oxidative stress resistance in the francisella tularensis live vaccine strain is associated with
TRANSCRIPT
University of Iowa University of Iowa
Iowa Research Online Iowa Research Online
Theses and Dissertations
Spring 2016
Oxidative stress resistance in the Francisella tularensis live Oxidative stress resistance in the Francisella tularensis live
vaccine strain is associated with genetic variability in the ferrous vaccine strain is associated with genetic variability in the ferrous
iron uptake gene feoB iron uptake gene feoB
Joshua Robert Fletcher University of Iowa
Follow this and additional works at: https://ir.uiowa.edu/etd
Part of the Genetics Commons
Copyright 2016 Joshua Robert Fletcher
This dissertation is available at Iowa Research Online: https://ir.uiowa.edu/etd/3083
Recommended Citation Recommended Citation Fletcher, Joshua Robert. "Oxidative stress resistance in the Francisella tularensis live vaccine strain is associated with genetic variability in the ferrous iron uptake gene feoB." PhD (Doctor of Philosophy) thesis, University of Iowa, 2016. https://doi.org/10.17077/etd.c5lqs7hk
Follow this and additional works at: https://ir.uiowa.edu/etd
Part of the Genetics Commons
Oxidative stress resistance in the Francisella tularensis Live Vaccine Strain is
associated with genetic variability in the ferrous iron uptake gene feoB
by
Joshua Robert Fletcher
A thesis submitted in partial fulfillment
of the requirements for the Doctor of Philosophy
degree in Genetics in the
Graduate College of
The University of Iowa
May 2016
Thesis Supervisor: Professor Bradley D. Jones
Copyright by
JOSHUA ROBERT FLETCHER
2016
All Rights Reserved
Graduate College The University of Iowa
Iowa City, Iowa
CERTIFICATE OF APPROVAL
____________________________
PH.D. THESIS
_________________
This is to certify that the Ph.D. thesis of
Joshua Robert Fletcher
has been approved by the Examining Committee for the thesis requirement for the Doctor of Philosophy degree in Genetics at the May 2016 graduation. Thesis Committee: ____________________________________________ Bradley D. Jones, Thesis Supervisor ____________________________________________ Lee-Ann H. Allen ____________________________________________ Craig D. Ellermeier ____________________________________________ Mary E. Wilson ____________________________________________ John R. Kirby
ii
To my family.
iii
ACKNOWLEDGEMENTS
No scientist’s endeavors occur in a vacuum, unless you happen to be a
computer scientist in the 1940’s or an astronaut. I am neither. I owe much
gratitude to: my friends and colleagues Matt Faron and Jed Rasmussen, two
fellow graduate students in the lab who were soundboards for my ideas more
often than they probably wanted to be, and who offered many a great
conversation about books, science, life, the universe and Rick Astley’s place in it;
everyone in the Francisella group who helped me learn the field and think critically
about the details of experiments; Bram Slutter from the Harty lab for his help and
expertise with T cell biology; our collaborator Katy Bosio from Rocky Mountain
Labs for being generous with data and strains that served as the foundation for
this document; my mentor Brad Jones for giving me space in the lab and the
freedom to explore it; the Genetics program for admitting me and allowing me to
develop as a scientist. Lastly, I owe an enormous debt of gratitude to my friends
and family for supporting me even though I often put science ahead of them.
iv
ABSTRACT
Francisella tularensis is a highly virulent bacterial pathogen with an
extremely low infectious dose (~10 CFU) and high rates of mortality if left
untreated (30-60%). F. tularensis has an extensive history as a bioweapon, and
there is no vaccine currently licensed. For these reasons the CDC considers F.
tularensis a Tier 1 Select Agent. The unlicensed F. tularensis subsp. holarctica
Live Vaccine Strain (LVS) provides moderate protection against virulent strains;
however, we have discovered that various “wild type” lab stocks differ in their
virulence and ability to confer immunity. Genome sequencing of high virulence
(RML, LD50 ~200 CFU) and low virulence (ATCC, LD50 ~9,000 CFU) strains has
identified nine differences, of which four are non-synonymous substitutions. One
such mutation occurs in the ferrous iron uptake gene feoB in RML. While iron is
required for cellular function, ferrous iron (Fe2+) can participate in the Fenton
reaction with H2O2, leading to inactivation of essential iron-sulfur cluster
enzymes. Part of the innate immune response involves mitochondria-derived
reactive oxygen species in the cytosol. Fully virulent strains of F. tularensis are
known to be highly resistant to such host defenses, and have low levels of
intracellular iron. Accordingly, the RML strain was highly resistant to exogenous
H2O2 in vitro relative to the ATCC strain. An iron-responsive lacZ reporter had
~2-fold higher induction in the RML strain relative to ATCC during iron limitation.
Overexpression of the functional feoB allele, but not the RML allele, leads to
significantly increased sensitivity to H2O2.
v
Given the connection of iron and H2O2 toxicity, I revisited a previously
published transposon screen to determine if any of the mutants identified had a
role in iron homeostasis and oxidative stress resistance. One such gene was
annotated as bacterioferritin (bfr), which in other bacteria forms a hollow,
spherical multimer that oxidizes Fe2+ to Fe3+ and stores the oxidized form in the
interior of the sphere. The Δbfr mutant was ~10-fold more sensitive to H2O2 and
was attenuated nearly 8-fold in murine intranasal infection in terms of LD50
relative to the parental RML strain. Importantly, the Δbfr mutant allowed us to
test the hypothesis that H2O2 resistance is critical for the RML LVS to stimulate
productive immunity. At six weeks post-infection, mice previously infected with
either RML or the Δbfr mutant were challenged with an infection of 25 CFU of the
fully virulent F. tularensis Schu S4 strain. All mice immunized with RML survived
this challenge, while all mice immunized with Δbfr succumbed; only displaying a
slight increase in time to death. These results are consistent with the hypothesis
that the H2O2 resistance of RML LVS mediates increased fitness in a host.
vi
PUBLIC ABSTRACT
Many organisms on Earth require oxygen and iron to support their growth
and metabolism, including many bacteria that can cause disease. One such
bacterium, Francisella tularensis, can cause severe pneumonia that may lead to
death. The current vaccine against infection comes from a live but weakened
version of this bacterium, though it is not very effective. One of the goals of this
thesis was to examine the DNA from different lots of this vaccine strain to see if
there were genetic differences between lots that were more or less effective at
providing immunity. This analysis found that the effective vaccines had a genetic
mutation that made these particular strains of the weakened bacteria less able to
gather iron from their environment. Certain forms of iron can be highly reactive
with oxygen, which can be seen as rust on metal. This form of iron is important
for certain metabolic reactions in most living things, but too much of it can be
toxic to life. Iron can react with hydrogen peroxide, a common antiseptic.
Hydrogen peroxide is also made by the cells of the immune system to kill
invading pathogens. My research found that the more effective vaccine strains of
Francisella tularensis were resistant to the toxic effects of hydrogen peroxide
because they had less iron. I showed that genetically manipulating this strain to
make it have more iron made it sensitive to hydrogen peroxide and made it a less
effective vaccine.
vii
Table of Contents
LIST OF FIGURES ...............................................................................................ix
TABLE: PRIMER SEQUENCES ........................................................................... 1
CHAPTER I .......................................................................................................... 3
INTRODUCTION .................................................................................................. 3 Francisella tularensis early history .................................................................... 3 Tularemia .......................................................................................................... 6 F. tularensis taxonomy and ecology .................................................................. 8 Francisella tularensis genetics ........................................................................ 10 Bacterial factors that mediate pathogenesis .................................................... 13 Iron uptake in F. tularensis .............................................................................. 17 Intracellular lifecycle ........................................................................................ 20 Immunity to Francisella species ...................................................................... 26
CHAPTER II ....................................................................................................... 30
A VIRULENT BIOVAR OF F. TULARENSIS LVS IS INTRINSICALLY MORE RESISTANT TO HYDROGEN PEROXIDE ........................................................ 30
Introduction ..................................................................................................... 30 Materials and Methods .................................................................................... 32 Results ............................................................................................................ 36
LVS biovars have similar PiglA-lacZ expression and lack low molecular weight O-antigen glycosylated proteins........................................................ 36 RML LVS has less intracellular Fe2+ than the ATCC or Iowa LVS ............... 38 FeoB D471Y does not complement E. coli ΔfeoB fhuF::λplac ..................... 42 Resistance to H2O2 correlates with feoB D471Y allele ................................. 43 Resistance to H2O2 also requires Dyp peroxidase ....................................... 45 Neither feoB nor dyp are essential for intracellular growth in A549 or BMM cells .................................................................................................... 47
Discussion ....................................................................................................... 62
CHAPTER III ...................................................................................................... 67
CHARACTERIZATION OF BACTERIOFERRITIN IN OXIDATIVE STRESS RESISTANCE ..................................................................................................... 67
Introduction ..................................................................................................... 67 Materials and methods .................................................................................... 69 Results ............................................................................................................ 72
A bacterioferritin mutant has modest reduction in PfslA-lacZ activity ............. 72 Bacterioferritin promoter activity is not significantly induced in iron limiting media ............................................................................................... 73 Bacterioferritin protects against H2O2 but is not required for intracellular growth in A549 cells ..................................................................................... 75
viii
Bacterioferritin contributes to virulence and immunogenicity of RML LVS in vivo ........................................................................................................... 75
CHAPTER IV ...................................................................................................... 92
CONCLUSIONS AND FUTURE DIRECTIONS .................................................. 92
APPENDIX A .................................................................................................... 109
DEVELOPMENT OF RECOMBINANT IGLC-LCMV EPITOPE T CELL REPORTER FUSIONS ..................................................................................... 109
Rationale ....................................................................................................... 109 Materials and methods .................................................................................. 110 Results and discussion .................................................................................. 112
APPENDIX B .................................................................................................... 120
OXYR DOES NOT INFLUENCE OXIDATIVE STRESS RESISTANCE IN F. TULARENSIS LVS ........................................................................................... 121
Rationale ....................................................................................................... 121 Results and discussion .................................................................................. 124
REFERENCES ................................................................................................. 129
ix
LIST OF FIGURES
Figure II.1. RML LVS has similar FPI reporter gene expression and lacks glycosylated proteins seen in virulent Type A and Type B strains of F. tularensis. ..................................................................................................... 49
Figure II.2. Location of the amino acid residue that is mutated in the RML LVS FeoB. .................................................................................................... 50
Figure II.3. RML LVS exhibits growth restriction under iron limitation on MMH agar. ................................................................................................... 51
Figure II.4. RML LVS has iron-related gene expression consistent with less intracellular Fe2+ in iron limiting media. ....................................................... 52
Figure II.5. Constitutive fur expression inhibits robust growth of RML LVS. ....... 53
Figure II.6. RML feoB allele does not complement an E. coli ΔfeoB iron reporter strain. .............................................................................................. 54
Figure II.7. RML LVS is significantly more resistant to H2O2 lethality. ............... 55
Figure II.8. RML LVS can be sensitized to H2O2 by overexpression of a functional feoB. ............................................................................................ 56
Figure II.9. FeoB-mediated sensitivity to H2O2 is independent of the RML LVS genetic background. ............................................................................. 57
Figure II.10. Dyp peroxidase protects against H2O2. ......................................... 58
Figure II.11. Fold growth of LVS biovars in A549 cells. ...................................... 59
Figure II.12. Contribution of feoB alleles to intracellular growth in A549 cells. ... 60
Figure II.13. Twenty-four-hour growth of LVS biovars and mutant derivatives in murine bone marrow-derived macrophages. ............................................ 61
Figure III.1. Bacterioferritin promoter activity is similar between LVS biovars grown in Chamberlain’s defined medium with high or low iron. .................... 78
Figure III.2. Bacterioferritin promoter activity in is upregulated in ∆feoB and downregulated in Δbfr in MMH broth. ........................................................... 79
Figure III.3. Mutants lacking bfr and feoB differ in their H2O2 sensitivity depending on phase of growth. .................................................................... 80
Figure III.4. The bacterioferritin mutant is proficient for intracellular growth in vitro. ............................................................................................................. 81
Figure III.5. RML Δbfr is modestly attenuated in murine infection via the intranasal route of inoculation. ..................................................................... 82
x
Figure III.6. RML Δbfr-infected mice lost less weight than mice infected with the wild type RML LVS. ................................................................................ 83
Figure III.7. RML requires bacterioferritin to elicit protection against an intranasal challenge of 25 CFU F. tularensis Schu S4. Mice were infected with either 121 CFU of RML LVS or 215 CFU of the bacterioferritin mutant and allowed to recover for fix weeks before challenge. ..................................................................................................... 84
Figure IV.I Model of iron import in F. tularensis subsp. holarctica Live Vaccine Strain (LVS). ................................................................................. 107
Figure IV.II Model of intracellular outcomes for high and low iron strains. ........ 108
Figure A.I. Low dose immunization survival curves. ......................................... 116
Figure A.II. High dose immunization survival curves. ....................................... 117
Figure A.III. IglC-LCMV epitope fusions are expressed in LVS and do not alter intracellular growth. ............................................................................ 118
Figure A.IV. LVS infected mouse lymphocyte response to LCMV antigens ex vivo. ....................................................................................................... 119
Figure A.V. Survival of LCMV immunized mice challenged with LVS expressing LCMV epitopes. ....................................................................... 120
Figure B.I. The RML LVS ∆oxyR mutant is as resistant to H2O2 as the wild type. ........................................................................................................... 127
Figure B.II. The ΔoxyR mutant has a mild intracellular growth defect in J774A.16 cells. ........................................................................................... 128
1
TABLE: PRIMER SEQUENCES
5’feoB.up.asc1.bamh1 – ggcgcgccggatccAGCATATCAAACACAAAGAAGAAGTATTG 3’feoB.up.avr2 – cctaggAATCAGTTTTCAGGAGTTTATAGTATAG 5’feoB.down.asc1.avr2 – ggcgcgcccctaggAATTCTAATTTGAATATACAGCTTTA 3’feoB.down.bgl2 – agatctTTAGCATTTCACTAAGATTTGCACC 5’feoB.mut.check – AGATGATTGCTCAGTAATAACAGCAACTTTGC 3’feoB.mut.check – CATTAATAAGAATAGTATTCTCATTTTAAAATACCTC 5’feoB.SNP – AGACTTGCGATATTTTCAGTATTTGC 3’feoB.SNP – TTTACCGGCATATTCAAGTGCTGTGG 5’dyp.up.asc1.bamh1 – ggcgcgccggatccCCATCATCAAAGCTTTGATTTGGG 3’dyp.up.avr2 – cctaggTTAACTTTTCCTTATAAAACGCTC 5’dyp.down.asc1.avr2 – ggcgcgcccctaggAACAAGCTACTTGAAATTCTTTATTTTATA 3’dyp.down.bgl2 - agatctCATAGTACTCAACAAACTTGCCAAC 5’feoBc.kpn1 – ggtaccATGAAATATGCTCTAGTTGGCAATCC 3’feoBc.sal1 – gtcgacATATTTAAAGCTGTATATTCAAATTAG 5’dyp.mut.check – AGTATCCATACGATAATAAAGTTAGTTATAGAAGC 3’dyp.mut.check - ATAGAGAATATTTCTTTCCTTTATCCTCTATTTGTAGC 5’dyp.comp.kpn1 – ggtaccGTGGAAATAATTAAATATCAATTAGGAATAG 3’dyp.comp.sal1 – gtcgacCTAATTTAGTAAAGAAATATCTAGTTTGCC 5’bferr.up.asc1.bamh1 – ggcgcgccggatccTTTTAGTGATACTTTTTGAGACAATTGTCCC 3’bferr.up.asc1.avr2 – ggcgcgcccctaggCATATTGTTACCTCCATTATTTAAAACTCTAATC 5’bferr.down.asc1 – ggcgcgccTAAAGGCTATTATCCTCGATGAGTTTTTCTTC 3’bferr.down.bamh1 – ggatccAAGTATTTATCTGTAGTTACAATGGTGG 5’bfr.mut.check – ATATCATTTTTATTAAAATATCTAGGTTG 3’bfr.mut.check – ATAAATACTTTAAGTCACTAAATATCTCG 5’Pbfr.bamh1 -ggatccATATCATTTTTATTAAAATATCTAGGTTG 3’Pbfr.kpn1 – ggtaccATTGTTACCTCCATTATTTAAAACTCTAATC 5’iglC.kpn1 – ggtaccATGATTATGAGTGAGATGATAACAAG
2
3’iglC.sal1 - gtcgacCTATGCAGCTGCAATATATCCTATTTTAGC 3’iglC – CTATGCAGCTGCAATATATCCTATTTTAGC 3’iglC.GP33.41 – gtcgacCTATAATATACCCATAGTAGCACCATTATATACAGCTTTTATTGAAGTTGCAGCTGCAATATATCCTATTTTAGCAACTCC gp61.80.1c - GCACAGATAAAGGAGTTGCTAAAATAGGATATATTGCAGCTGCAGGTATGTATGGTTTAAAAGGTCCTGATATATATAAAGG gp61.80.1r – CCTTTATATATATCAGGACCTTTTAAACCATACATACCTGCAGCTGCAATATATCCTATTTTAGCAACTCCTTTATCTGTGC Gp61.80.2c – GGTATGTATGGTTTAAAAGGTCCTGATATATATAAAGGTGTATATCAATTTAAATCAGTAGAATTTGATATGTCACATTAGgtcgac Gp61.80.2r – gtcgacCTAATGTGACATATCAAATTCTACTGATTTAAATTGATATACACCTTTATATATATCAGGACCTTTTAAACCATACATACC 5’oxyR.up – ACTATCCATACAGTCGACTCTAGAGGATCACCAGCTACAGACTTAAGATAAGCATTTGC 3’oxyR.up – AACTAAATCTTATAGTTACTATAAATATTAT 5’oxyR.down – GCTCACATAAATATCATCCAAATACCC 3’oxyR.down – GATGAATTCGAGCTCGGTACCCGGGTTAATTTTAAGGATGGTAAGCG 5’oxyR.up.JC84.ITA – ACTATCCATACAGTCGACTCTAGAGGATCggcgcgccggatccACCAGCTACAGACTTAAGATAAGCATTTGC 3’oxyR.down.JC84.ITA – GATGAATTCGAGCTCGGTACCCGGGGATCggatccGTTAATTTTAAGGATGGTAAGCG 5’oxyR.mut.check – GTTAAGCTACGCTATTATTGTATTACTCTTGC 3’oxyR.mut.check – AGGATTTTTATCATTAACAGAAAATATTATGATG
3
CHAPTER I
INTRODUCTION
Francisella tularensis early history
The bacterium that was eventually named Francisella tularensis was
isolated in 1911 from squirrels in California by George McCoy and Charles
Chapin, researchers working at the United States Public Health and Marine
Service who were investigating disease epidemiology in squirrels3. Infected
squirrels were described as having plague-like buboes, enlarged spleens, and
lesions in their livers. McCoy and Chapin eventually isolated the causative agent
and characterized it as being quite fastidious, requiring nutrient rich media or
passage through animals for propagation. The disease caused by this bacterium
was differentiated from that caused by Yersinia pestis by infecting plague-
immune guinea pigs with either Y. pestis or the newly discovered pathogen.
They observed that the guinea pigs infected with the latter quickly succumbed,
while the Y. pestis challenged animals survived, suggesting that the newly
discovered bacterium was not immunologically similar to Y. pestis. Not long after
their initial description of the disease and the causative agent, the first case of
human infection with Francisella tularensis was reported4. The infected
individual had recently prepared meat at a restaurant; he presented with
conjunctivitis and developed swollen lymph nodes and the infectious agent
isolated from his conjunctival ulcers matched the description of the organism
described by McCoy and Chapin.
4
Increased case reports over the following decade made it clear that a
major risk factor for contracting the disease was through the handling or
ingestion of infected meat or by insect bite, typically ticks or deer flies5,6. Edward
Francis, for whom the bacteria is named, characterized outbreaks of tularemia in
rural populations of the Pahvant Valley region of Utah, describing the
ulceroglandular form of the disease as occurring following the bite of a deer fly
that had previously bitten an infected rabbit; Dr. Francis also recorded an early
case of lethal pneumonic tularemia7. In his studies he found that the organism
can persist in lice associated with rodents, rabbits and hares, squirrel-associated
fleas, as well as bedbugs8,9. During the time period in which he was conducting
these studies, Dr. Francis also described the infection of six laboratory workers
who had been working with him to characterize the tularemia outbreaks in Utah.
Two of the individuals were performing field work and may have contracted the
infection naturally, but the remaining four became infected under then standard
laboratory conditions during propagation and characterization of the bacterium7.
Francis later notes that laboratory worker infection during routine passage of the
bacterium through animals was so dangerous that the Lister Institute of
Preventative Medicine in London, England suspended work with the organism7.
The symptoms experienced by both the initial patients and infected laboratory
personnel highlight two of the hallmark features of F. tularensis: 1) the organism
is highly infectious, and 2) the disease it causes is debilitating and can result in
fatality.
5
The extreme infectiousness of F. tularensis and the incapacitation caused
by tularemia did not escape the notice of the governments of Imperial Japan, the
Soviet Union, and the United States, who in the early and mid-20th century
devoted significant resources towards developing biological weapons
programs10-14. While the United States had forsworn offensive bioweapons
capabilities in 1972, evidence exists that the Soviet Union maintained active
bioweapons stockpiles and research programs through its dissolution in the early
1990s5. Given this information and the ubiquity of F. tularensis in the
environment throughout the Northern hemisphere, fears of the intentional release
of aerosols of the bacteria are not without merit. A model produced by the World
Health Organization estimated that aerosol release of F. tularensis over a large
metropolitan area would likely result in approximately 250,000 casualties and
upwards of 19,000 fatalities15. A more recent model put forth by the Centers for
Disease Control and Prevention estimated that an attack on a population center
of 100,000 persons could result in greater than 6,000 deaths and up to $5.4
billion in associated costs for treatments and prophylaxis16. Indeed, these
scenarios have recently spurred considerable research efforts that are focused
on understanding the pathogenesis of F. tularensis and developing effective
vaccines against the disease. To date there is no licensed vaccine for prevention
of tularemia, however an attenuated derivative of F. tularensis subsp. holarctica
called the Live Vaccine Strain or LVS has been used to vaccinate laboratory
workers in the past11. Vaccination proved useful in reducing incidences of
laboratory acquired infections, but LVS does not protect against moderate doses
6
of aerosolized F. tularensis subsp. tularensis, and the details of the strain’s
creation are unclear. Furthermore, as the nature of the attenuating mutations in
LVS were not fully understood, there were concerns about potential reversion or
suppression of these mutations that could make LVS virulent again, thus the
strain was discontinued for human use in the United States.
Tularemia
The disease caused by F. tularensis infection can manifest in multiple
ways depending on the route of infection. Typical cases of tularemia are
associated with general flu-like symptoms, such as fever, aches, sore throat, and
occasionally dry cough. The most common form of tularemia (~75-80% of
cases), ulceroglandular, presents with a cutaneous ulcer at the site of infection,
with malaise, chills and fever setting in approximately three or more days after
the infection17,18. This is followed by the swelling of draining lymph nodes, which
can continue to enlarge to the point of suppuration. The ulcerous tissue can
undergo necrosis and forms a structure that is not unlike that of an anthrax-like
eschar. While not associated with high rates of mortality, this form of the disease
can progress to secondary pneumonic tularemia or sepsis by dissemination of
the bacteria to other organs. Oculoglandular tularemia occurs when bacteria are
introduced to the mucus membranes of the eye, and is characterized by
inflammation and ulceration of the conjunctiva; the cervical lymph nodes may
became swollen18. Oropharyngeal tularemia, while rare, occurs after ingestion of
7
contaminated materials and is associated with oral ulcers, cervical lymph node
swelling, tonsillitis, gastrointestinal discomfort and occasionally diarrhea18,19.
The most serious form of the disease, pneumonic tularemia, results from
the inhalation of F. tularensis and is associated with high mortality in the absence
of antibiotic administration, as high as 60%17,20. This form of the disease is
characterized by symptoms typical of tularemia, as well as a non-productive
cough and chest pains that become worse as the disease progresses, bloody
sputum, pleural effusion, possible hemorrhage, and dissemination of the
organisms to the liver and spleen17,18. If untreated, this can lead to organ failure
and death. While rare, outbreaks of pneumonic tularemia occur in areas where
rodent or lagomorph populations have high rates of infection with F. tularensis,
such as Martha’s Vineyard in Massachusetts, rural areas of Scandinavia or areas
where sanitation and hygiene services have been discontinued due to war or
natural disaster21-25. Outbreaks in these locations are associated with aerosols
of F. tularensis that had been generated by running lawnmowers over infected
rabbit carcasses, and the movement of hay that likely had contact with infected
rodents or their carcasses, as well as contamination of water and food supplies.
Pneumonic tularemia would likely be the most common presentation of the
disease in the event of intentional release of F. tularensis, thus careful
surveillance is needed to distinguish between this and naturally occurring
outbreaks.
