. Genchi M, Prati P, Vicari N, Manfredini A, Sacchi L, Clementi E, Bandi C, Epis

S, and Fabbi M. 2015. Francisella tularensis: No Evidence for Transovarial

Transmission in the Tularemia Tick Vectors Dermacentor reticulatus and Ixodes

ricinus. PLoS One 10:e0133593.

335. Coburn J, Maier T, Casey M, Padmore L, Sato H, and Frank DW. 2015.

Reproducible and quantitative model of infection of Dermacentor variabilis with

the live vaccine strain of Francisella tularensis. Appl Environ Microbiol 81:386-

95.

336. Reese SM, Petersen JM, Sheldon SW, Dolan MC, Dietrich G, Piesman J, and

Eisen RJ. 2011. Transmission efficiency of Francisella tularensis by adult

american dog ticks (Acari: Ixodidae). J Med Entomol 48:884-90.

337. Goethert HK and Telford SR 3rd. 2011. Differential mortality of dog tick vectors

due to infection by diverse Francisella tularensis tularensis genotypes. Vector

Borne Zoonotic Dis 11:1263-8.

338. Desilva AM and Fikrig E. 1995. Growth and Migration of Borrelia-Burgdorferi in

Ixodes Ticks during Blood-Feeding. American Journal of Tropical Medicine and

Hygiene 53:397-404.

180 339. Sonenshine DE. 2018. Range Expansion of Tick Disease Vectors in North

America: Implications for Spread of Tick-Borne Disease. Int J Environ Res

Public Health 15.

340. Craig L, Pique ME, and Tainer JA. 2004. Type IV pilus structure and bacterial

pathogenicity. Nat Rev Microbiol 2:363-78.

341. Mattick J. 2002. Type IV pili and twitching motility. Annu Rev Microbiol 56:289-

314.

342. Salomonsson E, Forslund AL, and Forsberg A. 2011. Type IV Pili in Francisella

- A Virulence Trait in an Intracellular Pathogen. Front Microbiol 2:29.

343. Carbonnelle E, Helaine S, Prouvensier L, Nassif X, and Pelicic V. 2005. Type IV

pilus biogenesis in : PilW is involved in a step occurring

after pilus assembly, essential for fibre stability and function. Mol Microbiol

55:54-64.

344. Persat A, Inclan YF, Engel JN, Stone HA, and Gitai Z. 2015. Type IV pili

mechanochemically regulate virulence factors in Pseudomonas aeruginosa. Proc

Natl Acad Sci U S A 112:7563-8.

345. McKee R, Aleksanyan N, Garrett EM, and Tamayo R. 2018. Type IV Pili

Promote difficile Adherence and Persistence in a Mouse Model of

Infection. Infect Immun 86.

346. Gil H, Benach JL, and Thanassi DG. 2004. Presence of pili on the surface of

Francisella tularensis. Infect Immun 72:3042-7.

347. Larsson P, Oyston, PC, Chain P, Chu MC, Duffield M, Fuxelius HH, Garcia E,

Halltorp G, Johansson D, Isherwood KE, Karp PD, Larsson E, Liu Y, Michell S,

181 Prior J, Prior R, Malfatti S, Sjostedt A, Svensson K, Thompson N, Vergez L,

Wagg JK, Wren BW, Lindler LE, Andersson SG, Forsman M, and Titball RW.

2005. The complete genome sequence of Francisella tularensis, the causative

agent of tularemia. Nat Genet 37:153-9.

348. Ark N and Mann BJ. 2011. Impact of Francisella tularensis pilin homologs on

pilus formation and virulence. Microb Pathog 51:110-20.

349. Zogaj X, Chakraborty S, Liu J, Thanassi DG, and Klose KE. 2008.

Characterization of the Francisella tularensis subsp. novicida type IV pilus.

Microbiology 154:2139-50.

350. Forslund A, Kuoppa K, Svensson K, Salomonsson E, Johansson A, Bystrom M,

Oyston PC, Michell SL, Titball RW, Noppa L, Frithz-Lindsten E, Forsman M,

and Forsberg A. 2006. Direct repeat-mediated deletion of a type IV pilin gene

results in major virulence attenuation of Francisella tularensis. Mol Microbiol

59:1818-30.

351. Forslund A, Salomonsson EN, Golovliov I, Kuoppa K, Michell S, Titball R,

Oyston P, Noppa L, Sjostedt A, and Forsberg A. 2010. The type IV pilin, PilA, is

required for full virulence of Francisella tularensis subspecies tularensis. BMC

Microbiol 10:227.

352. Chakraborty S MM, Maier TM, Benach JL, Frank DW, and Thanassi DG. 2008.

Type IV pili in Francisella tularensis: roles of pilF and pilT in fiber assembly,

host cell adherence, and virulence. Infect Immun 76:2852-61.

182 353. Hager A, Bolton DL, Pelletier MR, Brittnacher MJ, Gallagher LA, Kaul R,

Skerrett SJ, Miller SI, and Guina T. 2006. Type IV pili-mediated secretion

modulates Francisella virulence. Mol Microbiol 62:227-37.

354. Forsberg A and Guina T. 2007. Type II secretion and type IV pili of Francisella.

Ann N Y Acad Sci 1105:187-201.

355. Kuchma S, Griffin EF, and O'Toole GA. 2012. Minor pilins of the type IV pilus

system participate in the negative regulation of swarming motility. J Bacteriol

194:5388-403.

356. Marko V, Kilmury SLN, MacNeil LT, and Burrows LL. 2018. Pseudomonas

aeruginosa type IV minor pilins and PilY1 regulate virulence by modulating

FimS-AlgR activity. PLoS Pathog 14:e1007074.

183 Appendix A

Reproduction of Figure 3-1 and Table 3.1

Figure A-1: Figure 3-1 and Table 3.1 were reproduced from Front Cell Infect Microbiol.

2019 May 8; 9:146. doi: 10.3389/fcimb.2019.00146. eCollection 2019 under the Creative

Commons Attribution 4.0 International Public License.

184