8
F. tularensis taxonomy and ecology
F. tularensis is a member of the γ-proteobacteria, though it is
phylogenetically distant from its most closely-related genera, Wolbachia,
Coxiella, and Legionella26. The Francisella genus has three recognized species:
F. tularensis, F. philomiragia, and F. novicida. The latter two species are rarely
associated with disease in humans; little is known about F. philomiragia, and only
immune compromised individuals are at increased risk of infection with F.
novicida27. The two disease-causing subspecies of F. tularensis are known as
Type A (subsp. tularensis) and Type B (subsp. holarctica), and can be found
across the northern hemisphere28. F. tularensis subsp. tularensis is endemic to
North America and causes the severe forms of tularemia that are associated with
high mortality rates, while subsp. holarctica causes significant but rarely lethal
disease, and can be found in North America, Europe, and Asia. Type A strains
exhibit relatively high genetic diversity within the subspecies, while Type B strains
tend to have lower diversity, consistent with a proposed bottleneck event
sometime in the past29-31. It is interesting to speculate that the reduced virulence
of F. tularensis subsp. holarctica may have contributed to its rapid spread across
the northern hemisphere, as prolonged windows of time for insect mediated
transmission events would likely occur in populations of sick animals where Type
B strains are endemic.
A comprehensive genotyping analysis has identified two major clades
within Type A F. tularensis, called Type A.I (itself further separated into A.Ia and
A.Ib) and A.II32. Type A.I and A.II strains have been shown to be slightly more
9
virulent than the prototypical virulent laboratory strain, F. tularensis subsp.
tularensis Schu S4 in mouse models of infection33. Epidemiological data suggest
that A.Ib strains cause the most morbidity and mortality among documented
human infections in the United States, while A.Ia and A.II strains are associated
with less mortality than even Type B strains32. A third subspecies of F. tularensis
has been isolated from Central Asia, F. tularensis subsp. mediasiatica that
exhibits some virulence in rabbit, but it is not clear if the organism is connected to
disease in man34.
The two subclades of subsp. tularensis strains exhibit differences in
geographic distributions, with the distribution of common vector species for each
subclade closely overlapping. The A.I strains are typically found in the eastern
portion of North America and are commonly associated with infections stemming
from tick (Dermacentor variabilis or Amblyomma americanum) bites or contact
with infected rabbit flesh30. As these tick species have been reported to harbor F.
tularensis for long periods of time throughout multiple stages of their life cycles, it
is not surprising that a majority (~60%) of recorded tularemia infections in the
central and southern regions of the United States reported contact with ticks near
the time of infection35-37. In contrast, the A.II strains are found in the western
United States; transmission and infection are associated with Dermacentor
andersoni ticks and the deer-fly Chysops discalis30. As with Type A strains, ticks
serve as important reservoirs and vectors for Type B strains in North America
and Europe. Typically, these strains are transmitted by ticks from the Ixodes and
10
Dermacentor genera, though others may be associated, as well as Laelaps mites
that prey on small rodent species34.
The transmission cycles of both subspecies involve infection of rodents
and lagomorphs through tick and fly vectors, and it is known that tularemia
outbreaks in small mammal populations are associated with outbreaks of human
tularemia38. Interestingly, F. tularensis has an enormous range of hosts that it
can infect, possibly greater than 250 species, including amoebae in soil and
aquatic environments, yet large scale human outbreaks are relatively rare39,40.
Some species are highly susceptible to infection, such as mice, and may not
survive long enough for significant transmission to insect vectors. It is also
possible that the many of the species that can be infected with F. tularensis do
not encounter any of the insect vectors in the wild, or do not contract significant
enough disease upon infection to transmit bacteria to insect vectors.
Francisella tularensis genetics
At approximately 1.89 million base pairs in length, the F. tularensis
genome is relatively small compared to other intracellular pathogens and is AT
rich, with approximately 33% GC content26. Genome comparison between the
two virulent subspecies and a species that is nonpathogenic for humans, F.
novicida, indicates that a high degree of similarity exists in the Francisella genus.
The F. novicida genome has approximately 97-98% identity among conserved
sequences between it and the F. tularensis species, and nearly 86% of the core
genome is common to all three41. F. novicida is an environmental species
11
usually associated with aquatic environments that exhibits virulence in mice,
though significant differences exist in terms of the disease when compared to
that caused by F. tularensis42. The high degree of similarity between Francisella
species is remarkable given the extreme difference in virulence and disease
phenotypes, especially when one considers the paucity of genes dedicated solely
to virulence, such as those encoding Type III or Type IV secretion systems used
by other intracellular pathogens like Salmonella or Legionella26. One notable
exception is the presence of an approximately 30 kilobase locus, termed the
Francisella Pathogenicity Island (FPI) that encodes genes with weak similarity to
Type VI secretion systems43,44. This locus encodes key functions for the
intracellular lifecycle of Francisella species, and interestingly, two functioning
copies of the FPI are encoded on the chromosomes of subsp. tularensis and
subsp. holarctica, highlighting the importance of these genes for the
pathogenesis of the bacteria. The FPI will be discussed further in a later section.
A key feature of the genomes of the virulent species of Francisella that
distinguishes them from that of F. novicida are the high number of insertion
sequence (IS) elements and genomic rearrangements that have occurred as a
consequence of homologous recombination between these IS elements. Indeed,
Rohmer, et. al. found that the genomes of subsp. holarctica LVS and subsp.
tularensis Schu S4 have 109 and 79 IS elements, respectively, while that of F.
novicida only contains 2641. A consequence of the expansion of these IS
elements is both an increase in the number of genes inactivated by insertion of
these elements and an increase in genomic rearrangements in the virulent
12
species relative to F. novicida. The organization within genomic blocks and the
presence of conserved sequences flanking the IS elements that have mediated
recombination events suggests that a significant number of these changes
occurred in the ancestor of subsp. tularensis and subsp. holarctica, though
further subspecies-specific changes have occurred since their divergence41. The
proliferation of these IS elements is consistent with what is observed in the
genomes of other pathogenic bacteria. For example, pathogenic species of
Bordetella, Shigella, Yersina, Burkholderia, as well as Mycobacterium leprae all
have genomic signatures of proliferation of IS elements and gene loss or decay,
suggesting that adaptation to the pathogenic lifestyle is associated with loss of
so-called “anti-virulence genes” that are not essential or even deleterious to the
infection process45-49. Many of the genes lost or non-functional in the virulent
species of Francisella encode proteins that are involved in metabolic pathways,
in particular those for amino acid biosynthesis41. In fact, F. tularensis requires
supplementation of thirteen amino acids and numerous other nutrients, such as
thiamine and β-nicotinamide adenine dinucleotide, to support its growth in
standard laboratory culture. This implies that F. tularensis readily obtains these
essential molecules in the preferred niche of the host cell environment.
The F. tularensis subsp. tularensis and subsp. holarctica genomes differ
primarily by the chromosomal location of large blocks of synteny that have
rearranged independently in each subspecies, however the gene order and
nucleotide identity within these blocks are highly conserved41. With the exception
of pseudogenes, the two subspecies differ very little genetically. Over half of the
13
annotated pseudogenes (160) in each subspecies appear to be inactivated by
the same mutation, which suggests that they occurred before the subspecies
diverged and may have been important for the adaptation to a pathogenic
lifestyle, though a further 94 and 143 pseudogenes have occurred since these
subspecies split41. It is unknown whether all of the pseudogenes unique to each
subspecies contribute to the differences in their virulence, but it is reasonable to
hypothesize that at least some do. The F. tularensis subsp. holarctica LVS
chromosome is highly similar to that of a prototypical subsp. holarctica strain in
terms of synteny blocks and pseudogenes, though LVS has fifteen unique
deletions and pseudogenes that likely account for the strain’s attenuation50. A
deletion that lead to the loss of the pilin gene pilA and the partial deletion of the 3’
end of FTL_0391 and the 5’ end of FTL_0392 that created a gene fusion have
been shown to account for most, if not all of the attenuation of LVS in mouse
models of infection51. Interestingly, low-level genetic variation also exists within
the LVS lineage, some of which is the focus of this thesis.
Bacterial factors that mediate pathogenesis
For a pathogen of such high virulence, F. tularensis lacks many of the
common virulence factors used by other pathogenic bacteria, such as toxins,
hemolysins, and Types III or IV secretion systems26,52,53. While transposon
screens have identified hundreds of genes that affect the ability of the bacteria to
replicate in a host cell or cause disease in vivo, many of the mutations disrupt
genes that are important for the general fitness of the organism, such as those
14
involved in ribosome biogenesis or protein folding, and relatively few appear to
be dedicated virulence factors54-56. It should be noted, though, that a significant
portion of mutants identified in a F. tularensis Schu S4 transposon screen
encoded genes of unknown function or those without obvious homologs in other
bacterial species, suggesting that they may represent novel virulence factors
unique to Francisella.
The genes within the Francisella Pathogenicity Island (FPI) are among the
few that appear to be specific to the pathogenesis of the bacteria. Studies into F.
tularensis LVS during growth in the cytoplasm of J774 macrophage-like cell line
identified a number of bacterial proteins as being upregulated, including a 23 kilo
Dalton protein that would be later described as belonging to the FPI57.
Interestingly, Golovliov, et. al. noted that this protein was also highly induced
when the bacteria were exposed to hydrogen peroxide, suggesting that oxidative
stress may be a signal within the host cell that the bacteria are able to detect and
respond to with changes in gene expression. Soon after this discovery, F.
novicida transposon mutants were identified as failing to grow in murine
macrophages; one of the mutants mapped to the gene encoding the 23 kilo
Dalton protein and another mapped to a gene upstream within the same operon
(called the intracellular growth locus, or igl; the 23 kilo Dalton protein was named
IglC)58. Subsequent development of tools for molecular genetic analysis and
genome sequencing demonstrated that this locus was duplicated in the
chromosomes of the virulent F. tularensis26,59. These early genetic studies also
found that iglC mutants fail to inhibit the production of proinflammatory cytokines
15
in J774 cells and human peripheral blood monocytic cells, indicating that a major
component of F. tularensis pathogenesis is to suppress host responses that
typically would result in recognition and clearance of a bacterium60,61.
Further genetic analysis expanded the number of genes linked to the igl
locus, and showed that they belonged to a pathogenicity island of approximately
30 kilo bases and 19 open reading frames43. The GC content of the FPI is
considerably lower than that of the rest of the Francisella chromosome,
suggesting that it is derived from a horizontal transfer event or is the remnant of a
phage. Indeed, bioinformatic analysis indicates that the FPI encodes a secretion
system very distantly related to the Type VI secretion systems of other bacteria,
such as Pseudomonas and Vibrio species, and these systems have similarity to
the tail spike of certain bacteriophages62-64. Numerous studies have found
evidence that FPI proteins can gain access to the cytoplasm of infected cells,
though evidence suggests that some FPI proteins can still reach the cytoplasm in
the absence of a functioning FPI65-69. Recent structural analysis indicates that
the FPI does indeed encode a Type VI secretion system that likely functions
similarly to those of other bacteria70. While the mechanistic details still need
further investigation, it is clear that F. tularensis mutants lacking many of the
genes of the FPI are profoundly attenuated for virulence, often remaining trapped
in the phagosome of host cells71,72.
The FPI genes are regulated largely by the transcription factors FevR
(Francisella effector of virulence regulator) and the Francisella homologs of the
stringent starvation proteins A and B, called MglA and MglB (macrophage growth
16
locus)73,74. Mutants in these genes show attenuation in vitro and in vivo, and they
show differential expression of FPI genes and approximately 70 genes relative to
the wild type strains, supporting the idea that FPI-independent virulence factors
may exist elsewhere on the chromosome74,75. Interactions with the MglA/B-RNA
polymerase complex and FevR are stabilized by the small molecule guanosine
tetraphosphate, or ppGpp76. Production of ppGpp is essential for FevR
association with the RNA polymerase complex, as mutants that fail to produce
significant amounts of ppGpp have minimal FevR-bound RNA polymerase and
expression of FPI genes is substantially reduced. Our lab has identified mutants
that contribute to ppGpp generation and have shown that each mutant has more
or less severe intracellular growth phenotypes depending on the type of cell
being infected77,78. It is not currently known what stimulates production of ppGpp
in Francisella species, but there may be a connection to metabolic processes.
The FPI genes are also regulated by iron levels in their local environment, as
both lacZ reporter and proteomics analysis of F. tularensis LVS grown in iron
limiting media showed upregulation of FPI genes79,80.
In addition to the FPI, a major F. tularensis virulence factor is the Group 4
capsule and lipopolysaccharide (LPS) structures, which share the same
polysaccharide subunits81-84. The LPS of F. tularensis is tetraacylated rather than
penta- or hexaacylated, and the acyl chains tend to contain 16-18 carbons, rather
than 12-14 like that of other Gram negative bacteria. This atypical structure
renders F. tularensis LPS remarkably inert with respect to activation of TLR4
signaling, which is typically highly activated upon stimulation with LPS from other
17
species, such as E. coli or Salmonella85-88. Not only is the LPS inert, but it
appears that virulent F. tularensis is able to suppress host signaling pathways
that are activated upon typical LPS engagement of TLR4, suggesting that
avoidance of this pathway is highly important to the pathogenesis of F.
tularensis89. The capsule and LPS prevent complement-mediated lysis and can
bind to complement to mediate uptake in human macrophages through
complement receptor 3, a process associated with inhibition of proinflammatory
signaling. In addition to this, the capsule and LPS are essential for growth in the
host cell cytoplasm, as mutants that lack these structure rapidly induce host cell
death56. These mutants are highly attenuated in vivo, yet cause significant
inflammation and tissue damage, suggesting that capsule and LPS effectively
shield the bacteria from intracellular immune surveillance systems90,91. Our lab
and others have identified multiple mutants that disrupt various aspects of LPS
and capsule biosynthesis but have yet to identify mutants deficient in only one or
the other of these structures.
Iron uptake in F. tularensis
The iron uptake systems of F. tularensis are not dedicated virulence
factors, per se, however they contribute to the ability of the pathogen to cause
disease. Interestingly, Francisella species do not have a homolog of the TonB
system that powers the active import of siderophores in other bacteria, and thus
far no candidate gene to place this function has been found in the Francisella
genome to date92. Francisella species import iron through siderophore-bound
18
ferric iron via the Fsl and Fup systems, and ferrous iron via FeoB and FupA/B
with the latter fusion protein being unique to LVS93-96. The fsl operon encodes
proteins for the biosynthesis and uptake of the Francisella siderophore, which is
structurally identical to the fungal siderophore rhizoferrin95,97-99. Loss of
siderophore production or uptake does not affect intracellular growth or virulence
in vivo, indicating that other iron uptake mechanisms are able to
compensate94,100. Characterization of a spontaneous mutant of F. tularensis
Schu S4 that was highly attenuated in murine infections found that a deletion had
occurred within ORFs FTT_0918 and FTT_0919, creating a gene fusion101. The
coding sequences of these genes, now called fupA and fupB (Fe utilization
protein), also underwent a similar fusion event in the creation of LVS, with the
fusion protein mostly mediating siderophore uptake. Further analysis of the fupA
showed that it is a paralog of the siderophore receptor gene fslE, and that the
protein is located at the outer membrane, but somewhat surprisingly mediates
ferrous iron uptake in F. tularensis Schu S4; however, these iron molecules
would still need to pass through the inner membrane ferrous iron transporter
FeoB94,96,100,102,103.
In other bacterial species, the feo operon typically consists of feoAB, and
less frequently feoABC; the Francisella chromosome encodes feoA and feoB
separately and lacks feoC94,104. FeoB is a large transmembrane protein that has
an N-terminal G-protein domain and a multi-pass transmembrane C-terminal
domain105-107. In other organisms the Fe2+ import activity of FeoB requires
interaction with the small protein FeoA, though the details of how FeoA
19
stimulates FeoB activity are not clear108-111. The importance of iron in infection is
reflected by the numerous ways that bacterial pathogens have evolved to obtain
iron in a host, and several works have demonstrated that feoB contributes to or is
required for full pathogenesis of Salmonella, Campylobacter, Helicobacter,
Legionella, and others112-117. Multiple studies have characterized the Fe2+ uptake
function of the F. tularensis FeoB and have linked this to its pathogenesis, though
its contribution to oxidative stress resistance has not been thoroughly
investigated94,96,118.
Nearly all life on earth requires iron for essential metabolic processes,
including bacterial pathogens, yet the utility of iron in metabolism due to its ability
to accept and donate electrons makes it a relatively dangerous molecule inside of
a cell, as it can participate in Fenton Reaction chemistry119,120. F. tularensis
virulence is connected to the ability of the organism to avoid the detrimental
effects of reactive oxygen species (ROS), such as H2O2. The bacterium couples
glutamate metabolism pathways to H2O2 neutralization, and also maintains
optimally low levels of intracellular Fe2+ such that Fenton Reaction-mediated
damage appears to be minimized121,122. When this threshold is breached,
numerous antioxidant enzymes are able to protect the organism by detoxification
of H2O2; mutants lacking enzymes like the Dyp peroxidase, superoxide
dismutases and catalase are more sensitive to H2O2, fail to inhibit host
inflammatory signaling, and can be attenuated in murine infection models123-129.
20
Intracellular lifecycle
F. tularensis is a facultative intracellular pathogen that has the capacity to
gain access to the cytosol of multiple cell types including macrophages,
polymorphonuclear leukocytes, dendritic cells, lung airway epithelia, endothelia,
and hepatocytes and it is there that the bacteria acquire nutrients and proliferate
to high numbers78,130-136. The process of parasitizing the host cell cytosol begins
with uptake of the bacteria, which can differ depending on the subspecies of
bacteria used for the infection, the host cell type being infected, both in terms of
host lineage (mouse vs. human) and developmental origin (professional
phagocytes vs. epithelial cells, for example)137. Briefly, the process initiates
when the bacteria come into contact with a host cell and interact with one or
more host cell receptors, such as mannose receptor, complement receptor 3,
scavenger receptor-A, Fcγ receptor, etc138. Human monocyte-derived
macrophages take up the bacteria by an actin-dependent process termed
“looping phagocytosis” that requires intact capsule and lipopolysaccharide
carbohydrates on the outer surface of the bacteria, while epithelial cells appear to
utilize macropinocytosis for uptake of F. tularensis139,140. Phagocytosed bacteria
temporarily reside in a phagosome that acquires markers of early and late
endosomes, such as EEA-1 and LAMP-1, before the organism induces
degradation of the phagosome membrane141-144. The bacteria then have access
to the nutrient-rich cytosol where they grow until the death of the cell, whereby
the bacteria can begin the process again in a new cell145,146.
21
Opsonization mediates increased uptake of the bacteria, and depending
on the opsonin can have profound influence on the fate of the bacteria in terms of
phagosome escape kinetics, growth in the host cell cytosol, and whether an
inflammatory response is stimulated or suppressed138. For example, while
unopsonized F. tularensis typically interacts with the mannose receptor to
mediate phagocytosis, serum opsonization can lead to uptake via complement
receptors138,147. The complement cascade is a complex and highly regulated part
of the innate immune system that can directly kill bacteria via lysis or target them
for receptor-mediated phagocytosis by neutrophils or macrophages, but
Francisella species are resistant to complement-mediated lysis via their LPS and
capsule polysaccharides148,149. Additionally, these structures appear to help the
bacteria to subvert this important host defense by binding and converting
complement component C3b to the inactive C3bi, which does not contribute to
formation of the lytic membrane attack complex, yet still allows for receptor-
mediated phagocytosis149. Natural IgM readily binds to F. tularensis capsule and
LPS and mediates activation of the classical complement pathway, suggesting
that the bacteria have evolved in such a way that interaction with the complement
system may increase its pathogenesis150. Indeed, Dai, et. al. found that
complement receptor 3-mediated phagocytosis of F. tularensis Schu S4 lead to
suppression of TLR2 induced proinflammatory cytokine production in human
monocyte-derived macrophages, arguing that this contributes to the stealthy
nature of F. tularensis infections151. Interestingly, they report that this
immunosuppressive phenotype was not observed when the experiments were
22
repeated with murine bone marrow-derived macrophages, highlighting the
complexity of outcomes between experiment design, as well as the differences
between human and mouse responses.
Antibody-mediated phagocytosis also leads to considerably more uptake
of the bacteria in murine macrophages; however, the growth of the organisms is
restricted in a NADPH oxidase-dependent manner138. Regardless of the route of
entry into the cell, uptake in vitro is surprisingly low when compared to the very
small infectious dose of approximately 10 CFU via the respiratory route,
suggesting that mechanisms of uptake in the host environment may be
considerably different, or that the bacteria may persist and possibly grow
extracellularly before encountering sufficient numbers of host cells to mediate the
relatively rapid dissemination and high organ burdens observed in mouse models
of tularemia.
Francisella-containing phagosomes initially progress from having
characteristics of early to late endosomes, but the bacteria somehow inhibit
phagosome fusion to lysosomes and avoid damage by the protease cathepsin D
or the acidic nature associated with these vesicles141,143,144,152. Once inside the
phagosome F. tularensis avoids the destructive activity of the respiratory burst of
neutrophils and macrophages. The integral membrane proteins gp91phox and
p22phox come together with the soluble subunits gp47phox, p67phox and p40phox at
the phagosome membrane to form the NADPH oxidase153. This complex is
responsible for the production of high levels of toxic reactive oxygen species,
such as the superoxide anion, hydrogen peroxide, and hypochlorous acid (the
23
latter being specific to neutrophils)154. These molecules severely damage or kill
ingested microbes, though F. tularensis has evolved as yet undefined
mechanisms for preventing the NADPH oxidase from assembling or functioning
appropriately135,155. Neutrophils infected with F. tularensis do not show
colocalization of internalized bacteria and components of the NADPH oxidase nor
do the infected cells generate a respiratory burst; in fact, infected cells can
prevent this response when treated with compounds that typically stimulate
robust respiratory bursts135,156. Indeed, mice lacking genes coding the gp91phox
subunit of the NADPH oxidase did not exhibit dramatic differences in survival or
organ burdens when infected with F. tularensis Schu S4, suggesting that reactive
oxygen species generated by the NADPH oxidase do not significantly contribute
to controlling infection in wild type mice157. Genetic screening for mutants that
fail to inhibit the burst identified mutants that were defective in various aspects of
uracil metabolism; how these mutants stimulate a burst is not clear, though it may
depend in part on expression levels of genes from the Francisella Pathogenicity
Island158. That the bacteria avoid the respiratory burst strongly argues that a
major mechanism of pathogenesis for F. tularensis is to avoid the toxic effects of
host-induced reactive oxygen species.
Escape from the phagosome occurs shortly after phagocytosis, typically
within hours depending on the host species, Francisella strain, and the presence
or absence of opsonins141-143,159. Acidification of the phagosome via the vacuolar
ATPase has been shown to occur in vitro when unopsonized bacteria infect
murine macrophages; this acidification was important for egress from the
24
phagosome141. These results are in contrast with those of Clemens, et. al. who
found that opsonized F. tularensis do not induce acidification of the phagosome
in human macrophages, nor do they require this process for escape143,160.
Interestingly, prevention of the acidification of the phagosome has also been
reported to make F. tularensis LVS less able to acquire iron in the phagosome161.
As mentioned previously, the disparities between reports highlights the fact that
interpretation of results of these types of experiments cannot be generalized, as
differing conditions used to examine F. tularensis intracellular growth in vitro lead
to multiple outcomes, and it is currently not known if one or more experimental
paradigms better recapitulate the complexity of the conditions encountered in
vivo.
Degradation of the phagosome occurs by an unknown mechanism;
however, it is clear that many of the genes of the Francisella Pathogenicity Island
are required for this process, as certain mutants remain trapped in the
phagosome75,162. Once in the cytoplasm the bacteria begin to acquire nutrients
and replicate significantly faster than is observed in rich laboratory media, with F.
tularensis Schu S4 doubling in population size in approximately one hour141.
Being an auxotroph for numerous amino acids, F. tularensis must obtain them
from the host cytoplasm. One mechanism by which the bacteria do this appears
to be through the upregulation of host amino acid importers to increase the
cytoplasmic amino acid concentration. Barel et. al. found that infection of an
immortalized macrophage cell line with F. tularensis LVS lead to upregulation of
the host SLC1A5 amino acid importer, and that siRNA knockdown of its transcript
25
restricted LVS growth in the cytosol163. This group and others have also
identified a number of bacterial factors that mediate uptake of various amino
acids that are essential for growth in the cytoplasm of host cells, consistent with
the idea that evolution towards the intracellular parasitism of F. tularensis
relieved selective pressure for the maintenance of biosynthetic pathways to
produce these molecules, as the bacteria readily obtain them from their host164-
166. Scavenging amino acids and other nutrients from the host also has the
additional survival advantage of providing some of the end products of metabolic
pathways that are inactivated by H2O2 via oxidation of iron-sulfur cluster-
containing enzymes167.
F. tularensis burden in the cytosol of infected host cells can reach high
numbers, as microscopy of in vitro infection shows cells that appear to be entirely
filled with bacteria56,78,168. Presumably, the bacteria exhaust the host cell of its
nutrients, which leads to its death145,146. As with phagosome acidification and
escape, the type of cell death that an infected cell undergoes appears to be
specific to each Francisella species, as well as the species of the host cell being
used for infection. For example, F. novicida and F. tularensis LVS stimulates
pyroptotic cell death in an AIM2 and caspase-1-dependent fashion and have
been fruitful as tools to explore the molecular mechanisms underlying the
function of this host surveillance system, but this mechanism does not appear to
be involved in cell death induced by the virulent species169-172. Rather, apoptotic
cell death pathways induced by subsp. tularensis strains appear to be mediated
by caspase-3 in vivo173,174. Regardless of the pathway responsible, F. tularensis-
26
infected cells die and facilitate the further infection of adjacent or newly arrived
cells.
Immunity to Francisella species
One of the key features of the murine pneumonic tularemia model is the
rapid time to death (~5-7 days after infection), which precludes the development
of a robust adaptive immune response. This aspect of Francisella virulence has
made vaccine development a daunting challenge, and many labs have used
Francisella tularensis subsp. holarctica LVS as a less virulent substitute to study
aspects of B and T cell biology in sub-lethal infection and immunization, though it
is clear that the host responds to Francisella in a strain-specific fashion175,176. A
vaccine against fully virulent Francisella strains will almost certainly require
stimulation of cellular immunity, as studies of individuals who have recovered
from tularemia indicates a clear Francisella-specific T cell response, even years
after infection177-180.
Protective antibodies were an early target of research into immunity to
Francisella. In early studies utilizing a Type A strain, Foshay, et. al. showed that
transfer of hyperimmune serum from horses or goats provided protection to ~70-
90% of infected rats, although recent studies utilizing the mouse model have
found contradictory results181,182. Kirimanjeswara, et. al. found that murine
alveolar macrophages stimulated with interferon-γ were able to kill Schu S4 that
had been opsonized with serum from LVS-immunized mice, but that the immune
serum did not provide protection against Schu S4 during infection in vivo182.
27
While the latter is intriguing, Crane, et. al. have shown that Schu S4 may be able
to avoid this fate by interacting with the host serine protease plasmin. Schu S4
can bind to active plasmin, which can degrade Francisella-specific antibodies183.
The authors found that while antibody opsonization of both Schu S4 and LVS
resulted in an increase in TNF-α, the presence of active plasmin on Schu S4, but
not LVS, inhibited the induction of TNF-α in infected cells183. These results
highlight a difference between the two strains that may have important
implications for how virulent Francisella species avoid early detection and
control, and highlight significant pathogenesis mechanisms that will need to be
accounted for when designing a next generation vaccine against F. tularensis.
As with antibodies, the role of B cell activity in providing immunity to
Francisella species is complex; however, it is clear that mice lacking B cells are
less able to control or survive a Francisella infection184. Data is emerging that a
specific subset of B cells have divergent roles depending on the infecting strain
and the model of infection185,186. Immunization with lipopolysaccharide from LVS
generated a B1a cell-dependent protective response against an intraperitoneal
challenge with approximately 103 CFU of LVS185. In contrast, Crane, et. al.
utilized a short-term low dose antibiotic treatment after intranasal infection with
Schu S4 (the convalescent model) and observed that mice largely deficient in
B1a cells (XID mice) survived better than did wild type mice186. The increased
survival in these mice was found to be associated with a reduction in the anti-
inflammatory cytokine IL-10, which is a potent inhibitor of IL-12. The latter
cytokine stimulates interferon-γ production, and is necessary for the survival of
28
tularemia184. Curiously, IL-10 -/- mice succumb to standard intranasal Schu S4
infection similar to wild type mice, though it should be noted that the time scales
of infection are different between these studies and direct comparisons may not
be appropriate187. The immune responses observed in the convalescent model
would not have time to develop in a standard intranasal infection, although it
would be of interest to see if IL-10 -/- mice phenocopy XID mice in the
convalescent model of Schu S4 infection.
One possible target of the anti-inflammatory activity of IL-10 may be at the
interface of antigen presenting cells and T lymphocytes. Hunt, et. al., identified a
factor released by Francisella infected cells that stimulated IL-10-dependent
degradation of MHC class II molecules in macrophages188. Importantly,
supernatants from Schu S4 infected macrophages also stimulated the
downregulation of MHC class II molecules189. These data suggest that antigen
presentation to CD4+ T cells may be reduced in vivo. The same research group
and others have also previously found that F. novicida and LVS induce
prostaglandin E2 (PGE2) production in infected macrophages189,190. PGE2 has
been shown to inhibit macrophage maturation, and might play a similar role in
downregulating pathways important for intracellular killing of Francisella, as in
Burkholderia pseudomallei infections191,192. Interestingly, they showed that PGE2
induced by Francisella infection inhibited T cell proliferation and interferon-γ
production in vitro. It has long been known that T cells are important for bacterial
control and immunity to Francisella179,184,193. Both CD4+ and CD8+ T cells were
required for survival of a primary Schu S4 infection in the convalescent model of
29
tularemia, but only partial protection against a secondary challenge was
observed for wild type mice (66% survival)184. It is intriguing to speculate that the
partial protection may be a function of a less than optimal T cell response
mediated by PGE2, both in terms of antigen presentation and T cell proliferation.
Considerable effort has been expended testing potential vaccines and
vaccine strains against infection with F. tularensis, yet as with the disparities
between in vitro infection outcomes, these studies all differ depending on the
details of experimental setup. The differences in protocols include route of
inoculation (intranasal, subcutaneous, or intradermal), immunizing agent (live
versus dead organisms, membrane fractions, purified LPS/capsule, passive
transfer of serum or antibodies), challenge strain (LVS or fully virulent F.
tularensis), and model system (mouse, rat, rabbit, etc.)194. As such, comparisons
between these studies are difficult to interpret. Thus, the major goal of the work
presented in this thesis was to develop a foundation for future understanding of
what makes F. tularensis LVS the “gold standard” live vaccine strain using the
tools of bacterial genetics.
30
CHAPTER II
A VIRULENT BIOVAR OF F. TULARENSIS LVS IS INTRINSICALLY MORE
RESISTANT TO HYDROGEN PEROXIDE
Introduction
Prior to initiating my thesis work I spent considerable time developing and
testing a low and moderate dose immunization experiment with two attenuated
capsule/O-antigen mutants, waaY and wbtA, that had been identified in
transposon screens from my lab56,91. In this experiment (Appendix A) I included
LVS as a positive control for protection against intranasal challenge against F.
tularensis Schu S4, and found that LVS significantly outperformed waaY and
wbtA, despite the latter strains’ attenuation and stimulation of an obvious host
response (ruffled fur, hunched shoulders, conjunctivitis, etc.). Concurrent with
this experiment, our collaborator at Rocky Mountain Laboratories, Catharine
Bosio, shared data indicating that two LVS stocks in her possession differed in
virulence and ability to provide protection against Schu S4. She graciously
shared these strains and their genome sequences, so I began to explore the
phenotypic differences between these and my lab strain in an effort to
understand the genetic requirements of LVS to cause disease and stimulate an
immune response in a mouse.
Despite significant differences in virulence, members of the Francisella
genus are very similar at the genetic level; many of the observed differences are
genomic rearrangements and single nucleotide polymorphisms (SNPs)41,50. The
similarities across the Francisella strains has made identification of the specific
31
factors that mediate the high level of virulence displayed by Type A and Type B
strains difficult, although genome comparisons between F. tularensis subsp.
tularensis Schu S4 and less virulent LVS have identified some of the attenuating
mutations. These include fusion of the fupA/B genes and loss of pilA102. Another
area of significant difference between the highly virulent Type A strains and the
intermediately virulent type B strains (and LVS) is the level of intracellular iron,
with virulence being inversely correlated with iron levels122. Iron is an essential
micronutrient, and numerous studies have shown that it is highly sought after by
bacterial pathogens, as host sequestration of iron (a part of nutritional immunity)
can restrict the growth of numerous pathogens113,195. Apparently, F. tularensis
may represent an exception to this paradigm as Lindgren, et al., found that
higher virulence subspecies had low levels of intracellular iron122.
Phenotypic variability exists between different LVS biovars in terms of
virulence and vaccine efficacy51,196. These differences provide a naturally
occurring genetic experiment to identify bacterial factors that correlate with
vaccine efficacy. To explore this, the genomes of two LVS biovars were
sequenced (Bosio lab), previously labeled RML (higher virulence; provides better
protection against Schu S4 challenge) and ATCC (lower virulence; poor
protection against Schu S4 challenge) and I sought to characterize the genetic
differences here. Interestingly, two mutations occurred in genes known to be
involved either in iron uptake (FTL_0133, feoB) or oxidative stress resistance
(FTL_1773, dyp peroxidase). FeoB is an inner membrane protein that imports
ferrous iron into the bacterial cytoplasm, and has previously been linked to
32
virulence as one of two major iron uptake pathways in LVS94,96,197. Dyp is a
member of a broad class of dye-decolorizing peroxidases with wide substrate
specificity and important industrial uses198. In agreement with the data reported
in this thesis, Binesse, et. al. characterized the LVS dyp allele with a 93 base pair
deletion as contributing to sensitivity to H2O2, and found that complementation
with a full length dyp (found in Schu S4 and some LVS lots) restored H2O2
resistance51,127. These data suggest that resistance to host-induced oxidative
stress is a significant component of the virulence of F. tularensis, and may
contribute to the vaccine efficacy of LVS given that the ATCC biovar of LVS has
the 93 base pair deletion and was relatively attenuated in murine infections and
stimulated weak immunity against a challenge of F. tularensis Schu S4196.
Materials and Methods
Bacterial strains and growth conditions: Three different biovars of F.
tularensis subsp. holarctica LVS were used in this work - the LVS strain used in
the Jones lab for several years, here called Iowa LVS (University of Iowa) and
the RML LVS and ATCC LVS, (Rocky Mountain Laboratories). Bacteria were
routinely cultured on modified Mueller-Hinton (supplemented with 1% glucose,
0.025% ferric pyrophosphate, and 2% IsoVitaleX) agar or in broth, with 50 μg/mL
of kanamycin or spectinomycin, as needed. Mueller-Hinton agar was also
prepared without ferric pyrophosphate supplementation to assay for growth in
lower iron conditions. Bacteria were also cultured in Chamberlain’s defined
medium with either 35 μM FeSO4 or 350 nM FeSO4, supplemented with
33
antibiotics as needed. Agar plates were incubated at 37°C with humidity and 5%
CO2 while broth cultures were grown at 37°C, shaken at 200 rpm.
Mutant construction and complementation: Deletion of feoB (FTL_0133) and
dyp (FTL_1773) gene was achieved by homologous recombination using
derivatives of the non-replicating plasmid pJC84. Primer sequences are shown
in Table 1. Briefly, upstream and downstream flanking DNA was amplified via
PCR, and amplicons were cloned into the multiple cloning site of pCR2.1 (Life
Technologies). A spectinomycin resistance cassette was cloned into the AvrII
site in the 3’ end of each upstream PCR fragment. The upstream-spectinomycin
resistance fragment was removed by digestion with AscI, and cloned into the
AscI site in the 5’ region of the downstream PCR plasmid. The entire upstream-
spectinomycin resistance-downstream fragment was cloned into the BamHI site
of the suicide plasmid pJC84. Finally, the spectinomycin resistance cassette was
removed by digestion with AvrII and the plasmid was re-ligated. Plasmids were
electroporated into LVS (2.5 kV, 25 μF, and 600 Ω), and the bacteria were plated
onto MMH agar with 50 μg/mL kanamycin after 2-3 hours of outgrowth.
Organisms were then grown overnight in broth lacking kanamycin, and were
plated onto MMH agar with 8% sucrose for sacB-mediated counter-selection.
Kanamycin sensitive colonies were screened by colony PCR to detect deletion of
the feoB or dyp gene. Primers were designed such that they flanked the coding
sequence of each gene such that an amplicon would be produced regardless of
genotype. For complementation of each mutant, the coding sequence of each
34
allele of feoB or dyp was PCR amplified using the Phusion high fidelity
polymerase. Primers were designed such that 5’ KpnI and 3’ SalI sites were
included for cloning purposes. The PCR amplicons were purified and A-tailed
with Taq polymerase (New England Biolabs) per the manufacturer’s instructions
and cloned into pCR2.1. Inserts were removed by digestion with KpnI and SalI,
and cloned into the respective sites downstream of the groE promoter in
pTrc99a; the groE promoter has a 5’ BamHI site. The entire PgroE-coding
sequence was moved to the BamHI and SalI sites of pBB103 to create the final
plasmids. The waaY::Trgtn mutants were constructed as described previously56.
Iron-regulated β-galactosidase reporter activity: To assess iron uptake-
regulated gene expression, Miller assays were performed with Iowa, RML, and
ATCC LVS carrying the fslA promoter-lacZ fusion on the pBB103 plasmid77.
Bacteria were grown in Chamberlain’s defined medium (supplemented with 50
μg/mL spectinomycin) to late-log phase, and β-galactosidase activity was
assayed using the standard Miller assay199.
FeoB function in a heterologous reporter system: The coding sequence of
feoB was PCR-amplified from both Iowa and RML LVS using the high fidelity
Phusion polymerase (New England Biolabs). Sanger sequencing was performed
to confirm that the SNP observed in the RML LVS background was present. The
feoA coding sequence was cloned into the KpnI/SalI sites of a derivative of
pTrc99a, immediately downstream of the groE promoter. The entire PgroE-feoA
35
fragment was removed by digestion with BamHI and SalI, and ligated into the
same sites in pBB103. Both this plasmid and the pWKS30 containing feoB were
transformed into the E. coli H1771 strain (MC4100 aroB feoB7 fhuF::plac Mu).
Control strains were also generated that carried only one of the plasmids. Miller
assays were performed to measure FeoB Fe2+ import activity via the readout of
Fur repression of fhuF::placZ. All strains were grown in LB supplemented with 50
μg/mL of spectinomycin and 100 μg/mL of ampicillin as necessary.
Hydrogen peroxide sensitivity: To measure resistance to H2O2-mediated
killing, mid- to late-log phase organisms were pelleted at 13,200 RPM for 5 min,
washed in PBS, and ~106 CFU were resuspended into 200 μL PBS with or
without 100 μM fresh hydrogen peroxide (H2O2) in a 96-well dish. The samples
were incubated at 37°C with humidity and 5% CO2 for thirty minutes or one hour.
The culture from each well was then serially diluted, plated onto modified
Mueller-Hinton agar (with antibiotic where necessary) and the number of
surviving organisms for each strain was enumerated after 2-3 days.
In vitro infections: Murine bone marrow-derived macrophages (BMM) were
generated by harvesting bone marrow from C57BL/6 mice, 6-10 weeks old. The
bone marrow was differentiated in DMEM supplemented with 10% heat-
inactivated fetal bovine serum (FBS), 10% L929 cell-conditioned DMEM and Pen
Strep (Gibco) for 5-6 days. Subsequently, non-adherent cells were removed by
PBS wash, and adherent cells were lifted by a brief incubation with Versene
36
(Gibco) for ~5-6 minutes at 37°C. Cells were enumerated and seeded into 24
well dishes at a density of 3x105 per mL in DMEM with 10% L929-conditioned
media, 10% heat-inactivated FBS. Multiplicity of infection (MOI) was estimated
by measuring the OD600 value of mid to late-log phase grown bacteria, and
confirmed via serial dilution onto modified Mueller-Hinton agar and enumeration
after 2-3 days growth at 37°C with humidity, 5% CO2. Bacteria were allowed to
associate with the BMMS for 3 hours, after which time the cells were washed 3x
with PBS, and incubated in 10% L929-conditioned DMEM supplemented with
100 μg/mL of gentamicin for one hour, cells were then washed 3x with PBS and
either lysed or incubated in antibiotic-free media for a further 20 hours. At the 4
and 24-hour time points 0.1% saponin was added to each well. Wells were
scraped and vigorously pipetted to disrupt the cells and liberate intracellular
bacteria for serial dilution and enumeration. Infection of A549 cells was
performed identically as BMM infections, however the cells were maintained in
DMEM with 10% heat-inactivated FBS with Pen Strep as needed, and cells were
lysed with saponin at the four-hour time point rather than at three hours.
Results
LVS biovars have similar PiglA-lacZ expression and lack low molecular weight O-
antigen glycosylated proteins
Prior to having access to the genome sequences of RML LVS and
ATCC LVS I investigated two phenotypes putatively connected to virulence that
our lab had identified as differing between virulent strains and LVS: FPI gene
37
reporter expression and the presence of protein species glycosylated with O-
antigen sugars. The Francisella Pathogenicity Island (FPI) encodes numerous
genes essential for the pathogenesis of Francisella, and as such the genes at
this locus are known to be highly expressed. Previous data from our laboratory
has shown that a PiglA-lacZ reporter has different activity levels between
subspecies, with expression in the Type A strain F. tularensis subsp. tularensis
Schu S4 being about two-fold higher than the Type B F. tularensis subsp.
holartica. The LVS exhibited the lowest level expression, approximately one-
third that of Schu S4 expression. While we do not have evidence that these iglA
expression level differences mediate the relative virulence of each strain, it
remains an intriguing observation that virulent strains have higher expression of
fundamental virulence factor genes. To test the hypothesis that the more virulent
RML LVS biovar had higher expression of a FPI gene relative to two less virulent
biovars, I introduced the PiglA-lacZ reporter plasmid to each strain and performed
Miller assays after overnight growth. As seen in Figure II.1A, no significant
difference was observed between the strains under these conditions, suggesting
that the difference in virulence between them is not likely due to differential
regulation of the igl operon.
Additionally, it was previously shown by our lab that the virulent
subspecies of F. tularensis produce low molecular weight O-antigen positive,
proteinase K sensitive protein species while the Iowa LVS does not81. To test the
hypothesis that the more virulent RML LVS biovar produce such species, I
constructed a waaY::TrgTn mutant in both RML LVS and ATCC LVS; an identical
38
mutant in the Iowa LVS background was constructed previously. Each sample
was prepared with or without overnight treatment with proteinase K, separated by
SDS-PAGE, and blotted with an antibody that recognizes F. tularensis O-antigen
sugars (FB11). As shown in Figure II.1B, the positive control F. tularensis Schu
S4 waaY::trgTn sample has several low molecular weight bands that are not
present when the sample is treated overnight with proteinase K. Like the Iowa
and ATCC LVS (latter strain not shown), the RML LVS sample did not have
these bands irrespective of treatment, suggesting that the higher virulence of this
strain is not likely due to the presence of O-antigen glycosylated proteins.
RML LVS has less intracellular Fe2+ than the ATCC or Iowa LVS
Genome sequences of the RML and ATCC LVS indicated a
nonsynonymous substitution in the RML LVS feoB allele that changed an
aspartate at residue 471 to a tyrosine (Fig. II.2). The feoB D471Y allele
represented an interesting target to identify the virulence difference between the
RML and the ATCC strain due to the well characterized nature of FeoB function
in Fe2+ uptake and the connections between iron content and virulence. The
D471Y mutation maps to a cytoplasmic loop, adjacent to a highly conserved
glycine residue that is predicted to be functionally important by ConSurf200,201.
The substitution of a large, hydrophobic tyrosine for an aspartate at residue 471
led us to hypothesize that the FeoB encoded by the RML allele imports Fe2+
poorly, or not at all. Although the genome of the lab stock of LVS from the Jones
lab (hereafter referred to as “Iowa”) has not been completely sequenced, the
39
strain was also included in the experiments described here. Sanger sequencing
confirmed that the Iowa feoB encodes an aspartate at residue 471 and a full
length dyp peroxidase, suggesting that it is potentially an intermediate between
the RML and ATCC strains.
We tested the hypothesis that the RML LVS has less intracellular iron than
either the Iowa LVS or the ATCC LVS by growing the strains overnight in
standard modified Mueller-Hinton (MMH) broth, serially diluting in PBS, and
spotting ten-fold dilutions onto MMH agar of varying concentrations of iron (Fig.
II.3A). Growth was identical among each strain when spotted onto the control
agar containing the typical concentration of iron (0.0025% ferric pyrophosphate)
routinely used for propagation of F. tularensis. When the iron concentration was
reduced by 50%, both the Iowa and ATCC strains had growth patterns similar to
that observed on the control plate, however, the RML strain exhibited growth
restriction at this concentration of iron. Fewer isolated colonies were observed
and lawn growth was less luxurious when compared to the Iowa and ATCC LVS.
On MMH agar lacking added iron, all three strains exhibited growth restriction,
however the reduced growth phenotype of RML was exacerbated, with extremely
poor growth even at the lowest dilution plated. Iowa and ATCC strains both grew
as lawns at these dilutions, indicating that they were able to scavenge sufficient
iron from the agar plate environment, while the RML strain could not. The
decrease in the RML strain growth was approximately two orders of magnitude
greater than that observed for either the Iowa or ATCC LVS, consistent with the
hypothesis that the RML strain has less intracellular iron under conditions of iron
40
limitation. In support of this, disc-diffusion assays on MMH agar lacking added
iron showed that the RML LVS grew smaller halos around discs soaked with
ferric pyrophosphate solutions of varying concentrations (Fig. II.3B).
As the RML strain grew less well under iron limiting conditions and the
feoB gene encodes a ferrous (Fe2+) iron importer, we hypothesized that the RML
strain would have less intracellular Fe2+. Levels of intracellular Fe2+ are tightly
regulated, as free iron can participate in the toxic Fenton reaction with H2O2,
causing cellular damage in the forms of lipid peroxidation, DNA damage, and
most importantly, poisoning of the iron-sulfur cluster enzymes of essential
metabolic pathways. To avoid this toxicity, bacteria have evolved transcriptional
regulatory mechanisms to maintain iron homeostasis. One such example is the
Fur system202. The Fur transcriptional repressor is an allosteric regulator that
uses Fe2+ as a co-repressor; when iron levels are sufficient, the Fur-Fe2+
complex binds to Fur Box sequence motifs in the promoters of iron uptake genes
to mediate their repression. When iron is limiting, less Fe2+ is bound to Fur,
decreasing its ability to bind DNA. This leads to de-repression of iron uptake
genes; the Fur regulatory system has been described to function similarly in
Francisella95,97,98. To assess F. tularensis LVS intracellular Fe2+ levels under iron
limiting conditions, a Fur regulated PfslA-lacZ construct was introduced into the
RML, Iowa, and ATCC strains. Iron limitation is known to relieve Fur repression
at the fslA promoter, so increasing gene expression correlates with decreasing
concentrations of intracellular Fe2+77. Each bacterial strain was grown in
Chamberlain’s Defined Medium (CDM) with 35 μM (high iron) or 350 nM (low
41
iron) FeSO4, and β-galactosidase activity was measured after overnight growth
(Fig. II.4). As expected, iron starvation induced high expression of PfslA-lacZ in all
three strains; however, induction in the RML LVS was approximately 2-fold
higher than that observed for Iowa LVS or ATCC LVS. The higher activity of the
PfslA-lacZ construct in the RML strain is consistent with the hypothesis that this
strain has lower levels of intracellular Fe2+ when iron is limiting in the growth
medium.
To assess iron levels in the RML LVS by an independent method, we
genetically manipulated the ability of each strain to acquire iron by
overexpression of Fur. The F. tularensis fur gene was placed under the control
of the groE promoter, and the construct introduced into each strain. Bacteria
were grown in MMH broth with normal iron concentrations overnight, serially
diluted in PBS and plated onto standard MMH agar (Fig. II.5). The Iowa and
ATCC strains, when overexpressing Fur, exhibited colony growth similar to that
of their vector controls, with no obvious decrease in the size of isolated colonies
or robustness of lawn growth at lower dilutions. The RML LVS, however, had
much smaller colony size and poor lawn formation, indicating a general growth
delay that was not observed in the RML vector control strain. We interpret these
results to indicate that Fur overexpression in RML LVS had a greater impact on
growth because the strain has significantly smaller pools of intracellular iron and
therefore inhibition of iron uptake systems has a greater impact on RML LVS
growth and survival than for the Iowa LVS or ATCC LVS, even though all strains
were grown in iron-rich media. This is consistent with the hypothesis that
42
intracellular Fe2+ pool of RML LVS are smaller than in Iowa LVS or ATCC LVS
and become limiting for growth in conditions that are not limiting for the other two
strains.
FeoB D471Y does not complement E. coli ΔfeoB fhuF::λplac
The low iron growth phenotypes and Fe2+-regulated gene expression of
RML LVS strongly suggest that the FeoB D471Y protein encoded by the strain
has significantly reduced Fe2+ import activity. To assess the ability of the FeoB
D471Y protein to transport iron, we made use of a well-characterized E. coli iron
reporter strain that lacks feoB (Fig. II.6)111,203. This strain, E. coli H1771, has a
chromosomal lacZ insertion in the Fur-regulated gene fhuF, and has high β-
galactosidase activity as a result of low Fe2+ levels. The feoB alleles from Iowa
and RML (D471Y) were introduced into H1771 on the low copy number plasmid
pWKS30, and Miller assays were performed. Initial experiments supplying only
feoB (either allele) showed no repression of fhuF::lacZ, indicating that FeoB
alone is not sufficient for Fe2+ import in this reporter system (data not shown). In
most bacterial species encoding a feo system, the small gene feoA is encoded
either in an operon with feoB or elsewhere on the chromosome and the encoded
protein is required to stimulate FeoB Fe2+ import activity104,109,111. When the
Francisella feoA was supplied on pBB103 in concert with feoB (Iowa/ATCC
allele) on pWKS30, significant repression of the fhuF-lacZ reporter was observed
(~ 80% reduction in expression relative to the empty vector control), indicating
that Fe2+ import was significantly increased. In contrast, complementation of
43
H1771 with feoA and feoB D471Y (RML allele) failed to repress fhuF::lacZ. This
result indicates that the FeoB D471Y protein failed to significantly import Fe2+
and further supports the hypothesis that RML has less intracellular Fe2+ under
conditions of iron limitation in vitro. I included the Schu S4 feoB allele in this
analysis as it differs from the annotated LVS allele by four amino acids, however
significant reduction in fhuF::lacZ reporter activity was observed, indicating that it
is competent to import ferrous iron.
Resistance to H2O2 correlates with feoB D471Y allele
Since the RML LVS had less intracellular Fe2+ and Lindgren et. al.
demonstrated that strains with lower intracellular iron concentrations were more
resistant to H2O2, we hypothesized that RML LVS would be more resistant to
killing by H2O2 than Iowa LVS or ATCC LVS. Bacteria were exposed to PBS
alone or PBS with 100 μM H2O2 in a 96-well dish for one hour at 37°C with 5%
CO2 and humidity. Following incubation, the bacteria were serially diluted in PBS
and plated onto MMH agar (Fig. II.7). The RML strain was approximately ten-fold
more resistant to H2O2 than the Iowa strain, and approximately one-hundred-fold
more resistant than ATCC. The genome of ATCC encodes a Dyp peroxidase
with a 93 base pair deletion; the contribution of Dyp function to H2O2 resistance is
described in a later section.
Given the increased sensitivity of Iowa LVS to H2O2, and the functional
complementation of an E. coli feoB mutant by the Iowa feoB allele, we next
tested if the RML LVS could be sensitized to H2O2 by increasing the intracellular
44
Fe2+ pool via overexpression of a functional feoB (Fig. II.8). This was achieved
by transforming the strain with a plasmid carrying the overexpression constructs
Pgro-feoB (Iowa/ATCC allele - high Fe2+ transport) and Pgro-feoB (D471Y; RML
allele - low Fe2+ transport). H2O2 sensitivity assays were performed as described
previously, however, a 30-minute time point was also included in case
overexpression lead to increased susceptibility such that no viable organisms
would be recovered at one hour. Overexpression of either feoB allele led to
increased sensitivity to H2O2; however, the effect mediated by the Iowa/ATCC
feoB was orders of magnitude greater than that of the poorly functional RML feoB
D471Y allele. Nearly one log of killing was observed at 30 minutes for the RML +
Pgro-feoB (Iowa/ATCC), while RML + Pgro-feoB (D471Y) only had a two-fold
reduction in viability. The effect was more pronounced at the one-hour time
point, with no viable bacteria recovered at the level of detection (102 CFU) from
RML + Pgro-feoB (Iowa/ATCC). In contrast, over 104 CFU were recovered at one
hour from the RML + Pgro-feoB (D471Y). These data demonstrate that the SNP
in feoB encoded in the RML LVS genome mediates increased resistance to
H2O2.
Although there were few differences between the genomes of RML LVS
and ATCC LVS, it was possible that the H2O2 sensitivity of RML LVS
overexpressing the Iowa/ATCC feoB allele may have been due to genetic
interactions between feoB and SNPs unique to the RML LVS and not feoB alone.
To confirm that the H2O2 resistance phenotype conferred by feoB D471Y is not
unique to the RML genetic background, I constructed a ΔfeoB mutant in the Iowa
45
LVS background and complemented it with the Pgro-feoB constructs (Fig. II.9).
While the genome of this strain is not sequenced, the Iowa LVS was chosen over
the ATCC LVS, as the latter strain was approximately 100 fold more sensitive to
H2O2 (see below). H2O2 sensitivity assays were performed as before and,
consistent with results in the RML strain background, overexpression of feoB
D471Y mediated significant resistance to H2O2 relative to that when wild type
feoB was overexpressed. This result strongly supports the hypothesis that the
feoB allele in the RML LVS genome encodes a FeoB protein with minimal ferrous
iron import and that this mediates higher resistance to the lethal effects of H2O2.
Resistance to H2O2 also requires Dyp peroxidase
The ATCC LVS was observed to be approximately two-to-three orders of
magnitude more sensitive to the killing effects of H2O2 than the RML strain. The
ATCC LVS genome, in addition to encoding a feoB that is competent for robust
Fe2+ import, encodes a dyp-type heme peroxidase that is missing nucleotides
737 to 830 relative to the first nucleotide of the coding sequence. This result in a
protein that lacks 31 amino acids near the C-terminus, and bioinformatic analysis
indicates that this deleted region encodes a conserved phenylalanine residue
that is part of the heme-binding pocket, a critical functional component for Dyp
enzymatic activity in other organisms198. Dyp peroxidases are known for having
a wide range of substrates, so ascertaining the natural function of the enzyme in
vivo has been difficult, though as a peroxidase Dyp uses H2O2 as a cofactor in its
reactions.
46
To assess the contribution of Dyp to H2O2 resistance, I constructed a ∆dyp
deletion mutant in the Iowa LVS and performed H2O2 sensitivity assays as
described in the materials and methods (Fig. II.10). The ∆dyp mutant was
significantly more sensitive relative to the parental Iowa strain, indicating that the
Dyp enzyme provides substantial protection against the killing activity of H2O2.
This indicates that the 93 base pair deletion observed in the ATCC LVS dyp
allele may disrupt its function in this capacity, rendering it highly sensitive to
H2O2. To confirm that the loss of these 93 base pairs is what abrogates Dyp
function in the ATCC LVS, I generated two plasmid-borne complementation
constructs with each dyp allele being expressed from the groE promoter. H2O2
sensitivity assays were performed as before and the ∆dyp mutant that had been
complemented with the full length allele exhibited both a reduction in overall
viability and an increase in sensitivity to H2O2, while the strain complemented
with the 93 base pair deletion allele exhibited modest partial complementation
and no obvious defect in overall viability. While surprising at first, this may be
consistent with the identity of Dyp as a heme-binding protein. Overexpression
may result in a concentration of Dyp protein that far exceeds its normal levels; in
fact, a Pdyp-lacZ reporter exhibits negligible β-galactosidase activity (data not
shown), suggesting that the gene is expressed at low levels. In this scenario the
full length protein may be binding and sequestering a larger percentage of the
cell’s available heme such that other enzymes requiring heme, such as the
catalase KatG, lack this essential cofactor and are thus catalytically inert. The
truncated protein encoded by the allele with the 93 base pair deletion is predicted
47
to lose a conserved residue in the heme-binding pocket and may bind negligible
quantities of heme when expressed naturally in the ATCC LVS, leading to an
inactive Dyp protein and H2O2 sensitivity in this strain. Alternatively,
overexpression of wild type Dyp may lead to levels of the enzyme that exceed its
natural substrate(s), and it may erroneously oxidize/reduce molecules important
for oxidative stress resistance.
Neither feoB nor dyp are essential for intracellular growth in A549 or BMM cells
After establishing the importance of both FeoB D471Y and Dyp in
resistance to the lethal effects of H2O2 I examined their contributions toward
intracellular growth in two cell types, the human adenocarcinoma A549 cell line
and primary murine bone-marrow-derived macrophages. The former cell line
was chosen for in vitro infection because a feoB mutant in the ATCC LVS lineage
was reported to exhibit poor intracellular growth over twenty-four hours, and the
latter cell type was chosen as growth within macrophages is a preferred niche for
F. tularensis96. No significant difference in growth in either cell type was
observed for the three LVS biovars, approximately 100-fold for A549 cells and
approximately 10-fold for BMMs (Fig. II.11 and Fig. II.13). These data suggest
that the genetic differences between these strains do not detectably alter
intracellular growth kinetics.
As it had already been established that the ATCC LVS feoB mutant
exhibited a growth defect in A549 cells, I generated clean deletion mutants in all
three parental strains and performed in vitro infections to see if this was a
48
general ∆feoB phenotype, or if the phenotype was unique to the ATCC LVS
genetic background (Fig. II.12). Surprisingly, only the ATCC LVS ∆feoB mutant
exhibited an intracellular growth defect, while the mutants in the Iowa LVS and
RML LVS genetic backgrounds grew similar to their respective parental strains.
During the course of these infections it was observed that approximately 1.5- to
2-fold higher CFUs were obtained at the four-hour time point for the ATCC LVS
∆feoB. Previous work in our lab has identified increased uptake as a phenotype
associated with genetic defects in capsule and LPS biosynthesis, so an
immunoblot for each of these structures was performed on the ATCC LVS ∆feoB
(not shown). This revealed that the the ATCC LVS ∆feoB indeed lacked capsule
and LPS, possibly due to an undefined second site mutation; it was not pursued
further.
49
0
20
40
60
80
100
120
140
160
Iowa RML ATCC fevR
Mill
er u
nits
(nor
m. t
o no
pro
mot
er
cont
rol)
PiglA-lacZ β-galactosidase activity
Figure II.1. RML LVS has similar FPI reporter gene expression and lacks glycosylated proteins seen in virulent Type A and Type B strains of F. tularensis. A) Promoter activity of the Francisella Pathogenicity Island gene iglA in three LVS biovars. Miller assays were performed as described in the materials and methods. No significant differences were observed between Iowa, RML, or ATCC LVS. Shown is combined data from three replicates. B) Immunoblot analysis of low molecular weight O-antigen positive, proteinase K sensitive bands observed in virulent F. tularensis subsp. tularensis and subsp. holarctica, but not in Iowa LVS. O-antigen ligase (FTT1238/FTL0706c) mutants were generated via the TargetTron method. Samples were treated with proteinase K (PK) overnight and separated by SDS-PAGE. Membranes were probed with the anti-O-antigen monoclonal antibody FB11. Shown is a representative blot from two replicates.
A B
50
Figure II.2. Location of the amino acid residue that is mutated in the RML LVS FeoB. The protein sequence was submitted to HMMTOP to predict transmembrane domains, then visualized using TMRPres2D1,2. The mutated residue is predicted to occur in a cytoplasmic loop near the membrane and its location is indicated by an arrow in the figure above.
51
Figure II.3. RML LVS exhibits growth restriction under iron limitation on MMH agar. A) Bacteria were grown in standard MMH broth overnight, then serially diluted in PBS and spotted onto MMH agar containing 0.025%, 0.0125 %, or no added ferric pyrophosphate. B) Bacteria were grown in standard MMH broth overnight, and then ~106 CFU were spread onto MMH agar lacking added ferric pyrophosphate. Sterile discs were then placed onto the plates and 10 μl of ten-fold dilutions of ferric pyrophosphate were spotted onto the discs: 0.025% (upper left), 0.0025% (upper right), 0.00025% (bottom left), or sterile water (bottom right). Shown are representative data from two similarly designed replicates.
A
B
52
0.00
500.00
1000.00
1500.00
2000.00
Iowa RML ATCC
Mill
er u
nits
PfslA-lacZ activity 1x Fe
0.01x Fe
Figure II.4. RML LVS has iron-related gene expression consistent with less intracellular Fe2+ in iron limiting media. Each strain carried the PfslA-lacZ plasmid which is responsive to changes in iron concentrations. Bacteria were grown overnight in Chamberlain’s defined medium with either 35 μM FeSO4 or 350 nM FeSO4 and Miller assays were performed as described in the materials and methods. Shown is a representative Miller assay from three replicates.
53
Figure II.5. Constitutive fur expression inhibits robust growth of RML LVS. Each strain was transformed with the vector control plasmid pBB103 or pBB103 with Pgro-fur and grown overnight in standard MMH broth. Bacteria were serially diluted in PBS and plated onto standard MMH agar. Shown is a representative assay.
54
0.00
100.00
200.00
300.00
400.00
500.00
600.00
Empty vector V. choleraefeoABC
Ft feoAB Ft feoAB D471Y Schu S4 feoAB
Mill
er u
nits
E. coli ΔfeoB fhuF::λplac Mu activity
Figure II.6. RML feoB allele does not complement an E. coli ΔfeoB iron reporter strain. E.coli H1771 was transformed with the control plasmids pWKS30 (empty vector), the positive control plasmid with the Feo system from V. cholerae in pWKS30, or pWKS30 expressing either LVS feoB allele, or the Schu S4 feoB allele. The latter three strains also carried derivatives of pBB103 with Pgro-feoA as FeoB is known to require FeoA in other bacterial species. Miller assays were performed as described in the materials and methods. Shown is a representative assay from three replicates.
55
1.E-01
1.E+00
1.E+01
1.E+02
Iowa RML ATCC
Perc
ent s
urvi
val (
norm
aliz
ed to
RM
L)
Survival 100 μM H2O2
Figure II.7. RML LVS is significantly more resistant to H2O2 lethality. Bacteria were grown in standard MMH broth overnight, centrifuged at 13,200 rpm for 5 minutes, and re-suspended in sterile PBS with or without 100 μM H2O2 in 200 μl total volume in a 96 well dish. These were incubated without shaking for one hour at 37° C, 5% CO2 before serial dilution in PBS and spotting onto standard MMH agar. Shown is a representative assay from >3 replicates.
56
1.E+00
1.E+01
1.E+02
RML+feoB (D471Y) RML+feoB (WT)
Perc
ent s
urvi
val 1
00 μ
M H
2O2
Survival 100 μM H2O2
30 min
1 hr
Figure II.8. RML LVS can be sensitized to H2O2 by overexpression of a functional feoB. RML LVS was transformed with pBB103 containing either feoB allele being expressed from the groE promoter. H2O2 sensitivity assays were performed as described in the materials and methods. No colonies were observed from RML + feoB (WT) after one hour H2O2 treatment at the level of detection of 102 CFU. Shown is a representative assay from two replicates.
57
1.E+00
1.E+01
1.E+02
1.E+03
Iowa Iowa ΔfeoB Iowa ΔfeoB+feoB (WT)
Iowa ΔfeoB+feoB (D471Y)
Perc
ent s
urvi
val (
norm
aliz
ed to
Iow
a)
Survival 100 μM H2O2
Figure II.9. FeoB-mediated sensitivity to H2O2 is independent of the RML LVS genetic background. A clean feoB deletion mutant was constructed in the Iowa LVS genetic background and the mutant was complemented with each allele expressed from the gro promoter. H2O2 sensitivity assays were performed as described in the materials and methods. Shown is a representative assay from two replicates.
58
0
20
40
60
80
100
120
Iowa IowaΔdyp IowaΔdyp + Pgro-dyp
IowaΔdyp + Pgro-dyp93
Perc
ent s
urvi
val (
norm
aliz
ed to
Iow
a)
Survival 100 μM H2O2
Figure II.10. Dyp peroxidase protects against H2O2. A clean deletion mutant of dyp was constructed and subsequently complemented with plasmid-borne Pgro-dyp and Pgro-dyp93 constructs. H2O2 sensitivity assays were performed as described in the materials and methods. Shown is a representative assay from two replicates.
59
1.E+00
1.E+01
1.E+02
1.E+03
Iowa RML ATCC
Fold
gro
wth
24 hour growth in A549 cells
Figure II.11. Fold growth of LVS biovars in A549 cells. Cells were infected at a multiplicity of infection of 100:1 and gentamicin protection assays were performed as described in the materials and methods. One replicate is shown.
60
1.E+00
1.E+01
1.E+02
1.E+03
Fold
gro
wth
24 hour growth A549 cells
1.E+00
1.E+01
1.E+02
RML ΔfeoB
Fold
gro
wth
24 hour growth A549 cells
Figure II.12. Contribution of feoB alleles to intracellular growth in A549 cells. Bacteria were centrifuged onto cells at 600 rpm for four minutes at a multiplicity of infection of 500:1. A) Gentamicin protection assays were performed as described in the materials and methods. Only ATCCΔfeoB and ATCCΔfeoB + Pgro-feoB (WT) were significantly different from all strains. Shown is the combined fold growth for three replicates. B) The RML LVS feoB mutant trends towards less intracellular growth than the parental strain, but the two are not statistically significant (P>0.5). Shown are combined fold growths from two replicates.
A
B
61
1.E+00
1.E+01
1.E+02
Iowa RML ATCC IowaΔfeoB IowaΔdyp
Fold
gro
wth
24 hour growth in BMM
Figure II.13. Twenty-four-hour growth of LVS biovars and mutant derivatives in murine bone marrow-derived macrophages. Bacteria were passively infected (no centrifugation) for three hours at a multiplicity of infection of 20:1. Gentamicin protection assays were performed as described in the materials and methods. Shown is the combined fold growth for three replicates. No significant differences were observed between strains.
62
Discussion
Francisella tularensis is a highly pathogenic bacterium for which no
effective licensed vaccine exists. Additionally, the extreme infectiousness of the
organism has raised the possibility of its intentional release, thus precipitating the
need for a vaccine. The Live Vaccine Strain of F. tularensis subsp. holarctica
has been used to vaccinate laboratory workers in the past, but the immunity
stimulated by it was not sufficient to protect against moderate levels of
aerosolized virulent subspecies11. Furthermore, the genetic mutations that
attenuated LVS were unknown, and it was not clear if reversion to virulence was
a possibility. While LVS is not likely to be re-licensed for human vaccination, it
has served as a standard in animal models by which to measure candidate
vaccines, and as a tool to discover correlates of immunity to F. tularensis.
Indeed, in a previous report it was found that the RML LVS biovar stimulated
protective immunity in a mouse model of vaccine against virulent F. tularensis
Schu S4 by inducing higher numbers of effector T cells196. Building on these
observations, I sought to characterize some of the genetic differences between
biovars of LVS and explore how they may affect the outcomes of mouse models
of virulence and immunity. Here I report that a difference between the RML LVS
and the Iowa or ATCC LVS biovars is in the homeostasis of ferrous iron.
Iron is one of the most abundant metals on the planet but is often found in
the environment in an insoluble oxidized state, thus bacteria have evolved
sophisticated molecular machinery to scavenge iron from their local
environment120. Bacterial pathogens must acquire iron within the host
63
environment, however they must also contend with the host immune response,
which can sequester iron and limit a pathogen’s growth204-207. F. tularensis also
requires iron for productive infection, as double mutants in LVS and F. tularensis
Schu S4 lacking both ferrous and ferric iron uptake pathways are incapable of
intracellular growth in vitro and are attenuated in murine infections94,100.
Additionally, F. tularensis induced upregulation of a number of host iron related
genes, including transferrin receptor, to increase the labile iron pool in the cytosol
of infected host cells208. Although F. tularensis requires iron for growth in a host,
it appears to be unique amongst pathogens in that the bacterium’s virulence may
depend in part on maintaining optimally low levels of iron within its cytoplasm, as
the highly virulent subsp. tularensis strains had approximately 4 to 5-fold less iron
than the moderately virulent subsp. holarctica strains122.
Consistent with these observations, I have shown that the more virulent
RML LVS biovar encodes a non-synonymous mutation in the feoB gene that
affects its ferrous iron levels. The strain grows similar to the Iowa and ATCC
LVS biovars in standard propagation media, yet is restricted for growth when the
iron concentrations are significantly reduced. This suggests that the intracellular
iron concentrations may be optimally low in all strains under normal growth, but
when extracellular iron levels in the media are significantly lowered, the RML
LVS is not as proficient at scavenging iron as the Iowa and ATCC LVS.
Numerous labs have reported that the fsl iron uptake operon genes are tightly
regulated by iron concentrations in the media, so we used a lacZ reporter the
measure the promoter activity of fslA, the first gene of the fsl operon in each
64
strain under rich and limiting iron conditions95,98,102. As with growth phenotypes,
there was no detectable difference in PfslA-lacZ activity under normal growth
conditions; however, the RML LVS had increased reporter activity relative to the
Iowa and ATCC LVS after growth in iron limiting media. This is consistent with
the strain having less intracellular ferrous iron under these conditions, as Fur
regulates this promoter; higher reporter activity is interpreted to mean that less
Fe2+ is available to bind Fur protein and mediate repression of the fslA
promoter95.
The PfslA-lacZ expression is likely specific to the feoB encoded by RML
LVS, as disruptive mutations in this gene are predicted to directly impact ferrous
iron levels and no genetic differences were detected in other iron related loci. It
was important, however, to show specificity to this allele. To achieve this, I
tested the ability of each feoB to functionally complement an E. coli ΔfeoB iron
reporter strain. The results showed that the RML LVS feoB allele did not
complement the reporter mutant, while the allele from Iowa/ATCC LVS did.
Thus, the FeoB D471Y protein did not significantly import ferrous iron in a
heterologous system. While little or no function was observed in E. coli, we
believe that the FeoB D471Y does retain minimal function in LVS, as the PfslA-
lacZ reporter activity is higher in a RML LVS ΔfeoB mutant than in the parental
RML LVS (data not shown). To our knowledge this is the first direct genetic
evidence that explains differences in iron content between F. tularensis strains,
though it is almost certainly unique to LVS and the effect that we observed is
65
likely subtler than that when comparing between subsp. tularensis and subsp.
holarctica strains.
The low ferrous iron phenotypes exhibited by subsp. tularensis have been
reported to correlate with resistance to the lethal effects of H2O2, so I sought to
characterize the relative sensitivity of each LVS biovar in vitro. Indeed, the RML
LVS exhibited nearly ten-fold more resistance than the Iowa LVS, which encodes
a functional FeoB, and one hundred-fold or more resistance than the ATCC LVS.
The latter strain is significantly more sensitive to H2O2 as it is competent for Fe2+
uptake through its functional FeoB, and because it encodes a Dyp peroxidase
lacking 31 amino acids near the C-terminal portion of the protein. The mutant
Dyp has been shown previously to increase H2O2 sensitivity, and a clean deletion
of dyp in the Iowa LVS recapitulates this phenotype127. I found that the RML LVS
could be sensitized to the H2O2 lethality by supplying the functional feoB allele
under constitutive expression on a plasmid; this effect was much less
pronounced when the allele supplied was its own. I further showed that the
increased H2O2 lethality associated with overexpression of a functional feoB was
not unique to the RML LVS genetic background, as an Iowa LVS ΔfeoB mutant
complemented with the mutant allele was more resistant to H2O2 than that
complemented with the functional allele. I think these sets of experiments are
significant, as the strains only differed by a single nucleotide in feoB, yet
exhibited nearly ten-fold differences in their sensitivities to H2O2, irrespective of
genetic background. By virtue of its lower function FeoB, the RML LVS may be
better at avoiding Fe2+ associated toxicity in the context of the host environment.
66
Somewhat surprisingly, I found that the RML and Iowa LVS ΔfeoB strains
did not exhibit the same intracellular growth defect in A549 cells as I did with the
ATCC ΔfeoB; this defect had originally been reported by Thomas-Charles, et.
al.96. Unfortunately, further analysis of this mutant showed that it had lost
capsule and LPS structures, likely through a second, unspecified mutation.
Phase variation with these structures has been reported for LVS before, so the
feoB mutation may somehow increase the probability of phase variation
occurring209,210. The mutagenesis strategy that I employed makes use of a two-
step homologous recombination process, and the mutagenesis vector is
integrated onto the chromosome during the first step. Immunoblotting of the
parental ATCC strain with the integrated vector revealed no capsule or LPS
defects, so I may have been unlucky when picking an ATCC ΔfeoB colony for
long-term storage. It is not known if the intracellular growth defect of the ATCC
LVS feoB mutant reported by Thomas-Charles, et. al. is due to the same
phenomenon as I observed, or if the differences between mutagenesis schemes
would result in different phenotypes. The mutant that I constructed lacked the
entire coding sequence of feoB, while that of Thomas-charles, et. al. only
removed 410 amino acids from the middle of the protein, which is 747 amino
acids in length. Curiously, I did observe modest recovery of intracellular growth
when the ATCC ΔfeoB was complemented on a plasmid with the low function
RML LVS feoB allele, though there was significant variability between assays.
67
CHAPTER III
CHARACTERIZATION OF BACTERIOFERRITIN IN OXIDATIVE STRESS
RESISTANCE
Introduction
A growing body of literature has shown that F. tularensis utilizes numerous
strategies to avoid activation of ROS-dependent host signaling pathways and
killing by host-generated reactive oxygen and nitrogen species125-129,156,158,211,212.
Furthermore, at least one phenotypic difference between the virulent F. tularensis
Schu S4 and the non-pathogenic F. novicida is that the latter is considerably
more sensitive to H2O2; this sensitivity is associated with activation of the AIM2
inflammasome via ROS likely derived from the mitochondria213. Given these
reports I mined a previously published transposon screen to look for genes
associated with iron homeostasis and oxidative stress resistance to assess the
role these processes play in pathogenesis and immunogenicity. This screen
identified the bacterioferritin (bfr) gene as important for F. tularensis Schu S4
growth in human monocyte-derived macrophages infected with pools of
transposon mutants56.
Bacterioferritin is a widely conserved member of the ferritin family of
proteins found in bacteria and archea, but its essentiality in iron storage and
H2O2 detoxification appears to vary cross taxa214,215. The bacterioferritin protein
exists as a spherical 24-mer, typically in the cytosol where it can sequester iron
within its hollow cavity216. While structurally and functionally similar to bacterial
ferritins, most bacterioferritins have been reported to bind heme, which is not
68
observed with the former215,216. This associated heme may be important for the
controlled release of iron via electron transfer and the activity of accessory
proteins like the bacterioferritin-associated ferredoxin to meet cellular iron needs,
although it is unknown if this is a widely conserved function of bacterioferritin
among bacteria217.
Evidence exists for some species that bacterioferritin may be upregulated
by iron starvation and can protect against H2O2 toxicity214,218-221. For example, a
Pseudomonas aeruginosa mutant lacking the bacterioferritin gene bfrA exhibited
increased sensitivity to H2O2, yet showed no significant difference in total iron
content relative to the wild type222. Interestingly, while the iron levels were
similar, it was found that the BfrA protein may serve as a specific source of iron
for the catalase enzyme, as its activity was reduced by approximately half in the
bfrA mutant. In contrast, mutation of bfr alone did not enhance sensitivity of E.
coli to H2O2, but when mutated in combination with the ferric uptake regulator fur,
did provide a synergistic effect over that of the fur mutant alone. This study also
determined that ferritin served as the main reservoir for iron storage in E. coli,
with bacterioferritin contributing minimally in this capacity. Importantly,
Mycobacterium tuberculosis mutants lacking one or more bfr genes were more
susceptible to H2O2 than wild type, grew poorly in THP-1 cells, and were
attenuated in mouse and guinea pig models of infection223,224.
Bacterioferritin has appeared in the Francisella literature as being one of
the targets of the cellular and antibody immune responses in LVS immunized
mice, however, passive protection is not conferred by transfer of anti-Bfr
69
antibodies225,226. Interestingly, a LVS sodB mutant upregulated bacterioferritin
almost 2.5-fold, suggesting that it may play a role in the oxidative stress
response of F. tularensis by sequestering free Fe2+ or by detoxifying H2O2225. It
is unknown if the F. tularensis bacterioferritin functions similarly to homologs in
other bacteria, and to my knowledge a mutant has not been reported beyond that
from Lindemann, et. al56. I report here that bacterioferritin provides protection
against H2O2 and modestly attenuates and alters the immunogenicity of a more
virulent biovar of LVS in a model of challenge with the fully virulent F. tularensis
Schu S4.
Materials and methods
Bacterial strains and growth conditions: Iowa LVS (University of Iowa) and
the RML LVS were routinely cultured on modified Mueller-Hinton (supplemented
with 1% glucose, 0.025% ferric pyrophosphate, and 2% IsoVitaleX) agar or in
broth, with 50 μg/mL of kanamycin or spectinomycin, as needed. Agar plates
were incubated at 37°C with humidity and 5% CO2 while broth cultures were
grown at 37°C, shaken at 200 rpm.
Mutant construction: Deletion of bfr (FTL_0617) was achieved by homologous
recombination using derivatives of the non-replicating plasmid pJC84. Briefly,
upstream and downstream flanking DNA was amplified via PCR, and amplicons
were cloned into the multiple cloning site of pCR2.1 (Life Technologies). A
spectinomycin resistance cassette was cloned into the AvrII site in the 3’ end of
70
each upstream PCR fragment. The upstream-spectinomycin resistance fragment
was removed by digestion with AscI, and cloned into the AscI site in the 5’ region
of the downstream PCR plasmid. The entire upstream-spectinomycin resistance-
downstream fragment was cloned into the BamHI site of the suicide plasmid
pJC84. Finally, the spectinomycin resistance cassette was removed by digestion
with AvrII and the plasmid was re-ligated. Plasmids were electroporated into
LVS (2.5 kV, 25 μF, and 600 Ω), and the bacteria were plated onto MMH agar
with 50 μg/mL kanamycin after 2-3 hours of outgrowth. Organisms were then
grown overnight in broth lacking kanamycin, and were plated onto MMH agar
with 8% sucrose for sacB-mediated counter-selection. Kanamycin sensitive
colonies were screened by colony PCR to detect deletion of the bfr gene.
Primers were designed such that they flanked the coding sequence of each gene
such that an amplicon would be produced regardless of genotype.
β-galactosidase reporter activity: To assess iron uptake-regulated gene
expression, Miller assays were performed with Iowa, RML, and ATCC LVS, as
well as the RML ∆feoB and RML ∆bfr strains carrying the bfr promoter-lacZ
fusion on the pBB103 plasmid77. RML and its feoB and bfr mutant derivatives
were also assayed for PfslA-lacZ activity to estimate relative intracellular Fe2+
levels. Bacteria were grown in Chamberlain’s defined medium (supplemented
with 50 μg/mL spectinomycin) or standard MMH broth overnight, and β-
galactosidase activity was assayed using the standard Miller assay199.
71
Hydrogen peroxide sensitivity: To measure resistance to H2O2-mediated
killing, mid- to late-log phase organisms were pelleted at 13,200 RPM for 5 min,
washed in PBS, and ~106 CFU were resuspended into 200 μL PBS with or
without 100 μM fresh hydrogen peroxide (H2O2) in a 96-well dish. The samples
were incubated at 37°C with humidity and 5% CO2 for one hour. The culture
from each well was then serially diluted, plated onto modified Mueller-Hinton agar
(with antibiotic where necessary) and the number of surviving organisms for each
strain was enumerated after 2-3 days.
In vitro infections: A549 cells were cultivated in DMEM supplemented with
10% heat-inactivated fetal bovine serum (HI-FBS). Cells were enumerated and
seeded into 24 well dishes at a density of 3x105 per mL in DMEM with 10%
L929-conditioned media, 10% heat-inactivated FBS. Multiplicity of infection
(MOI) was estimated by measuring the OD600 value of mid to late-log phase
grown bacteria, and confirmed via serial dilution onto modified Mueller-Hinton
agar and enumeration after 2-3 days’ growth at 37°C with humidity, 5% CO2.
Bacteria were allowed to associate with the cells for 3 hours, after which time the
cells were washed 3x with PBS, and incubated in DMEM supplemented with HI-
FBS and 100 μg/mL of gentamicin for one hour, cells were then washed 3x with
PBS and either lysed or incubated in antibiotic-free media for a further 20 hours.
At the 4 and 24-hour time points 0.1% saponin was added to each well. Wells
were scraped and vigorously pipetted to disrupt the cells and liberate intracellular
bacteria for serial dilution and enumeration.
72
Murine infections: 8-10 week old female C57BL/6 mice were maintained on
corncob bedding for one week prior to infection. Mice were intranasally infected
with 25 μL of various doses of F. tularensis LVS strains and mutant derivatives
that had been resuspended in PBS. Inocula were calculated by measuring the
OD600 value of mid- to late-log phase grown organisms, and were confirmed by
serial dilution onto modified Mueller-Hinton agar and enumeration after 2-3 days’
growth at 37°C with humidity, 5% CO2. Moribund animals (defined as having lost
25% of the initial body weight) were sacrificed in accordance with the protocol
approved by the University of Iowa Institutional Animal Care and Use Committee.
Results
A bacterioferritin mutant has modest reduction in PfslA-lacZ activity
Data from the second chapter of this thesis highlighted the connection
between ferrous iron homeostasis and resistance to toxicity associated with
exposure to H2O2. To further extend these findings, I mined a published
transposon screen from our lab to look for mutants that may have defects in iron
metabolism and oxidative stress resistance. One gene fit these criteria, a
putative bacterioferritin encoded at the FTT_1441 locus56. A significant
proportion of the genes identified as important for growth in human macrophages
were those with unknown function or no obvious homologs, so it is not clear if
any of these has an association with iron metabolism.
73
Bacterioferritin forms large spherical multimers that store iron via a
reaction that oxidizes Fe2+ to Fe3+, making a mutant in F. tularensis an attractive
target for this study215. I constructed a clean Δbfr mutant in the RML LVS
background and supplied the PfslA-lacZ reporter construct on a plasmid to test for
differences in iron regulated promoter activity between the mutant and parental
strains (Fig. III.2). After overnight growth in standard MMH broth, reporter activity
was approximately 15-20% lower in the RML LVS Δbfr mutant than in the
parental strain, consistent with the hypothesis that a mutant lacking the
bacterioferritin protein complex may have an increased intracellular pool of Fe2+
in standard laboratory culture conditions. Construction of a complemented
mutant is underway at the time of writing, however, the bacterioferritin gene is not
in an operon and the ~600 base pairs of sequence immediately downstream is
not annotated to encode a gene or other functional element, so it seems unlikely
that the mutation has polar effects.
Bacterioferritin promoter activity is not significantly induced in iron limiting media
Although the bfr promoter has been reported as having high basal activity,
it was of interest to see if there was differential regulation amongst the LVS
biovars227. Sequence analysis shows that there is a Fur box motif (ATAATGAT)
36 base pairs upstream of the annotated start codon, suggesting that bfr
expression may be responsive to iron levels. Differential expression of an iron
storage gene could represent another phenotypic difference between the lower
iron RML LVS and the Iowa and ATCC LVS biovars. A plasmid construct
74
containing 150 base pairs upstream of the bfr coding sequence fused to lacZ was
introduced into each LVS biovar and promoter activity was assayed in both
standard MMH and CDM with high or low iron. Unlike the PfslA-lacZ reporter, no
strong induction in low iron CDM was observed in any of the strains, suggesting
that the bacterioferritin promoter may not be regulated by intracellular iron levels
directly (Fig. III.1). Reporter activity in the RML and ATCC LVS tended to be
slightly higher than in the Iowa LVS in low iron CDM, but significance was only
reached for the RML LVS, though the minimal fold induction in low iron that was
observed in each experiment was such that it is not clear that these are
physiologically significant differences between each strain.
Interestingly, the bacterioferritin promoter showed a modest change in
activity in the RML LVS ΔfeoB and Δbfr mutant backgrounds after overnight
growth in standard MMH broth. Relative to RML LVS the ΔfeoB mutant
consistently exhibited ~20-30% increased expression from the bacterioferritin
promoter, while the Δbfr mutant had ~20% less expression. These gene
expression patterns match that seen from the PfslA-lacZ reporter in each mutant
background, indicating that there may be subtle regulation at the bacterioferritin
promoter in wild type strains that is not detectable using Miller assays under the
conditions used here.
75
Bacterioferritin protects against H2O2 but is not required for intracellular growth in
A549 cells
To assay the role of bacterioferritin in F. tularensis oxidative stress
resistance, I performed H2O2 sensitivity assays as described in the materials and
methods. As before, the RML LVS exhibited relatively high level resistance to
H2O2, while the Δbfr mutant had ~10-fold fewer viable organisms, levels similar to
the Iowa LVS (Fig. III.3). This data suggests that bacterioferritin protects against
oxidative stress and is consistent with what is known for Neisseria gonorrhea and
Mycobacterium tuberculosis. I next performed in vitro infections with RML LVS
and the Δbfr mutant in the human adenocarcinoma A549 cell line to model
intracellular growth (Fig. III.4). These experiments revealed that the Δbfr mutant
is competent to escape the phagosome and grow in the cytosol of these cells,
thus bacterioferritin is dispensable for this aspect of F. tularensis pathogenesis.
Bacterioferritin contributes to virulence and immunogenicity of RML LVS in vivo
Strains with increased sensitivity to H2O2 are known to have altered
virulence properties, so we sought to determine if this phenotype resulted in
attenuation in a mouse model of infection125,127,212. The RML LVS was previously
reported to have a relatively low LD50 (174 CFU) in C57BL6 mice; this was
confirmed in an independent experiment with two groups of mice (n = 5 per
group) infected with either 133 CFU or 1.33x103 CFU of RML LVS (Fig. III.5)196.
All mice succumbed by day 7 to infection with 1,330 CFU RML LVS. Sixty
percent survival was observed in the group infected with 133 CFU; the average
76
weight loss was approximately 23 percent of the pre-infection weights. To test if
bacterioferritin was important for virulence in vivo three groups of mice (n = 5 per
group) were intranasally infected with 215 CFU, 2.15x103, or 2.15x104 CFU of
the RML LVS Δbfr mutant, and their weight and health status was measured. All
mice that received 2x102 CFU survived (average weight loss of ~12 percent)
while forty percent survived infection with 2x103 CFU (average weight loss of ~22
percent); none survived 2x104 CFU (Fig. III.5 and III.6). Importantly, it took
>2,000 CFU Δbfr to lead to survival curves and weight loss that approximated
that seen for mice infected with only 133 CFU of RML LVS, a greater than 15-fold
difference in dose. This is consistent with the hypothesis that bacterioferritin
contributes to the fitness of F. tularensis LVS in the host environment.
It was previously shown that an H2O2 sensitive biovar, ATCC LVS,
stimulates relatively weak protection against challenge with F. tularensis Schu
S4, while the H2O2 resistant RML LVS provided considerably better protection.
To test the hypothesis that the efficacy of RML LVS immunization depended on
H2O2 resistance, we challenged surviving Δbfr-infected mice (n = 5 for 215 CFU
Δbfr, n = 2 for 2,150 CFU Δbfr) and a control group of mice that had been
intranasally infected with 121 CFU of RML LVS (n = 3) coincident with the Δbfr
infections (Fig. III.7). Approximately six weeks after the initial infection the
groups were challenged with 25 CFU F. tularensis Schu S4. All mice previously
infected with RML survived the challenge, while all mice previously infected with
either 215 CFU or 2,150 CFU Δbfr succumbed with only a slight increase in time-
77
to-death relative to naïve controls, indicating that bacterioferritin is necessary for
RML LVS to stimulate an optimal immune response.
78
0
20
40
60
80
100
120
140
160
Iowa RML ATCC
Mill
er u
nits
nor
mal
ized
to Io
wa
LVS
Pbfr-lacZ activity
1x iron0.01x iron
Figure III.1. Bacterioferritin promoter activity is similar between LVS biovars grown in Chamberlain’s defined medium with high or low iron. The Pbfr-lacZ construct was introduced to each strain and reporter activity was determined via Miller assay, as described in the materials and methods. All values were normalized to those from Iowa LVS. Shown are combined data from four replicates.
79
020406080
100120140160180200
RML RMLΔfeoB RMLΔbfr
Mill
er u
nits
nor
mal
ized
to R
ML
Promoter activity of fslA and bfr
fslA-lacZ
bfr-lacZ
Figure III.2. Bacterioferritin promoter activity in is upregulated in ∆feoB and downregulated in Δbfr in MMH broth. The PfslA-lacZ and Pbfr-lacZ constructs were introduced to each strain and reporter activity was determined via Miller assay, as described in the materials and methods. Data were normalized to the RML LVS. Shown is combined data from five experiments with RML and Δbfr, and four experiments with ∆feoB for PfslA-lacZ activity, and four experiments for Pbfr-lacZ.
80
1.E+00
1.E+01
1.E+02
1.E+03
RML RMLΔfeoB RMLΔbfr
Perc
ent s
urvi
val n
orm
aliz
ed to
RM
L
Survival 100 μM H2O2 Log
Overnight
Figure III.3. Mutants lacking bfr and feoB differ in their H2O2 sensitivity depending on phase of growth. Strains were subjected to H2O2 sensitivity assays at ~mid-log phase and after overnight growth. Data were normalized to RML. Shown is combined data from four experiments.
81
1.E+00
1.E+01
1.E+02
1.E+03
RML RMLΔbfr
Fold
gro
wth
24 hour growth A549 cells
Figure III.4. The bacterioferritin mutant is proficient for intracellular growth in vitro. Gentamicin protection assays were performed as described in the materials and methods. Shown is combined data from four experiments.
82
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post-infection
RML intranasal infection
RML 133 CFU (n = 5)
RML 1,330 CFU (n = 5)
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post-infection
RML Δbfr intranasal infection
RMLΔbfr 215 CFU (n = 5)
RMLΔbfr 2,150 CFU (n = 5)
RMLΔbfr 21,500 CFU (n = 4)
Figure III.5. RML Δbfr is modestly attenuated in murine infection via the intranasal route of inoculation. Mice were infected ten-fold dilutions of RML LVS or the Δbfr mutant that had been suspended in 25 μl sterile PBS. Health was monitored for the duration of infection and mice were humanely sacrificed if ≥25% of initial weight was lost. Data shown are from one replicate.
83
0
5
10
15
20
25
30
35
RML 133 CFU RML 1,330CFU
Δbfr 215 CFU Δbfr 2,150 CFU Δbfr 21,500 CFU
Aver
age
perc
ent w
eigh
t los
t Average lowest weight
Figure III.6. RML Δbfr-infected mice lost less weight than mice infected with the wild type RML LVS. Shown is the average lowest weight for each group of infected mice from Figure III.5. Animals were humanely sacrificed if ≥25% of initial weight was lost. Data shown is from one replicate, n = 5 mice per experimental condition.
84
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post-challenge
Schu S4 challenge - 25 CFU
121 CFU RML (n = 3)
215 CFU RMLΔbfr (n = 5)
2,150 CFU RMLΔbfr (n = 2)
Naïve (n = 10)
Figure III.7. RML requires bacterioferritin to elicit protection against an intranasal challenge of 25 CFU F. tularensis Schu S4. Mice were infected with either 121 CFU of RML LVS or 215 CFU of the bacterioferritin mutant and allowed to recover for fix weeks before challenge. The challenge inoculum was suspended in 25 μl sterile PBS and administered intranasally. Health was monitored for the duration of infection and mice were humanely sacrificed if ≥25% of initial weight was lost. Data shown is from one replicate.
85
Discussion
That Francisella tularensis has numerous enzymes for the avoidance and
detoxification of H2O2 is not surprising, given that all aerobic organisms must
avoid the damaging and potentially lethal consequences of the Fenton
reaction228. Some of the ways that bacteria have evolved to limit this toxicity is to
produce antioxidant enzymes like catalase and superoxide dismutase, and these
enzymes function similarly in F. tularensis125,211. Bacteria also avoid Fenton
chemistry by limiting the availability of redox active metals like ferrous iron by
oxidizing and storing them in complexes formed by ferritin or bacterioferritin,
usually both, though one typically predominates. Here I provide the first
description of a bacterioferritin mutant in F. tularensis and provide preliminary
data suggesting a role in virulence.
Bacterioferritin is known to oxidize ferrous iron and store the ferric oxide
mineral in the hollow cavity formed by the 24-mer sphere215. I reasoned that a F.
tularensis RML LVS mutant lacking bacterioferritin may be unable to sequester
Fe2+ as effectively as the wild type strain, and that this may be reflected in iron
related gene expression. The PfslA-lacZ reporter indeed showed a modest
decrease in expression in the Δbfr mutant, indicating that the strain may have
more Fe2+ available to interact with Fur and repress expression of the reporter.
This result is consistent with the hypothesis that F. tularensis LVS uses
bacterioferritin to alter the pools of intracellular Fe2+ levels by oxidation and
storage. The reporter expression decreased by only ~15-20%, implying that the
86
bacteria may also store iron via other mechanisms, possibly the ferritin protein; a
similar mutant analysis may prove useful.
The activity of the bacterioferritin promoter was not dramatically altered by
growth in low iron media, suggesting that the Fur box motif may not be important
under this level of iron limitation. Early proteomic studies compared the
proteomes of LVS grown in high and low iron conditions found either ~2-fold
induction of bacterioferritin in iron limiting media, or that the protein spot was only
observed in iron-starved bacteria79. Interestingly, a promoter trap library from
Zaide, et. al. found that the bacterioferritin promoter was approximately 10-fold
more active than that from the groEL promoter, itself known to be highly active in
Francisella species227. These two observations suggest that there may be more
complex regulation of bacterioferritin, possibly post-transcriptional. Francisella
encodes small RNAs in its genome, but their contribution to the physiology of
Francisella has been little studied229. Still, the RML LVS had a slight but
significant increase in bacterioferritin promoter activity relative to the Iowa LVS in
iron limiting media, and the RML LVS ∆feoB mutant had approximately 30%
higher expression than that observed in the parental strain. These data suggest
a link between FeoB activity, iron levels and the bacterioferritin promoter, though
more work is needed to flesh this relationship out. The ATCC LVS also had
somewhat increased activity, but tended to exhibit variability between replicates
and was not significantly different from either the RML or Iowa LVS biovars. This
may be similar to what is known for the oxidant-sensitive LVS superoxide
dismutase mutant, which upregulated bacterioferritin nearly 2.5-fold225. I interpret
87
this to mean that LVS may be able to upregulate bacterioferritin in response to
oxidative stress and that the ATCC LVS, being one hundred to one thousand-fold
more sensitive to H2O2 than RML LVS, may increase bacterioferritin expression
to sequester free Fe2+. Alternatively, the bacterioferritin protein complex may be
required to supply the requisite iron for antioxidant enzymes, such catalase,
which may be upregulated to respond to the increased H2O2 stress experienced
by the ATCC LVS217,230.
As the iron regulated reporter indicated that the Δbfr mutant had increased
levels of intracellular ferrous iron, I hypothesized that the strain would be more
sensitive to H2O2. I found that the Δbfr mutant had approximately 75-90%
reduction in viability relative to the parental strain, indicating that bacterioferritin
has a role in resistance to oxidative stress in F. tularensis LVS. Interestingly, in
the course of these experiments I also observed that the RML LVS ∆feoB mutant
exhibited almost complete resistance to the levels of H2O2 used in these assays.
Coupled with the observation that this mutant also had significantly increased
levels of bacterioferritin promoter activity, these data suggest that the high level
resistance of the feoB mutant may be due both to having less intracellular Fe2+
and an increase in the H2O2 protective activity of bacterioferritin. This is
consistent with the observation that the RML LVS (with low FeoB activity) had a
small but significant increase in bacterioferritin promoter activity under iron
limitation. Attempts to construct a double ∆feoB Δbfr mutant are underway at the
time of writing.
88
Surprisingly, I did not detect an intracellular growth defect from the RML
LVS Δbfr mutant, at least in A549 cells. Certainly, A549 cells are not likely to
replicate the cytosolic environment of a primary human macrophage, but the
transposon screen from which bacterioferritin was mined was based on the
decreased intracellular growth of a F. tularensis Schu S4 bacterioferritin
mutant56. Multiple possibilities exist to explain this discrepancy, and they are not
mutually exclusive. First, A549 cells are immortalized adenocarcinoma cells, not
professional phagocytes, and may not provide a sufficiently challenging growth
environment, thus the mild H2O2 sensitivity of the mutant may have negligible
effect on the mutant’s fitness. Secondly, the requirement of bacterioferritin for
efficient growth in the cytosol may be more important for the fitness of Schu S4
than it is for LVS. In this scenario LVS may encode factors that sufficiently
compensate for loss of bacterioferritin that may not be present or may not
function similarly in Schu S4. Indeed, Lee et. al. report that proteomics analysis
of LVS and a virulent F. tularensis subsp. tularensis clinical isolate exhibited
differences in bacterioferritin isoforms, suggesting subtle differences in protein
processing226. Lastly, the transposon screen in which bacterioferritin was
identified relied on infecting human macrophages with pools of transposon
mutants. It is possible that the Schu S4 bacterioferritin transposon mutant may
have been exposed to a more stressful intracellular environment if macrophages
were co-infected with one or more mutants that induced a host response that
increased bactericidal activities. Two replicates of murine BMM infection suggest
that an Iowa LVS Δbfr exhibits ~30% less fold growth than Iowa LVS (not
89
shown), though time and resources have precluded me from rigorously testing
the RML LVS Δbfr in BMMs.
In the previous chapter I showed that a major phenotypic difference
between the RML and ATCC LVS biovars is the ability to resist H2O2 mediated
killing, due in part to differences in intracellular Fe2+ pools. I reasoned that the
increased H2O2 resistance phenotype of the RML LVS was critical to the
enhanced virulence it displays relative to ATCC LVS. The Δbfr mutant allowed
me to test the virulence of a RML LVS strain with increased H2O2 sensitivity
mediated by higher intracellular Fe2+ levels. Intranasal infections of mice
confirmed that bacterioferritin was required for full virulence of the strain, with the
mutant’s LD50 estimated to be ~1,500 CFU, nearly an eight-fold increase relative
to the parental strain (<200 CFU). Additionally, the mice exhibited differences in
the severity of disease, as mice infected with 133 CFU lost considerable weight
during the course of the infection, approximately twenty-three percent on
average. In contrast, the mice infected with 215 CFU of the Δbfr mutant lost
significantly less body weight during infection (approximately twelve percent,
P=0.0004), even though the dose was sixty percent higher. Only at a ten-fold
higher dose of the Δbfr mutant did the mice lose similar amounts of weight as
those infected with the lowest dose of the parent strain. I interpret this to mean
that the increased intracellular iron pool of this mutant makes it more sensitive to
host defense mechanisms that depend, at least in part, on oxidative stress.
Additionally, the RML LVS is known to be more efficacious than the ATCC
LVS at stimulating productive immunity, so survivors from the LD50 experiment
90
were challenged with Schu S4 to test the hypothesis that RML LVS stimulates
protective immunity by a mechanism that requires resistance to H2O2. As in
Griffin, et. al., I found that mice given a sublethal infection of RML LVS all
survived challenge with 25 CFU Schu S4, while naïve controls succumbed to
infection. Consistent with my hypothesis, the RML LVS Δbfr did not stimulate
robust immunity in mice infected with this strain, as all succumbed from Schu S4
infection by day eight. These data suggest that H2O2 resistance somehow
contributes to the efficacy of the RML LVS as a vaccine against tularemia in
mice, however it is possible that other interpretations of the data are valid. For
example, the bacterioferritin protein is known to stimulate antibodies and T cell
proliferation in LVS immunized mice, however it isn’t clear that antibody
responses are effective at controlling infection with virulent F. tularensis Schu
S4225. Schu S4 is able to bind the active form of the host protease plasmin in
vitro, which can cleave Francisella specific antibodies to avoid the ensuing
bactericidal and inflammatory consequences of phagocytosis via antibody
mediated opsonization183. Furthermore, transfer of monoclonal antibodies to
Francisella bacterioferritin did not provide protection against lethal LVS
challenge, even though antibodies can to be highly effective at protecting against
this strain185,231. Taken together, this suggests that the lack of a bacterioferritin-
specific antibody response is not likely to explain the poor efficacy of the RML
LVS Δbfr at protecting against Schu S4.
It is possible that the absence of bacterioferritin and the lack of
subsequent antigen-specific T cell responses to the protein explain the poor
91
protection of mice immunized with Δbfr, rather than the increased H2O2 sensitivity
of the strain. One way to test this hypothesis would be to complement a Δbfr
mutant with a mutant bfr allele with site-directed mutations at conserved residues
important for the H2O2 protective activity. Assuming protein structure and
stability were not affected, this could help clarify interpretation of the data in Fig.
II.7. For example, if the H2O2 protective activity of bacterioferritin is required for
RML LVS induced productive immunity, then a future rationally designed vaccine
strain might be designed such that its method of attenuation preserves H2O2
resistance. Alternatively, if the antigen provided by the bacterioferritin protein is
required but its activity is dispensable, then the data from Fig. III.7 would be
strong evidence that bacterioferritin is a major T cell antigen in vivo that is
required for an effective immune response to eliminate virulent F. tularensis.
This is important, as the difference in survival between mice immunized with
RML LVS or the Δbfr mutant was significant, and to my knowledge most studies
of F. tularensis T cell antigens typically utilize purified proteins for in vitro cell
stimulation, or vaccination of mice. The latter approach is specific, but does not
fully recapitulate the effectiveness of vaccination with LVS, as immunization of
mice with multiple Francisella specific T cell antigens failed to provide significant
protection against F. tularensis Schu S4232. If the alternative hypothesis is
correct, then a future vaccine strain may be designed to have constitutively high
expression of bacterioferritin.
92
CHAPTER IV
CONCLUSIONS AND FUTURE DIRECTIONS
Francisella tularensis is a fascinating bacterium of extreme infectiousness
and virulence, yet the mechanisms governing these processes are still not well
defined. That inhalation of as few as 10 CFU is sufficient to cause a significant
and potentially life-threatening disease is quite remarkable, even among other
pathogens of similar stature. The bacterium apparently has few dedicated
virulence factors, and does not encode complex regulatory or signal transduction
systems. Even less well understood are the ways in which the bacteria prevent
the generation of robust immune responses, though it seems to be and active
process that contributes to the stealthy nature of the bacteria. This is likely due
to the combination of physical shielding of antigenic structures on the bacterium
via its capsule and the atypical LPS, as well as mechanisms that may depend on
the activity of the FPI system. Evidence is accumulating, however, that the
bacteria can actively influence host signaling systems to inhibit the production of
certain inflammatory cytokines, or to prevent antigen presentation to
lymphocytes233,234.
Although the Live Vaccine Strain of F. tularensis subsp. holarctica has
been used to vaccinate humans in the past, the strain retains some of the
mechanisms of virulence and immunosuppression. The bacterial factors
underlying these mechanisms may contribute to the inability of the LVS to induce
sterilizing immunity against fully virulent F. tularensis in a host, especially at
moderate to high dose challenge. It should be noted, though, that LVS is still
93
among the most effective vaccines in mouse models for the study of immunity to
tularemia, and it represents a “gold standard” that future potential vaccines must
surpass, both in terms of efficacy but also safety.
This thesis has its roots in a desire to identify mutant strains of F.
tularensis that are attenuated and stimulate protective immunity in a mouse
model of immunization and disease. Our lab had previously identified and
described a number of mutants that lacked capsule and had truncated LPS
structures, and found that they stimulate a response that triggers host cell death
and significant inflammation in vivo, thus we reasoned that they may elicit a
productive immune response. These strain induced significant host tissue
damage at high inoculum, ~106 CFU, indicating that lower immunizing doses
would be needed. This is consistent with other immunization paradigms in the
Francisella field, and models of T cell development in environments with high or
low levels of inflammation235,236. Accordingly, I infected mice with two doses of
either waaY::trgtn or wbtA::Tn5 mutants or one dose of Iowa LVS and
subsequently challenged the mice with two doses of virulent F. tularensis Schu
S4. Surprisingly, the LVS immunized mice survived both doses of Schu S4
challenge significantly better than mice immunized with low doses of either
waaY::trgtn or wbtA::Tn5. Mice immunized with moderate doses of the mutants
had survival rates approaching that of LVS immunized mice, though they did not
surpass them. It should be reiterated that the waaY::trgtn or wbtA::Tn5 mutants
stimulate significant inflammation and tissue damage, so future evaluation of the
94
ability of these mutants will need to account for that possibility when deciding on
the dose and route of inoculation.
As LVS was more effective at protecting mice against virulent F. tularensis
challenge than attenuated F. tularensis mutants, I began to focus my attention on
what attributes of LVS might be important for immunity. Through collaboration
with the Bosio lab at Rocky Mountain Laboratories I obtained two phenotypically
different biovars of LVS (RML and ATCC) as well as their genomes. The RML
LVS was reported to have a 50-fold lower LD50 than the ATCC LVS, yet it
elicited significantly better protection against subsequent challenge with F.
tularensis Schu S4. These biovars differ very little at the genetic level, though
two important differences exist. The first and most obvious was a 93 base pair
deletion in FTL_1773 of the ATCC LVS genome, which encodes a Dyp-type
heme peroxidase. This deletion had been reported before, and while I pursued a
genetic approach to characterize this locus another group published phenotypic
data on a dyp mutant using a similar approach127. The second genetic difference
between the LVS biovars was a non-synonymous substitution in the feoB
(FTL_0133) gene in the RML LVS genome; this changed the aspartate at residue
471 to a tyrosine. As it is known that virulent subspecies of F. tularensis have
lower iron levels and higher resistance to H2O2 than less virulent subspecies, I
hypothesized that the RML LVS was more virulent and a better vaccine strain
than the ATCC LVS as a consequence of it having lower levels of intracellular
iron and a fully functional Dyp-type peroxidase.
95
I tested this hypothesis in a number of ways, including growth under
conditions where iron was limiting or in excess. I found that decreasing the iron
levels in MMH agar lead to growth restriction in all three of the LVS biovars
tested, including the LVS that had been used in our lab. The RML LVS exhibited
a much greater inability to grow under severe iron limitation than did either the
Iowa or ATCC LVS strains. This growth defect was also recapitulated in disc
diffusion assays, where the RML LVS grew in considerably smaller haloes
surrounding sterile discs spotted with iron solution. Interestingly, the RML LVS
does not exhibit iron limitation growth phenotypes in broth, suggesting that the
diffusion of iron and siderophores is sufficient for the strain to grow in this
condition. To date, iron-related growth defects reported in the literature have
been comparisons of mutants to wild type or between F. tularensis subspecies,
making the genetic data included here the first report of variable iron-related
growth phenotypes within nominally wild type bacteria from the same
subspecies96,103,122.
I extended this by showing that a lacZ reporter fusion to the promoter of
fslA was significantly more active in the RML LVS than in the other biovars when
the bacteria were grown in iron limiting media. The fslA gene is the first in the
siderophore biosynthesis and uptake operon, and is regulated by the Fur
transcriptional repressor97. The higher expression level in RML LVS is consistent
with the hypothesis that this biovar has less intracellular iron under iron limiting
conditions. Furthermore, the PfslA-lacZ reporter data indicate that the RML LVS
specifically has less ferrous iron, the reduced form of the metal that is imported in
96
a FeoB-dependent manner in other bacteria. Consistent with this, artificially
increasing the intracellular levels of Fur via overexpression rendered RML LVS
growth on MMH agar sickly compared to the vector control and the Iowa and
ATCC biovars. These data strongly implicate the RML LVS FeoB D471Y as the
causative factor for the biovar’s iron and Fur-related growth restriction and
increased fslA promoter activity.
The functionality of FeoB from each strain was compared by
complementation of an E. coli feoB mutant with a chromosomal lacZ insertion at
the iron-regulated fhuF locus. The basal β-galactosidase activity of this strain is
high due to the lack of FeoB-mediated ferrous iron uptake, though it can be
repressed by complementation with a functional feo system111.
Complementation with the Iowa/ATCC LVS feoB allele mediated significant
repression of the β-galactosidase activity while complementation with the RML
LVS feoB did not. This indicates that the low iron growth restriction and high fslA
promoter activity of RML LVS is likely due to poor ferrous iron import mediated by
FeoB D471Y. The assay does not, however, give any indication on the stability
or localization of FeoB D471Y within the heterologous host, so it is possible that
the protein may be degraded by proteases, or insert in the inner membrane
improperly. Raising an antibody to FeoB or creation of an epitope-tagged
version of FeoB D471Y could be useful for addressing these possibilities.
Though not novel for the field of bacterial iron homeostasis, these experiments
also confirmed the requirement of FeoA for Francisella FeoB function, as no
repression of the iron-regulated reporter was observed in the absence of FeoA.
97
As it is established that the Type A strains of F. tularensis have
approximately 4 to 5-fold less intracellular iron than the Type B strains, I included
the Schu S4 feoB allele in the E. coli iron reporter experiments. The encoded
protein differs from that of the annotated LVS protein by four amino acids,
suggesting that there may be functional differences that could account for the
strain’s low iron levels. Surprisingly, the Schu S4 feoB allele complemented the
E. coli reporter strain nearly as well as the Iowa/ATCC feoB allele, indicating that
the extremely low iron levels in this prototypical Type A strain are not due to a
decreased capacity of the Schu S4 FeoB to import iron, though it should be
noted that the FeoA supplied in these experiments was from F. tularensis LVS
and it differs by one amino acid. It is possible that the amino acid difference in
Schu S4 FeoA may decrease its ability to stimulate FeoB import activity.
I have also compared the promoter sequences of two known Fur-
regulated genes between Schu S4 and LVS to see if mutations may have
occurred that might decrease expression in the Type A strains. The promoters
for fslA and feoB are highly similar, though some nucleotide differences exist in
the latter. I created PfeoB-lacZ reporters for each strain’s promoter, but neither
exhibited much β-galactosidase activity so I did not pursue this further (data not
shown). Negligible color change was observed in Miller assay samples over the
course of an hour. There are multiple possible reasons for why minimal activity
was observed. First, the promoter sequences may not have contained all of the
relevant regulatory elements for transcription to occur, or may have included
elements that became inhibitory outside of their native chromosomal context due
98
to the lack of associated upstream elements. Another possibility is that the basal
promoter activity is inherently low, which would be consistent with the potentially
dangerous effects of excess FeoB-mediated ferrous iron import. Differential
expression of feoB has been reported in the form of quantitative RT-PCR, so it is
possible that Miller Assays are not sensitive enough, or there may be post-
transcriptional of feoB mRNA via sRNA or RNAse-mediated degradation.
Whatever the case, the low iron status of Type A strains is likely due to complex
factors that will require the analysis of gene expression and protein activity of
both iron uptake pathways.
The low iron Type A strains were also shown to be more resistant to H2O2
in vitro, thus I hypothesized that the RML LVS is also more resistant as a result
of its low iron levels. Ferrous iron can participate in the Fenton reaction and lead
to lipid peroxidation, DNA damage, and inhibition of essential metabolic
enzymes. Indeed, I observed that the RML LVS was as much as 10-fold more
resistant than the Iowa LVS, which encodes a functional feoB in its genome. The
RML LVS was 100 to 1,000-fold than the ATCC LVS. As mentioned above, the
ATCC LVS encodes both a functional FeoB and a truncated Dyp-type peroxidase
that lacks a conserved heme-binding residue that likely renders the truncated
enzyme less functional. I confirmed that the Dyp enzyme protects against H2O2
by constructing a dyp deletion mutant in the Iowa LVS, which encodes a full
length dyp, and testing its sensitivity. Curiously, I was unable to construct a dyp
mutant in the RML LVS, though repeated attempts were made. The biological
significance of this is not clear, as Dyp peroxidases are known to have an
99
extraordinarily broad substrate range and a potential genetic interaction with a
low function FeoB has no obvious explanation198.
The connection between the low functioning FeoB D471Y and H2O2
resistance was shown by overexpression of each feoB allele in the wild type RML
LVS and repeating the H2O2 sensitivity assays. As predicted, overexpression of
the Iowa/ATCC LVS feoB allele rendered the RML LVS significantly more
sensitive, while overexpression of the D471Y allele only modestly increased
sensitivity. This means that a single nucleotide change in feoB is responsible for
the enhanced resistance to H2O2 displayed by the RML LVS. This was not
unique to RML LVS, as an Iowa LVS ∆feoB mutant complemented with the
overexpression construct displayed similar phenotypes. These experiments also
showed that the feoB from RML LVS is not likely a null mutant, as its
overexpression did slightly increase sensitivity. Likewise, data from chapter III
shows that a ∆feoB mutant in RML LVS exhibits approximately 30-40% higher
PfslA-lacZ activity than the parent, suggesting that FeoB D471Y retains some
minimal level of function in F. tularensis LVS.
None of the mutants or overexpression strains exhibited intracellular
growth defects in vitro, save the ATCC ∆feoB mutant. As discussed in chapter II,
this strain had an undefined second site mutation that led to loss of capsule and
O-antigen polysaccharides. While it may be interesting that overexpression with
the feoB D471Y allele led to modest partial complementation, the undefined
nature of the causative mutation(s) that affected the capsule and LPS makes
interpretation of this result difficult. That the feoB mutants in RML LVS and Iowa
100
LVS did not display growth defects in in vitro infections is largely consistent with
reports in the literature, save one96. As discussed in chapter II, Thomas-Charles,
et. al. report that a F. tularensis ATCC LVS feoB mutant fails to grow in the
human adenocarcinoma A549 cell line, though I see no such defects from the
Iowa or RML LVS. If this phenotype is due to a similar capsule/LPS defect to
what I observe with an ATCC LVS ∆feoB mutant, then it would be expected that
their mutant has growth defects in more than one cell type. This is not the case,
as their mutant replicated in human monocyte-derived macrophages, which have
been shown to undergo early cell death upon infection with capsule/LPS
mutants. Given that our mutant construction protocols are different, it is possible
that their mutant (internal deletion of ~400 amino acids) and my mutant (clean
deletion of coding sequence) simply have different phenotypes.
Given that the original report of multiple biovars of LVS demonstrated
differences in virulence and vaccine efficacy phenotypes, I wanted to perform
mouse studies with strains that had different combinations of feoB and dyp
alleles complemented on the chromosomes. Unfortunately, I have been
unsuccessful in my attempts to create these strains as of the time of writing,
though not for lack of effort. Complementation on the chromosome would allow
for regulation by the native promoters and remove the problem of gene dosage
effects introduced by constitutive expression on a multi-copy plasmid.
While working on these strains, I also created a bacterioferritin mutant in
the RML LVS. Lindemann, et. al identified a F. tularensis Schu S4 bacterioferritin
transposon mutant as being less fit in the environment of human monocyte-
101
derived macrophages in vitro, suggesting a possible connection to iron, H2O2
resistance, and virulence. I confirmed that the RML LVS ∆bfr mutant had lower
PfslA-lacZ levels, suggesting that this strain had an increase in the intracellular
ferrous iron pool. The mutant was also more sensitive to H2O2, indicating that
the protein can protect F. tularensis LVS from H2O2 possibly by using H2O2 as a
cofactor to oxidize ferrous iron during the process of storing it as has been
reported for bacterioferritin-mediated iron storage mechanisms in other
organisms. Biochemical approaches to dissect Francisella bacterioferritin are
needed to describe its mechanism of action.
As discussed in chapter III, the RML LVS ∆bfr mutant did not exhibit
intracellular growth defects, suggesting that the transposon mutant identified in F.
tularensis Schu S4 may be more severely affected by loss of bacterioferritin
function, or that the way the screen through cells was performed may have
affected its viability. Interestingly, an early proteomics study of F. tularensis
reported that bacterioferritin protein expression was specific to subsp. holarctica,
but a later study by the same group found that Schu S4 expressed the protein,
but had reduced quantities of it in the membrane relative to Type B strains237,238.
The significance of this is not clear, but it suggests that bacterioferritin expression
patterns may differ in Schu S4 compared to Type B strains, or that its localization
is different. If this is the case, it is interesting to speculate that the increased
expression/localization of bacterioferritin in Type B strains may be a detriment to
the bacteria, as it is known that bacterioferritin is a potent Francisella antigen226.
This would imply that Schu S4 and other Type A strains may express
102
bacterioferritin only in environments where it is absolutely necessary for the
survival of the bacterium, such as a host cell. Further work is needed with a
defined deletion mutant of bacterioferritin in the Schu S4 background examined
outside of the context of a mutant pool in vitro infection.
Despite the lack of an intracellular growth defect, the RML LVS ∆bfr
mutant allowed me to test the hypothesis that a key component of the virulence
of RML LVS and its ability to stimulate immunity in a mouse is its resistance to
H2O2 due to increased intracellular ferrous iron. Murine infections were
consistent with this hypothesis, as the mutant was mildly attenuated relative to
the parent strain. More importantly, subsequent challenge of ∆bfr immunized
mice with F. tularensis Schu S4 showed that bacterioferritin is a necessary factor
for RML LVS to elicit protective immunity, as all challenged mice succumbed
within eight days. In comparison, one-hundred percent survival was observed in
control mice that had been immunized with wild type RML LVS.
As discussed in chapter III, it is possible that the absence of the protein
and not necessarily its H2O2 protective enzymatic activity that leads to a less than
optimal immune response. Future work with site-directed mutants at conserved
residues critical for enzyme function should be performed to examine this
possibility. However, I interpret these results to mean that the H2O2 sensitivity of
the mutant stimulates a qualitatively different host response. It is known that the
H2O2 sensitive ATCC LVS elicits less poor protection relative to the resistant
RML LVS, even though its bacterioferritin gene is intact and the protein is likely
expressed appropriately196. I suspect that the mild to moderate H2O2 sensitivity
103
of these strains leads to occasional loss of bacterial membrane integrity inside a
host cell that could lead to host surveillance systems recognizing bacterial factors
from these low level “leaky” events, such as bacterial DNA, and responding by
induction of proinflammatory cytokines. F. novicida is highly sensitive to H2O2
relative to F. tularensis and it activates the AIM2 inflammasome in murine
macrophages by a process that depends in part on H2O2 from the host’s
mitochondria213. My working model is that inflammation derived from a similar
process coupled to the as yet undefined immunosuppressive mechanisms that
inhibit production of certain proinflammatory cytokines and antigen presentation
lead to a smaller population of Francisella-specific effector memory T cells in
mice immunized with H2O2 sensitive strains of LVS.
This model is necessarily oversimplified, as the full scope of intracellular
events that are actively modulated by F. tularensis is still unknown. Additionally,
further experiments need to be performed, as the model is built on assumptions
that need testing. First, I am assuming that strains and mutants with excess
intracellular ferrous iron exhibit occasional instability in their membranes that
results in detectable loss of integrity due to Fenton reactions between the ferrous
iron and H2O2 generated from endogenous metabolism. In the cytoplasm of a
host cell there would be a contribution of H2O2 from host metabolism as well.
This loss of integrity could be tested via commonly used live/dead stain assays.
Secondly, I assume that occasional loss of integrity occurs in the cytosolic
environment of infected cells and that this correlates with activation of an
intracellular immune surveillance system and production of proinflammatory
104
cytokines. This could be addressed with confocal microscopy of macrophages
infected with strains and mutants generated in this thesis. Similar experiments
include infection of host cells with bacteria harboring a plasmid with luciferase
driven by a eukaryotic promoter; it is thought that loss of membrane integrity
leads to leakage of the plasmid from the bacteria to the host cell cytoplasm, and
that the host then transcribes and translates the luciferase239. Thus, I could
infect cells with my strains and mutants and measure relative luciferase activity.
The model predicts that high sensitivity to H2O2 would correlate with high
luciferase activity.
In support of this model, studies of how F. tularensis physiology and
pathogenicity change after growth in modified Mueller-Hinton broth (MMH) or
Brain-Heart Infusion (BHI) have linked growth in MMH to the AIM2-dependant
stimulation of proinflammatory cytokines in infected murine macrophages240,241.
Importantly, BHI-grown mutants that are deficient in oxidative stress resistance,
katG and mglA, failed to inhibit cytokine production. I mention this because I
have compared PfslA-lacZ iron reporter activity and H2O2 sensitivity in strains
grown in MMH or BHI (data not shown) and found that bacteria grown in BHI
have five to seven-fold higher iron reporter activity and negligible sensitivity to
H2O2 at levels that cause significant lethality from bacteria grown in MMH. It is
important to note that the MMH used for F. tularensis propagation incudes
0.025% ferric pyrophosphate, making it an iron-rich medium. That the iron
reporter activity after growth in BHI is so high is striking; it is comparable to that
of bacteria grown in Chamberlain’s defined medium with 350 nM FeSO4, a
105
concentration 100-fold less than typically used in this medium for standard
Francisella propagation. Taken together, this shows that wild type bacteria
grown in MMH have higher intracellular iron levels, increased instances of
membrane instability as shown by live/dead stain of broth grown organisms, and
they stimulate significantly high levels of proinflammatory cytokines from infected
murine macrophages via AIM2. Interestingly, the authors report that BHI is a low
iron media, but they do not pursue the role of high iron concentrations in MMH in
mediating membrane instability or induction of cytokines. My model predicts that
strains with excess intracellular ferrous iron may stimulate an inflammatory
response that likely leads to clearance of the bacteria and polarization of naïve T
cells to a short-lived effector state, rather than longer-lived memory state.
The ultimate goal of this thesis was to both identify the bacterial factors
responsible for the variability between F. tularensis biovars with respect to their
virulence and to characterize the host responses both in vitro and in vivo due to
these factors. Unfortunately, I was not able to accomplish this; however, the low
and moderate dose immunization experiments and the IglC-LCMV epitope fusion
tools that I developed in appendix A can clearly be used to supplement the
findings in this thesis. For example, the IglC-LCMV epitope fusion could be used
to characterize antigen-specific T cell populations in mice immunized with strains
generated in this thesis. These tools could test the hypothesis that H2O2
sensitive strains weakly stimulate proliferation of antigen-specific T cells, and that
resistant strains are more effective at this process. These tools could also be
used in conjunction with mouse lines that lack components of intracellular
106
surveillance mechanisms, as well as mice that lack various inflammatory
cytokines to characterize the contribution of each in mediating the increased or
decreased vaccine efficacy of each strain. In the broader context, I had hoped
that this work and the tools built herein could help clarify the some of the
requirements of a live vaccine strain to elicit immunity against infection with F.
tularensis so that a future rationally designed vaccine strain could incorporate
and build upon the strengths and avoid the weaknesses of F. tularensis LVS.
107
Figure IV.I Model of iron import in F. tularensis subsp. holarctica Live Vaccine Strain (LVS). The RML LVS feoB allele has a nucleotide substitution that changes the aspartate at position 471 to a tyrosine; this change decreases the ferrous iron import activity of FeoB in the RML LVS, which correlates with enhanced resistance to H2O2. Once in the cytosol, ferrous iron can likely be oxidized by and stored within the central cavity of the bacterioferritin complex.
108
Figure IV.II Model of intracellular outcomes for high and low iron strains. RML LVS (left half of cell model) is more resistant to the bactericidal activity of IFNγ stimulated murine macrophages due in part to having less intracellular ferrous iron content. Strains that have excess ferrous iron, such as ATCC, are more sensitive to such bactericidal activities.
109
APPENDIX A
DEVELOPMENT OF RECOMBINANT IGLC-LCMV EPITOPE T CELL
REPORTER FUSIONS
Rationale
While developing a mouse model of low and moderate dose immunization
with F. tularensis Schu S4 mutants with capsule and LPS defects, I sought to
create tools to facilitate the study of T cell responses of mice immunized within
this model. To do this I made plasmids that carried the Francisella iglC gene with
the coding sequence of either of two well-known epitopes (GP33-41 and GP61-80)
from the lymphocytic choriomeningitis virus, or LCMV242. IglC was chosen for C-
terminal fusion of the epitopes because it is among the most highly induced
genes within the macrophage environment and the protein is among those that
are T cell antigens in LVS immunized mice243,244.
The rationale for this approach was to use these tools with mouse lines
that are T cell receptor restricted to these specific epitopes to characterize T cell
proliferation and maturation after immunization with a F. tularensis strain, as well
as use the strength of mouse genetics to identify host factors that are necessary
for generation of a T cell response to F. tularensis. Additionally, as the immune
responses to these epitopes are well described, their use in F. tularensis during
vaccination may also help identify mechanisms of immune suppression that
interfere with the generation of a robust and long-lived cellular immune response,
and/or be useful in identifying correlates of immunity.
110
Materials and methods
Bacterial strains and growth conditions: Iowa LVS (University of Iowa) and F.
tularensis Schu S4 and were routinely cultured on modified Mueller-Hinton
(supplemented with 1% glucose, 0.025% ferric pyrophosphate, and 2%
IsoVitaleX) agar or in broth, with 50 μg/mL of spectinomycin, as needed. Agar
plates were incubated at 37°C with humidity and 5% CO2 while broth cultures
were grown at 37°C, shaken at 200 rpm. Plasmids were electroporated into LVS
(2.5 kV, 25 μF, and 600 Ω), and the bacteria were plated onto MMH agar with 50
μg/mL spectinomycin after 2-3 hours of outgrowth.
Low and moderate dose mouse immunization and Schu S4 challenge: 6-8
week old female BALB/c mice were inoculated via the intranasal with either the
~500 CFU Iowa LVS, ~85 or 850 CFU F. tularensis waaY::Trgtn, or ~22 or 107
CFU wbtA::Tn5 that had been suspended in sterile PBS. Inoculum dose from
each strain was serially diluted and plated for enumeration. Mice were monitored
for signs of distress daily. Six weeks after immunization the mice were
challenged with either ~66 or 660 CFU of F. tularensis Schu S4 and monitored
for survival.
Recombinant IglC-LCMV epitope construction: The coding sequence of iglC
was PCR amplified and subsequently used as template in splice-overlap-
extension PCR (SOE-PCR) with large primers that included the 3’ end of iglC
and GP33-41 or GP61-80 that had been codon-optimized for the Francisella genome
111
(http://www.jcat.de). All PCRs used the high-fidelity Phusion polyermase per the
manufacturer’s recommendations. The 5’ primers included a KpnI site and the
final 3’ primers included a SalI site. The SOE-PCR product was A-tailed using
Taq polyermase and cloned into pCR2.1. Inserted sequences were excised by
digestion with KpnI and SalI, then ligated into the respective sites of pTrc99a,
immediately downstream of the iglA promoter that had been cloned previously.
Similar plasmids were created with the groE promoter. The promoter-iglC
fragments were removed by digestion with BamHI and SalI and cloned into the
respective sites of the final vector, pBB103.
In vitro infections: HepG2 cells were cultivated in DMEM supplemented with
10% heat-inactivated fetal bovine serum (HI-FBS). Cells were enumerated and
seeded into 24 well dishes at a density of 3x105 per mL in DMEM with 10%
L929-conditioned media, 10% heat-inactivated FBS. Multiplicity of infection
(MOI) was estimated by measuring the OD600 value of mid to late-log phase
grown bacteria, and confirmed via serial dilution onto modified Mueller-Hinton
agar and enumeration after 2-3 days’ growth at 37°C with humidity, 5% CO2.
Bacteria were allowed to associate with the cells for 3 hours, after which time the
cells were washed 3x with PBS, and incubated in DMEM supplemented with HI-
FBS and 100 μg/mL of gentamicin for one hour, cells were then washed 3x with
PBS and either lysed or incubated in antibiotic-free media for a further 20 hours.
At the 4 and 24-hour time points 0.1% saponin was added to each well. Wells
112
were scraped and vigorously pipetted to disrupt the cells and liberate intracellular
bacteria for serial dilution and enumeration.
T cell response to antigen ex vivo: Briefly, mice were infected intraperitoneally
with ~500 CFU of the Iowa LVS harboring either of the plasmids encoding the
IglC-LCMV epitope fusion and splenic lymphocytes were incubated with antigen
peptides, and the proportion of cells responding, as measured by interferon γ
production, were quantified via FACS protocol similar to Badovinac et. al245.
LCMV immunization and epitope-expressing LVS challenge: 6-8 week old
female BALB/c mice were inoculated intraperitoneally with 2.5x105 PFU LCMV.
Six weeks after inoculation, these mice and unimmunized controls were infected
via the intranasal route with either Iowa LVS carrying the control plasmid
expressing iglC alone, iglC-GP33-41, or iglC-GP61-80. Weight loss was monitored
daily and mice were humanely sacrificed if ≥25% of initial weight was lost.
Results and discussion
The low and moderate dose immunization experiment included Iowa LVS
as a control and a benchmark against which to compare the protection afforded
mice by intranasal inoculation with two F. tularensis Schu S4 mutants with
defects in capsule and LPS that our lab had previously characterized as
attenuated. Each group of mice was challenged with two doses of F. tularensis
113
Schu S4 and the survival of each group was compared to that of the LVS
immunized group (Fig. A.1 and A.2). From the group immunized with low dose
waaY::trgtn or wbtA::Tn5, no mice survived challenge with F. tularensis Schu S4,
but 80% of the LVS immunized group survived 66 CFU, though only an increase
in time to death was observed when challenged with 660 CFU. Mice immunized
with higher doses of the mutants survived the 66 CFU challenge better, with 20%
and 60% survival in mice from the waaY::trgtn and wbtA::Tn5 groups,
respectively, though still not as well as the LVS immunized mice. Survival of the
moderate immunization dose group after challenge with 660 CFU was roughly
comparable to that of the control LVS group; in fact one wbtA::Tn5 immunized
mouse survived the challenge for the duration of the experiment, though it should
be noted that this individual was considerably larger than its littermates.
To begin to characterize the differences in T cell responses between LVS
and the F. tularensis Schu S4 capsule/O-antigen mutants, I built IglC-LCMV
epitope fusion constructs. Fusion of the LCMV epitopes to IglC resulted in
detectable upward shifts of the IglC signal in Western blots from E. coli, LVS, and
the LVS fevR::trgtn mutant, indicating that the fusion constructs were being
expressed specifically from the groE promoter; no signal would be predicted from
E. coli lacking the plasmids and very low signal is expected from the fevR
mutant. The signal strength appears to be weaker from proteins with the
epitopes, indicating that they may be inherently less stable or are made targets of
proteases with the additional amino acids. Importantly, the presence of the
protein fusions did not appear to alter the intracellular growth kinetics of LVS in
114
HepG2 cells, suggesting that if the fusion proteins do negatively impact the FPI
function within the bacteria, it is not a dominant phenotype.
Mice infected with LVS expressing the fusion proteins generate a low level
response to IglC-GP33-41, as treatment of their lymphocytes ex vivo with purified
GP33-41 peptide stimulated only a four-fold increase in the number of interferon γ
positive cells relative to untreated cells. No GP61-80 response was detected in this
assay, suggesting that the decreased stability of the fusion protein may reduce
the protein levels to below a threshold where enough antigen is presented to be
detectable with this approach. It is also possible that the process of antigen
presentation is decreased or inhibited in vivo by F. tularensis as it is in vitro190.
To test the hypothesis that antigen-specific T cells can contribute to an
immune response to LVS expressing these epitopes, mice were given a sub-
lethal dose of LCMV that is known to stimulate cellular immunity. These mice
were challenged with the LVS strains approximately six weeks later and
monitored for weight loss and survival. Mice infected with the control LVS
expressing native IglC exhibited the most mortality, with only one mouse
surviving the intranasal challenge with ~6,000 CFU. In both naïve control mice
and LCMV-immunized mice there was a decrease in mortality when challenged
with both GP33-41 and GP61-80 expressing strains, indicating that the epitopes alter
the virulence of the bacteria, possibly as a result of the strength of the response
to the antigenic epitopes. It will be of interest to describe the events that occur in
naïve mice in response to these strains, as they may give insight into some of the
115
host response pathways that may be engaged for the control of intracellular
Francisella, and thus may aid in identifying correlates of immunity.
Ultimately, these tools were built in an effort to dissect the host response
to attenuated strains of F. tularensis that could contribute towards a vaccine
against tularemia. These tools were intended to be integrated into the high and
moderate dose immunization experimental paradigm to compare the T cell
responses to F. tularensis in environments with differing levels of inflammation,
which is can affect the developmental fate of naïve T cells246.
116
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post challenge
Low CFU immunization - 66 CFU Schu Challenge
Naïve - 19 CFU Schu (n=5)
LVS 500 CFU - 66 CFU Schu (n=5)
85 CFU waaY::Trgtn - 66 CFU Schu(n=5)22 CFU wbtA::Tn5 - 66 CFU Schu (n=5)
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post challenge
Low CFU immunization - 660 CFU Schu Challenge
Naïve - 19 CFU Schu (n=5)
LVS 500 CFU - 660 CFU Schu (n=5)
85 CFU waaY::Trgtn - 660 CFU Schu (n=5)
22 CFU wbtA::Tn5 - 660 CFU Schu (n=5)
Figure A.I. Low dose immunization survival curves. Groups of mice were inoculated via the intranasal route with LVS or two different F. tularensis Schu S4 capsule/O-antigen mutants six weeks prior to challenge with two doses of wild type Schu S4, as per the materials and methods. All data are from one replicate.
117
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14Perc
ent s
urvi
val
Days post challenge
High CFU immunization - 66 CFU Schu Challenge
Naïve - 19 CFU Schu (n=5)
LVS 500 CFU - 66 CFUSchu850 CFU waaY::Trgtn - 66CFU Schu (n=5)107 CFU wbtA::Tn5 - 66CFU Schu (n=5)
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Perc
ent s
urvi
val
Days post challenge
High CFU immunization - High Schu Challenge
Naïve - 19 CFU Schu (n=5)
LVS 500 CFU - 660 CFU Schu
850 CFU waaY::Trgtn - 660 CFUSchu (n=4)107 CFU wbtA::Tn5 - 660 CFU Schu(n=4)
Figure A.II. High dose immunization survival curves. Groups of mice were inoculated via the intranasal route with LVS or two different F. tularensis Schu S4 capsule/O-antigen mutants six weeks prior to challenge with two doses of wild type Schu S4, as per the materials and methods. All data are from one replicate.
118
1
100
LVS LVS C LVS C33 LVS C61
Fold
gro
wth
24 hour growth HepG2 cells
LVS
E. c
oli +
em
pty
vect
or
E. c
oli +
Pgr
o-ig
lC
E. c
oli +
Pgr
o-ig
lC-G
P33
E. c
oli +
Pgr
o-ig
lC-G
P61
LVS
LVS
fevR
::Trg
Tn
LVS
fevR
::Trg
Tn +
Pgr
o-ig
lC
LVS
fevR
::Trg
Tn +
Pgr
o-ig
lC-G
P33
LVS
fevR
::Trg
Tn +
Pgr
o-ig
lC-G
P61
LVS
+ P
gro-
iglC
LVS
+ P
gro-
iglC
-GP
33
LVS
+ P
gro-
iglC
-GP
61
Figure A.III. IglC-LCMV epitope fusions are expressed in LVS and do not alter intracellular growth. A) Plasmids with the coding sequence for expression of the IglC fusion proteins were transformed into E. coli, LVS, and the LVS fevR mutant. Lysates of each strain were subjected to Western blot and probed with the anti-IglC mAb. B) HepG2 cells were infected in vitro with either LVS or LVS carrying one of the recombinant IglC plasmids. LVS C = LVS expressing IglC with no epitope; LVS C33 = LVS expressing IglC-GP33-41; and LVS C61 = LVS expressing IglC-GP61-80. All data shown are from one replicate.
A
B
119
CD4+ CD4+
IFNγ
IFNγ
CD8+ CD8+
Figure A.IV. LVS infected mouse lymphocyte response to LCMV antigens ex vivo. Naïve mice were infected via the intraperitoneal route with ~500 CFU of LVS expressing either IglC-GP33-41 or IglC-GP61-81. Splenic lymphocyte responses to purified antigens were measured seven days later, as described in the materials and methods. Daa are from one replicate.
120
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12
Perc
ent s
urvi
val
Days post-infection
Naïve mice
LVS iglC (n=5)
LVS iglC-GP33 (n=4)
LVS iglC-GP61 (n=5)
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12
Perc
ent s
urvi
val
Days post infection
LCMV immunized
LCMV + LVS iglC (n=4)
LCMV + LVS iglC-GP33 (n=5)
LCMV + LVS iglC-GP61 (n=5)
Figure A.V. Survival of LCMV immunized mice challenged with LVS expressing LCMV epitopes. Mice were given 2.5x105 PFU of LCMV intraperitoneally and allowed to convalesce for six weeks, after which they were challenged by the intranasal route with ~6,000 CFU of LVS with or without epitopes fused to IglC, as described in the materials and methods. Data are from one replicate.
121
APPENDIX B
OXYR DOES NOT INFLUENCE OXIDATIVE STRESS RESISTANCE IN F.
TULARENSIS LVS
Rationale
Analysis of the genomes of Francisella indicates that a well conserved
homolog of the LysR transcription factor OxyR can be found. OxyR is known to
be an allosteric regulator of genes in the oxidative stress resistance pathways of
many bacteria, changing its DNA binding ability upon oxidation of cysteine
residues by H2O2 and activating transcription of protective genes like catalase
and the alkylhydroperoxide reductase247-250. Additionally, a transposon screen of
F. novicida identified oxyR as a mutant with increased sensitivity to H2O2 and
decreased fitness in a Drosophila model of infection. For these reasons I
hypothesized that the F. tularensis homolog of OxyR may be important for
upregulation of H2O2 protective factors, thus I created a mutant in F. tularensis
LVS and tested its role in resistance to H2O2.
Materials and methods
Bacterial strains and growth conditions: RML LVS was routinely cultured on
modified Mueller-Hinton (supplemented with 1% glucose, 0.025% ferric
pyrophosphate, and 2% IsoVitaleX) agar or in broth, with 25 or 50 μg/mL of
kanamycin or spectinomycin as needed, respectively. Agar plates were
122
incubated at 37°C with humidity and 5% CO2 while broth cultures were grown at
37°C, shaken at 200 rpm.
Mutant construction: Deletion of oxyR (FTL_1014) was achieved by
homologous recombination using derivatives of the non-replicating plasmid
pJC84. Briefly, upstream and downstream flanking DNA was amplified via PCR,
and amplicons were cloned into the multiple cloning site of pCR2.1 (Life
Technologies). A spectinomycin resistance cassette was cloned into the AvrII
site in the 3’ end of each upstream PCR fragment. The upstream-spectinomycin
resistance fragment was removed by digestion with AscI, and cloned into the
AscI site in the 5’ region of the downstream PCR plasmid. The entire upstream-
spectinomycin resistance-downstream fragment was cloned into the BamHI site
of the suicide plasmid pJC84. In contrast to the mutants made in chapters II and
II, the spectinomycin resistance cassette was not removed. Plasmids were
electroporated into LVS (2.5 kV, 25 μF, and 600 Ω), and the bacteria were plated
onto MMH agar with 50 μg/mL kanamycin after 2-3 hours of outgrowth.
Organisms were then grown overnight in broth lacking kanamycin, and were
plated onto MMH agar with 8% sucrose for sacB-mediated counter-selection.
Kanamycin sensitive colonies were screened by colony PCR to detect deletion of
the oxyR gene.
Hydrogen peroxide sensitivity: To measure resistance to H2O2-mediated
killing, mid- to late-log phase organisms were pelleted at 13,200 RPM for 5 min,
123
washed in PBS, and ~106 CFU were resuspended into 200 μL PBS with or
without 100 μM fresh hydrogen peroxide (H2O2) in a 96-well dish. The samples
were incubated at 37°C with humidity and 5% CO2 for one hour. The culture
from each well was then serially diluted, plated onto modified Mueller-Hinton agar
(with antibiotic where necessary) and the number of surviving organisms for each
strain was enumerated after 2-3 days. This assay was also modified in two
ways to test for the production and/or secretion of an H2O2 protective factor into
the culture media. First, a standard quantity of bacteria were added to fresh
media and allowed to outgrow for one hour, followed by addition of 500 µM H2O2
directly to the culture, from which samples were removed and serially diluted
onto MMH agar every hour for five hours; and second, the ∆oxyR mutant was
suspended in spent media from wild type RML LVS or itself with or without 1 mM
H2O2 for 30 minutes before serial dilution onto MMH agar.
In vitro infections: J774.A16 cells were cultivated in DMEM supplemented with
10% heat-inactivated fetal bovine serum (HI-FBS). Cells were enumerated and
seeded into 24 well dishes at a density of 3x105 per mL in DMEM with 10%
L929-conditioned media, 10% heat-inactivated FBS. Multiplicity of infection
(MOI) was estimated by measuring the OD600 value of mid to late-log phase
grown bacteria, and confirmed via serial dilution onto modified Mueller-Hinton
agar and enumeration after 2-3 days’ growth at 37°C with humidity, 5% CO2.
Bacteria were allowed to associate with the cells for 3 hours, after which time the
cells were washed 3x with PBS, and incubated in DMEM supplemented with HI-
124
FBS and 100 μg/mL of gentamicin for one hour, cells were then washed 3x with
PBS and either lysed or incubated in antibiotic-free media for a further 20 hours.
At the 4 and 24-hour time points 0.1% saponin was added to each well. Wells
were scraped and vigorously pipetted to disrupt the cells and liberate intracellular
bacteria for serial dilution and enumeration.
Results and discussion
The F. tularensis LVS OxyR protein shares 36% identity to that of E. coli
K12 and 35% identity to its homolog in Salmonella (via BLASt analysis),
significantly high values for such a phylogenetically distant organism. Such
conservation and the clear link of OxyR to oxidative stress resistance in many
bacteria, including other pathogens, made the F. tularensis LVS OxyR an
attractive target for further study. Furthermore, the oxyR gene is in a divergent
genomic orientation with the alkylhydroperoxide reductase gene ahpC, consistent
with the orientation of other LysR family genes, suggesting that this locus may
represent a functional unit in Francisella251. The mutant exhibits no obvious
growth defects in broth culture, though the colony size on an agar plate is smaller
than that of RML LVS; the significance of this is not clear. Quite surprisingly, the
∆oxyR mutant did not display increased H2O2 sensitivity in any of the three
assays that were used, though it did exhibit a very modest intracellular growth
defect in J774A.16 cells; its twenty-four hour fold growth was approximately 30%
that of the parental RML LVS.
125
Multiple possibilities exist to explain why the ∆oxyR mutant was not more
sensitive to H2O2, though only further experimentation can confirm or rule them
out. The first possibility is that OxyR may be a molecular fossil and no longer
regulates genes involved in oxidative stress resistance. This could occur by
mutations that affect the ability of OxyR to be oxidized by H2O2, mutations that
affect its ability to bind DNA, or by mutations within the promoters of oxidative
stress resistance genes. This would mean that the genes previously regulated
by OxyR may be constitutively expressed, possibly to the bacteria’s benefit.
The second possibility is that the F. tularensis LVS OxyR responds to a
different stress than H2O2, such as superoxide or reactive nitrogen species.
Indeed, Binesse, et. al. report that while a F. tularensis Schu S4 ahpC mutant
was not more sensitive to H2O2, it did have increased sensitivity to paraquat and
the compound SIN-1, which ultimately generates peroxynitrite127. If the oxyR-
ahpC divergent gene locus is a functional LysR-like regulatory unit, then it is
possible that the lack of H2O2 sensitivity displayed by the F. tularensis LVS
∆oxyR mutant would be consistent with the data from Binesse, et. al. Preliminary
experiments with this mutant and SIN-1 are inconclusive and need optimization
as of the time of writing.
A third possibility is redundancy in transcription factors that regulate H2O2
protective genes. Many bacteria have multiple pathways that can overlap in
terms of the genes they activate or repress, including those for oxidative stress
resistance252. For example, the FTL_1634 ORF encodes a LysR-like
transcription factor that shares 42% identity with OxyR at the amino acid level. If
126
this ORF is transcribed and encodes a functional homolog of OxyR, then it could
compensate and mask H2O2 associated phenotypes in the ∆oxyR mutant. There
are other, less conserved LysR-like genes encoded elsewhere on the
chromosome that could compensate for loss of OxyR, as well.
127
020406080
100120
10 uM 50 uM 100 uM
Perc
ent s
urvi
val
H2O2 in PBS
RML
ΔoxyR
0.001
0.01
0.1
1
1 2 3 4 5
Perc
ent s
urvi
val
Hours post H2O2
500 uM H2O2 in broth
RML
ΔoxyR
0.1
1
10
MMH RCM OCM
Perc
ent s
urvi
val
1 mM H2O2 in conditioned media
A
B
C
Figure B.I. The RML LVS ∆oxyR mutant is as resistant to H2O2 as the wild type. A) The bacteria were exposed to three doses of H2O2 in PBS for one hour. B) Bacteria were sub-cultured into fresh media for one hour prior to addition of 500 µM H2O2. Samples were taken at one hour intervals. C) ∆oxyR was exposed to 1 mM H2O2 for thirty minutes in either fresh MMH broth or filtered spent media from the wild type RML LVS or the ∆oxyR mutant. MMH = modified Mueller-Hinton; RCM = RML LVS conditioned media; OCM = ∆oxyR conditioned media.
128
0
20
40
60
80
100
120
RML ΔoxyR
Fold
gro
wth
nor
mal
ized
to
RM
L 24 hour growth in J774.A16 cells
Figure B.II. The ΔoxyR mutant has a mild intracellular growth defect in J774A.16 cells. Infections were performed as described in the materials and methods. Combined fold growths are shown from four independent experiments.
129
REFERENCES 1 Tusnady, G. E. & Simon, I. The HMMTOP transmembrane topology
prediction server. Bioinformatics 17, 849-850 (2001). 2 Spyropoulos, I. C., Liakopoulos, T. D., Bagos, P. G. & Hamodrakas, S. J.
TMRPres2D: high quality visual representation of transmembrane protein models. Bioinformatics 20, 3258-3260, doi:10.1093/bioinformatics/bth358 (2004).
3 McCoy, G. W. & Chapin, C. W. Studies of plague, a plague-like disease,
and tuberculosis among rodents in California. I. A study of the virulence of cultures of Bacillus pestis of various sources and ages. II. The virulence of Bacillus pestis of ground squirrel origin. III. Immunity of wild rats (Mus norvegicus) to plague infection. IV. Susceptibility of a ground squirrel (Ammospermophilus leucurus, Merriam) to plague. V. Bacterium tularense the cause of a plague-like disease of rodents. VI. A note on the susceptibility of ground squirrels (Citellus beechyi, richardson) to tuberculosis. (Govt. print. off., 1912).
4 Wherry, W. B. & Lamb, B. H. Infection of man with Bacterium tularense.
1914. J Infect Dis 189, 1321-1329 (2004). 5 Sjostedt, A. Tularemia: history, epidemiology, pathogen physiology, and
clinical manifestations. Ann N Y Acad Sci 1105, 1-29, doi:10.1196/annals.1409.009 (2007).
6 Chapin, C. W. Review of Our Knowledge of Bacterium Tularense.
American journal of public health 11, 529-532 (1921). 7 Francis, E. Landmark article April 25, 1925: Tularemia. By Edward
Francis. Jama 250, 3216-3224 (1983). 8 Francis, E. & Lake, G. C. Transmission of Tularaemia by the Mouse Louse
Polyplax Serratus (Burm.). Public health reports 37, 96-101 (1922). 9 Francis, E. & Lake, G. C. Transmission of Tularaemia by the Bedbug,
Cimex Lectularius. Public health reports 37, 83-95, doi:Doi 10.2307/4576251 (1922).
10 McLendon, M. K., Apicella, M. A. & Allen, L. A. Francisella tularensis:
taxonomy, genetics, and Immunopathogenesis of a potential agent of biowarfare. Annu Rev Microbiol 60, 167-185, doi:10.1146/annurev.micro.60.080805.142126 (2006).
130
11 Saslaw, S., Eigelsbach, H. T., Prior, J. A., Wilson, H. E. & Carhart, S. Tularemia vaccine study. II. Respiratory challenge. Archives of internal medicine 107, 702-714 (1961).
12 Saslaw, S., Eigelsbach, H. T., Wilson, H. E., Prior, J. A. & Carhart, S.
Tularemia vaccine study. I. Intracutaneous challenge. Archives of internal medicine 107, 689-701 (1961).
13 Harris, S. Japanese biological warfare research on humans: a case study
of microbiology and ethics. Ann N Y Acad Sci 666, 21-52 (1992). 14 Christopher, G. W., Cieslak, T. J., Pavlin, J. A. & Eitzen, E. M., Jr.
Biological warfare. A historical perspective. Jama 278, 412-417 (1997). 15 World Health Organization. Health aspects of chemical and biological
weapons. Report of a WHO Group of Consultants. (1970). 16 Kaufmann, A. F., Meltzer, M. I. & Schmid, G. P. The economic impact of a
bioterrorist attack: are prevention and postattack intervention programs justifiable? Emerg Infect Dis 3, 83-94, doi:10.3201/eid0302.970201 (1997).
17 Dennis, D. T. et al. Tularemia as a biological weapon: medical and public
health management. Jama 285, 2763-2773 (2001). 18 Foley, J. E. & Nieto, N. C. Tularemia. Veterinary microbiology 140, 332-
338, doi:10.1016/j.vetmic.2009.07.017 (2010). 19 Aktas, D. et al. Oropharyngeal Tularemia Outbreak Associated with
Drinking Contaminated Tap Water, Turkey, July-September 2013. Emerg Infect Dis 21, 2194-2196, doi:10.3201/eid2112.142032 (2015).
20 McCrumb, F. R. Aerosol Infection of Man with Pasteurella Tularensis.
Bacteriol Rev 25, 262-267 (1961). 21 Grunow, R., Kalaveshi, A., Kuhn, A., Mulliqi-Osmani, G. & Ramadani, N.
Surveillance of tularaemia in Kosovo, 2001 to 2010. Euro Surveill 17 (2012).
22 Syrjala, H., Kujala, P., Myllyla, V. & Salminen, A. Airborne transmission of
tularemia in farmers. Scand J Infect Dis 17, 371-375 (1985). 23 Teutsch, S. M. et al. Pneumonic tularemia on Martha's Vineyard. N Engl J
Med 301, 826-828, doi:10.1056/NEJM197910113011507 (1979).
131
24 Feldman, K. A. et al. An outbreak of primary pneumonic tularemia on Martha's Vineyard. N Engl J Med 345, 1601-1606, doi:10.1056/NEJMoa011374 (2001).
25 Johansson, A. et al. An outbreak of respiratory tularemia caused by
diverse clones of Francisella tularensis. Clin Infect Dis 59, 1546-1553, doi:10.1093/cid/ciu621 (2014).
26 Larsson, P. et al. The complete genome sequence of Francisella
tularensis, the causative agent of tularemia. Nat Genet 37, 153-159, doi:10.1038/ng1499 (2005).
27 Brett, M. E. et al. Outbreak of Francisella novicida bacteremia among
inmates at a louisiana correctional facility. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 59, 826-833, doi:10.1093/cid/ciu430 (2014).
28 Olsufjev, N. G. & Emelyanova, O. S. Immunological relationships between
Old and New World varieties of tularaemic bacteria. J Hyg Epidemiol Microbiol Immunol 7, 178-187 (1963).
29 Vogler, A. J. et al. Phylogeography of Francisella tularensis: global
expansion of a highly fit clone. Journal of bacteriology 191, 2474-2484, doi:10.1128/JB.01786-08 (2009).
30 Farlow, J. et al. Francisella tularensis in the United States. Emerging
infectious diseases 11, 1835-1841, doi:10.3201/eid1112.050728 (2005). 31 Johansson, A. et al. Worldwide genetic relationships among Francisella
tularensis isolates determined by multiple-locus variable-number tandem repeat analysis. Journal of bacteriology 186, 5808-5818, doi:10.1128/JB.186.17.5808-5818.2004 (2004).
32 Kugeler, K. J. et al. Molecular Epidemiology of Francisella tularensis in the
United States. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 48, 863-870, doi:10.1086/597261 (2009).
33 Molins, C. R. et al. Virulence difference between the prototypic Schu S4
strain (A1a) and Francisella tularensis A1a, A1b, A2 and type B strains in a murine model of infection. BMC infectious diseases 14, 67, doi:10.1186/1471-2334-14-67 (2014).
34 Keim, P., Johansson, A. & Wagner, D. M. Molecular epidemiology,
evolution, and ecology of Francisella. Ann N Y Acad Sci 1105, 30-66, doi:10.1196/annals.1409.011 (2007).
132
35 Taylor, J. P. et al. Epidemiologic characteristics of human tularemia in the
southwest-central states, 1981-1987. Am J Epidemiol 133, 1032-1038 (1991).
36 Hopla, C. E. The ecology of tularemia. Advances in veterinary science and
comparative medicine 18, 25-53 (1974). 37 Hopla, C. E. Experimental studies on tick transmission of tularemia
organisms. Am J Hyg 58, 101-118 (1953). 38 Morner, T. The ecology of tularaemia. Rev Sci Tech 11, 1123-1130
(1992). 39 Abd, H., Johansson, T., Golovliov, I., Sandstrom, G. & Forsman, M.
Survival and growth of Francisella tularensis in Acanthamoeba castellanii. Applied and environmental microbiology 69, 600-606 (2003).
40 Ellis, J., Oyston, P. C., Green, M. & Titball, R. W. Tularemia. Clin Microbiol
Rev 15, 631-646 (2002). 41 Rohmer, L. et al. Comparison of Francisella tularensis genomes reveals
evolutionary events associated with the emergence of human pathogenic strains. Genome biology 8, R102, doi:10.1186/gb-2007-8-6-r102 (2007).
42 Kingry, L. C. & Petersen, J. M. Comparative review of Francisella
tularensis and Francisella novicida. Front Cell Infect Microbiol 4, 35, doi:10.3389/fcimb.2014.00035 (2014).
43 Nano, F. E. et al. A Francisella tularensis pathogenicity island required for
intramacrophage growth. J Bacteriol 186, 6430-6436, doi:10.1128/JB.186.19.6430-6436.2004 (2004).
44 Ludu, J. S. et al. The Francisella pathogenicity island protein PdpD is
required for full virulence and associates with homologues of the type VI secretion system. J Bacteriol 190, 4584-4595, doi:10.1128/JB.00198-08 (2008).
45 Brotcke, A. et al. Identification of MglA-regulated genes reveals novel
virulence factors in Francisella tularensis. Infection and immunity 74, 6642-6655, doi:10.1128/IAI.01250-06 (2006).
46 Maurelli, A. T., Fernandez, R. E., Bloch, C. A., Rode, C. K. & Fasano, A.
"Black holes" and bacterial pathogenicity: a large genomic deletion that enhances the virulence of Shigella spp. and enteroinvasive Escherichia coli. Proc Natl Acad Sci U S A 95, 3943-3948 (1998).
133
47 Parkhill, J. et al. Comparative analysis of the genome sequences of
Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35, 32-40, doi:10.1038/ng1227 (2003).
48 Bliven, K. A. & Maurelli, A. T. Antivirulence genes: insights into pathogen
evolution through gene loss. Infect Immun 80, 4061-4070, doi:10.1128/IAI.00740-12 (2012).
49 Moore, R. A. et al. Contribution of gene loss to the pathogenic evolution of
Burkholderia pseudomallei and Burkholderia mallei. Infect Immun 72, 4172-4187, doi:10.1128/IAI.72.7.4172-4187.2004 (2004).
50 Rohmer, L. et al. Potential source of Francisella tularensis live vaccine
strain attenuation determined by genome comparison. Infect Immun 74, 6895-6906, doi:10.1128/IAI.01006-06 (2006).
51 Salomonsson, E. et al. Reintroduction of two deleted virulence loci
restores full virulence to the live vaccine strain of Francisella tularensis. Infect Immun 77, 3424-3431, doi:10.1128/IAI.00196-09 (2009).
52 Costa, T. R. et al. Secretion systems in Gram-negative bacteria: structural
and mechanistic insights. Nat Rev Microbiol 13, 343-359, doi:10.1038/nrmicro3456 (2015).
53 Michalska, M. & Wolf, P. Pseudomonas Exotoxin A: optimized by
evolution for effective killing. Front Microbiol 6, 963, doi:10.3389/fmicb.2015.00963 (2015).
54 Qin, A. & Mann, B. J. 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, doi:10.1186/1471-2180-6-69 (2006).
55 Weiss, D. S. et al. In vivo negative selection screen identifies genes
required for Francisella virulence. Proc Natl Acad Sci U S A 104, 6037-6042, doi:10.1073/pnas.0609675104 (2007).
56 Lindemann, S. R. et al. Francisella tularensis Schu S4 O-antigen and
capsule biosynthesis gene mutants induce early cell death in human macrophages. Infect Immun 79, 581-594, doi:10.1128/IAI.00863-10 (2011).
57 Golovliov, I., Ericsson, M., Sandstrom, G., Tarnvik, A. & Sjostedt, A.
Identification of proteins of Francisella tularensis induced during growth in
134
macrophages and cloning of the gene encoding a prominently induced 23-kilodalton protein. Infect Immun 65, 2183-2189 (1997).
58 Gray, C. G., Cowley, S. C., Cheung, K. K. & Nano, F. E. The identification
of five genetic loci of Francisella novicida associated with intracellular growth. FEMS Microbiol Lett 215, 53-56 (2002).
59 Golovliov, I., Sjostedt, A., Mokrievich, A. & Pavlov, V. A method for allelic
replacement in Francisella tularensis. FEMS Microbiol Lett 222, 273-280 (2003).
60 Telepnev, M., Golovliov, I., Grundstrom, T., Tarnvik, A. & Sjostedt, A.
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 (2003).
61 Telepnev, M., Golovliov, I. & Sjostedt, A. Francisella tularensis LVS
initially activates but subsequently down-regulates intracellular signaling and cytokine secretion in mouse monocytic and human peripheral blood mononuclear cells. Microbial pathogenesis 38, 239-247, doi:10.1016/j.micpath.2005.02.003 (2005).
62 Pukatzki, S. et al. Identification of a conserved bacterial protein secretion
system in Vibrio cholerae using the Dictyostelium host model system. Proc Natl Acad Sci U S A 103, 1528-1533, doi:10.1073/pnas.0510322103 (2006).
63 Hachani, A. et al. Type VI secretion system in Pseudomonas aeruginosa:
secretion and multimerization of VgrG proteins. J Biol Chem 286, 12317-12327, doi:10.1074/jbc.M110.193045 (2011).
64 Cianfanelli, F. R., Monlezun, L. & Coulthurst, S. J. Aim, Load, Fire: The
Type VI Secretion System, a Bacterial Nanoweapon. Trends Microbiol 24, 51-62, doi:10.1016/j.tim.2015.10.005 (2016).
65 de Bruin, O. M., Ludu, J. S. & Nano, F. E. The Francisella pathogenicity
island protein IglA localizes to the bacterial cytoplasm and is needed for intracellular growth. BMC Microbiol 7, 1, doi:10.1186/1471-2180-7-1 (2007).
66 Barker, J. R. et al. The Francisella tularensis pathogenicity island encodes
a secretion system that is required for phagosome escape and virulence. Mol Microbiol 74, 1459-1470 (2009).
67 Broms, J. E., Lavander, M., Meyer, L. & Sjostedt, A. IglG and IglI of the
Francisella pathogenicity island are important virulence determinants of
135
Francisella tularensis LVS. Infect Immun 79, 3683-3696, doi:10.1128/IAI.01344-10 (2011).
68 Broms, J. E., Meyer, L., Sun, K., Lavander, M. & Sjostedt, A. Unique
substrates secreted by the type VI secretion system of Francisella tularensis during intramacrophage infection. PLoS One 7, e50473, doi:10.1371/journal.pone.0050473 (2012).
69 Hare, R. F. & Hueffer, K. Francisella novicida pathogenicity island
encoded proteins were secreted during infection of macrophage-like cells. PLoS One 9, e105773, doi:10.1371/journal.pone.0105773 (2014).
70 Clemens, D. L., Ge, P., Lee, B. Y., Horwitz, M. A. & Zhou, Z. H. Atomic
structure of T6SS reveals interlaced array essential to function. Cell 160, 940-951, doi:10.1016/j.cell.2015.02.005 (2015).
71 Nano, F. E. & Schmerk, C. The Francisella pathogenicity island. Ann N Y
Acad Sci 1105, 122-137, doi:10.1196/annals.1409.000 (2007). 72 Long, M. E., Lindemann, S. R., Rasmussen, J. A., Jones, B. D. & Allen, L.
A. Disruption of Francisella tularensis Schu S4 iglI, iglJ, and pdpC genes results in attenuation for growth in human macrophages and in vivo virulence in mice and reveals a unique phenotype for pdpC. Infect Immun 81, 850-861, doi:10.1128/IAI.00822-12 (2013).
73 Baron, G. S. & Nano, F. E. MglA and MglB are required for the
intramacrophage growth of Francisella novicida. Mol Microbiol 29, 247-259 (1998).
74 Brotcke, A. & Monack, D. M. Identification of fevR, a novel regulator of
virulence gene expression in Francisella novicida. Infect Immun 76, 3473-3480, doi:10.1128/IAI.00430-08 (2008).
75 Bonquist, L., Lindgren, H., Golovliov, I., Guina, T. & Sjostedt, A. MglA and
Igl proteins contribute to the modulation of Francisella tularensis live vaccine strain-containing phagosomes in murine macrophages. Infect Immun 76, 3502-3510, doi:10.1128/IAI.00226-08 (2008).
76 Charity, J. C., Blalock, L. T., Costante-Hamm, M. M., Kasper, D. L. &
Dove, S. L. Small molecule control of virulence gene expression in Francisella tularensis. PLoS Pathog 5, e1000641, doi:10.1371/journal.ppat.1000641 (2009).
77 Buchan, B. W., McCaffrey, R. L., Lindemann, S. R., Allen, L. A. & Jones,
B. D. Identification of migR, a regulatory element of the Francisella tularensis live vaccine strain iglABCD virulence operon required for normal
136
replication and trafficking in macrophages. Infect Immun 77, 2517-2529, doi:10.1128/IAI.00229-09 (2009).
78 Faron, M. et al. The Francisella tularensis migR, trmE, and cphA genes
contribute to F. tularensis pathogenicity island gene regulation and intracellular growth by modulation of the stress alarmone ppGpp. Infect Immun 81, 2800-2811, doi:10.1128/IAI.00073-13 (2013).
79 Lenco, J. et al. Proteomics analysis of the Francisella tularensis LVS
response to iron restriction: induction of the F. tularensis pathogenicity island proteins IglABC. FEMS Microbiol Lett 269, 11-21, doi:10.1111/j.1574-6968.2006.00595.x (2007).
80 Buchan, B. W., McLendon, M. K. & Jones, B. D. Identification of
differentially regulated francisella tularensis genes by use of a newly developed Tn5-based transposon delivery system. Appl Environ Microbiol 74, 2637-2645, doi:10.1128/AEM.02882-07 (2008).
81 Jones, B. D., Faron, M., Rasmussen, J. A. & Fletcher, J. R. Uncovering
the components of the Francisella tularensis virulence stealth strategy. Frontiers in cellular and infection microbiology 4, 32, doi:10.3389/fcimb.2014.00032 (2014).
82 Vinogradov, E., Perry, M. B. & Conlan, J. W. Structural analysis of
Francisella tularensis lipopolysaccharide. Eur J Biochem 269, 6112-6118 (2002).
83 Thomas, R. M. et al. The immunologically distinct O antigens from
Francisella tularensis subspecies tularensis and Francisella novicida are both virulence determinants and protective antigens. Infect Immun 75, 371-378, doi:10.1128/IAI.01241-06 (2007).
84 Apicella, M. A. et al. Identification, characterization and immunogenicity of
an O-antigen capsular polysaccharide of Francisella tularensis. PLoS One 5, e11060, doi:10.1371/journal.pone.0011060 (2010).
85 Beutler, B. & Poltorak, A. Positional cloning of Lps, and the general role of
toll-like receptors in the innate immune response. Eur Cytokine Netw 11, 143-152 (2000).
86 Miller, S. I., Ernst, R. K. & Bader, M. W. LPS, TLR4 and infectious disease
diversity. Nat Rev Microbiol 3, 36-46, doi:10.1038/nrmicro1068 (2005). 87 Barker, J. H., Weiss, J., Apicella, M. A. & Nauseef, W. M. Basis for the
failure of Francisella tularensis lipopolysaccharide to prime human
137
polymorphonuclear leukocytes. Infect Immun 74, 3277-3284, doi:10.1128/IAI.02011-05 (2006).
88 Hajjar, A. M. et al. Lack of in vitro and in vivo recognition of Francisella
tularensis subspecies lipopolysaccharide by Toll-like receptors. Infection and immunity 74, 6730-6738, doi:10.1128/IAI.00934-06 (2006).
89 Walters, K. A. et al. Prior infection with Type A Francisella tularensis
antagonizes the pulmonary transcriptional response to an aerosolized Toll-like receptor 4 agonist. BMC Genomics 16, 874, doi:10.1186/s12864-015-2022-2 (2015).
90 Rasmussen, J. A. et al. Francisella tularensis Schu S4 lipopolysaccharide
core sugar and O-antigen mutants are attenuated in a mouse model of tularemia. Infect Immun 82, 1523-1539, doi:10.1128/IAI.01640-13 (2014).
91 Rasmussen, J. A., Fletcher, J. R., Long, M. E., Allen, L. A. & Jones, B. D.
Characterization of Francisella tularensis Schu S4 mutants identified from a transposon library screened for O-antigen and capsule deficiencies. Front Microbiol 6, 338, doi:10.3389/fmicb.2015.00338 (2015).
92 Crosa, L. M., Crosa, J. H. & Heffron, F. Iron transport in Francisella in the
absence of a recognizable TonB protein still requires energy generated by the proton motive force. Biometals 22, 337-344, doi:10.1007/s10534-008-9170-7 (2009).
93 Ramakrishnan, G. & Sen, B. The FupA/B protein uniquely facilitates
transport of ferrous iron and siderophore-associated ferric iron across the outer membrane of Francisella tularensis live vaccine strain. Microbiology 160, 446-457, doi:10.1099/mic.0.072835-0 (2014).
94 Perez, N. M. & Ramakrishnan, G. The reduced genome of the Francisella
tularensis live vaccine strain (LVS) encodes two iron acquisition systems essential for optimal growth and virulence. PLoS One 9, e93558, doi:10.1371/journal.pone.0093558 (2014).
95 Ramakrishnan, G., Meeker, A. & Dragulev, B. fslE is necessary for
siderophore-mediated iron acquisition in Francisella tularensis Schu S4. J Bacteriol 190, 5353-5361, doi:10.1128/JB.00181-08 (2008).
96 Thomas-Charles, C. A. et al. FeoB-mediated uptake of iron by Francisella
tularensis. Infection and immunity 81, 2828-2837, doi:10.1128/IAI.00170-13 (2013).
97 Sullivan, J. T., Jeffery, E. F., Shannon, J. D. & Ramakrishnan, G.
Characterization of the siderophore of Francisella tularensis and role of
138
fslA in siderophore production. J Bacteriol 188, 3785-3795, doi:10.1128/JB.00027-06 (2006).
98 Deng, K., Blick, R. J., Liu, W. & Hansen, E. J. Identification of Francisella
tularensis genes affected by iron limitation. Infection and immunity 74, 4224-4236, doi:10.1128/IAI.01975-05 (2006).
99 Kiss, K., Liu, W., Huntley, J. F., Norgard, M. V. & Hansen, E. J.
Characterization of fig operon mutants of Francisella novicida U112. FEMS Microbiol Lett 285, 270-277, doi:10.1111/j.1574-6968.2008.01237.x (2008).
100 Perez, N., Johnson, R., Sen, B. & Ramakrishnan, G. Two parallel
pathways for ferric and ferrous iron acquisition support growth and virulence of the intracellular pathogen Francisella tularensis Schu S4. Microbiologyopen, doi:10.1002/mbo3.342 (2016).
101 Twine, S. et al. 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. Infection and immunity 73, 8345-8352, doi:10.1128/IAI.73.12.8345-8352.2005 (2005).
102 Lindgren, H. et al. The 58-kilodalton major virulence factor of Francisella
tularensis is required for efficient utilization of iron. Infection and immunity 77, 4429-4436, doi:10.1128/IAI.00702-09 (2009).
103 Ramakrishnan, G., Sen, B. & Johnson, R. Paralogous outer membrane
proteins mediate uptake of different forms of iron and synergistically govern virulence in Francisella tularensis tularensis. J Biol Chem 287, 25191-25202, doi:10.1074/jbc.M112.371856 (2012).
104 Cartron, M. L., Maddocks, S., Gillingham, P., Craven, C. J. & Andrews, S.
C. Feo--transport of ferrous iron into bacteria. Biometals : an international journal on the role of metal ions in biology, biochemistry, and medicine 19, 143-157, doi:10.1007/s10534-006-0003-2 (2006).
105 Andrews, S. C., Robinson, A. K. & Rodriguez-Quinones, F. Bacterial iron
homeostasis. FEMS microbiology reviews 27, 215-237 (2003). 106 Hantke, K. Is the bacterial ferrous iron transporter FeoB a living fossil?
Trends in microbiology 11, 192-195 (2003). 107 Marlovits, T. C., Haase, W., Herrmann, C., Aller, S. G. & Unger, V. M. The
membrane protein FeoB contains an intramolecular G protein essential for Fe(II) uptake in bacteria. Proceedings of the National Academy of
139
Sciences of the United States of America 99, 16243-16248, doi:10.1073/pnas.242338299 (2002).
108 Kim, H., Lee, H. & Shin, D. The FeoA protein is necessary for the FeoB
transporter to import ferrous iron. Biochemical and biophysical research communications 423, 733-738, doi:10.1016/j.bbrc.2012.06.027 (2012).
109 Lau, C. K., Ishida, H., Liu, Z. & Vogel, H. J. Solution structure of
Escherichia coli FeoA and its potential role in bacterial ferrous iron transport. Journal of bacteriology 195, 46-55, doi:10.1128/JB.01121-12 (2013).
110 Su, Y. C. et al. Structure of Stenotrophomonas maltophilia FeoA
complexed with zinc: a unique prokaryotic SH3-domain protein that possibly acts as a bacterial ferrous iron-transport activating factor. Acta crystallographica. Section F, Structural biology and crystallization communications 66, 636-642, doi:10.1107/S1744309110013941 (2010).
111 Weaver, E. A., Wyckoff, E. E., Mey, A. R., Morrison, R. & Payne, S. M.
FeoA and FeoC are essential components of the Vibrio cholerae ferrous iron uptake system, and FeoC interacts with FeoB. Journal of bacteriology 195, 4826-4835, doi:10.1128/JB.00738-13 (2013).
112 Nagy, T. A., Moreland, S. M. & Detweiler, C. S. Salmonella acquires
ferrous iron from haemophagocytic macrophages. Molecular microbiology 93, 1314-1326, doi:10.1111/mmi.12739 (2014).
113 Cassat, J. E. & Skaar, E. P. Iron in infection and immunity. Cell host &
microbe 13, 509-519, doi:10.1016/j.chom.2013.04.010 (2013). 114 Aranda, J. et al. Contribution of the FeoB transporter to Streptococcus
suis virulence. International microbiology : the official journal of the Spanish Society for Microbiology 12, 137-143 (2009).
115 Velayudhan, J. et al. Iron acquisition and virulence in Helicobacter pylori: a
major role for FeoB, a high-affinity ferrous iron transporter. Molecular microbiology 37, 274-286 (2000).
116 Robey, M. & Cianciotto, N. P. Legionella pneumophila feoAB promotes
ferrous iron uptake and intracellular infection. Infection and immunity 70, 5659-5669 (2002).
117 Naikare, H., Palyada, K., Panciera, R., Marlow, D. & Stintzi, A. Major role
for FeoB in Campylobacter jejuni ferrous iron acquisition, gut colonization, and intracellular survival. Infection and immunity 74, 5433-5444, doi:10.1128/IAI.00052-06 (2006).
140
118 Lindgren, H., Lindgren, L., Golovliov, I. & Sjostedt, A. Mechanisms of
heme utilization by Francisella tularensis. PloS one 10, e0119143, doi:10.1371/journal.pone.0119143 (2015).
119 Imlay, J. A. The molecular mechanisms and physiological consequences
of oxidative stress: lessons from a model bacterium. Nat Rev Microbiol 11, 443-454, doi:10.1038/nrmicro3032 (2013).
120 Crichton, R. R. & Pierre, J. L. Old iron, young copper: from Mars to Venus.
Biometals : an international journal on the role of metal ions in biology, biochemistry, and medicine 14, 99-112 (2001).
121 Ramond, E. et al. Glutamate utilization couples oxidative stress defense
and the tricarboxylic acid cycle in Francisella phagosomal escape. PLoS pathogens 10, e1003893, doi:10.1371/journal.ppat.1003893 (2014).
122 Lindgren, H. et al. Iron content differs between Francisella tularensis
subspecies tularensis and subspecies holarctica strains and correlates to their susceptibility to H(2)O(2)-induced killing. Infection and immunity 79, 1218-1224, doi:10.1128/IAI.01116-10 (2011).
123 Bakshi, C. S. et al. Superoxide dismutase B gene (sodB)-deficient mutants
of Francisella tularensis demonstrate hypersensitivity to oxidative stress and attenuated virulence. Journal of bacteriology 188, 6443-6448, doi:10.1128/JB.00266-06 (2006).
124 Llewellyn, A. C., Jones, C. L., Napier, B. A., Bina, J. E. & Weiss, D. S.
Macrophage replication screen identifies a novel Francisella hydroperoxide resistance protein involved in virulence. PloS one 6, e24201, doi:10.1371/journal.pone.0024201 (2011).
125 Melillo, A. A. et al. Identification of Francisella tularensis live vaccine strain
CuZn superoxide dismutase as critical for resistance to extracellularly generated reactive oxygen species. Journal of bacteriology 191, 6447-6456, doi:10.1128/JB.00534-09 (2009).
126 Melillo, A. A., Bakshi, C. S. & Melendez, J. A. Francisella tularensis
antioxidants harness reactive oxygen species to restrict macrophage signaling and cytokine production. The Journal of biological chemistry 285, 27553-27560, doi:10.1074/jbc.M110.144394 (2010).
127 Binesse, J., Lindgren, H., Lindgren, L., Conlan, W. & Sjostedt, A. Roles of
reactive oxygen species-degrading enzymes of Francisella tularensis SCHU S4. Infect Immun 83, 2255-2263, doi:10.1128/IAI.02488-14 (2015).
141
128 Shakerley, N. L., Chandrasekaran, A., Trebak, M., Miller, B. A. & Melendez, J. A. Francisella tularensis catalase restricts immune function by impairing TRPM2 channel activity. The Journal of biological chemistry, doi:10.1074/jbc.M115.706879 (2015).
129 Rabadi, S. M. et al. Antioxidant defenses of Francisella tularensis
modulate macrophage function and production of proinflammatory cytokines. The Journal of biological chemistry, doi:10.1074/jbc.M115.681478 (2015).
130 Anthony, L. D., Burke, R. D. & Nano, F. E. Growth of Francisella spp. in
rodent macrophages. Infection and immunity 59, 3291-3296 (1991). 131 Ben Nasr, A. et al. Critical role for serum opsonins and complement
receptors CR3 (CD11b/CD18) and CR4 (CD11c/CD18) in phagocytosis of Francisella tularensis by human dendritic cells (DC): uptake of Francisella leads to activation of immature DC and intracellular survival of the bacteria. Journal of leukocyte biology 80, 774-786, doi:10.1189/jlb.1205755 (2006).
132 Buddingh, G. J. & Womack, F. C. Observations on the Infection of Chick
Embryos with Bacterium Tularense, Brucella, and Pasteurella Pestis. The Journal of experimental medicine 74, 213-222 (1941).
133 Lindemann, S. R., McLendon, M. K., Apicella, M. A. & Jones, B. D. An in
vitro model system used to study adherence and invasion of Francisella tularensis live vaccine strain in nonphagocytic cells. Infection and immunity 75, 3178-3182, doi:10.1128/IAI.01811-06 (2007).
134 Faron, M., Fletcher, J. R., Rasmussen, J. A., Apicella, M. A. & Jones, B.
D. Interactions of Francisella tularensis with Alveolar Type II Epithelial Cells and the Murine Respiratory Epithelium. PLoS One 10, e0127458, doi:10.1371/journal.pone.0127458 (2015).
135 McCaffrey, R. L. & Allen, L. A. Francisella tularensis LVS evades killing by
human neutrophils via inhibition of the respiratory burst and phagosome escape. J Leukoc Biol 80, 1224-1230, doi:10.1189/jlb.0406287 (2006).
136 Hall, J. D. et al. Infected-host-cell repertoire and cellular response in the
lung following inhalation of Francisella tularensis Schu S4, LVS, or U112. Infect Immun 76, 5843-5852, doi:10.1128/IAI.01176-08 (2008).
137 Celli, J. & Zahrt, T. C. Mechanisms of Francisella tularensis intracellular
pathogenesis. Cold Spring Harb Perspect Med 3, a010314, doi:10.1101/cshperspect.a010314 (2013).
142
138 Geier, H. & Celli, J. Phagocytic receptors dictate phagosomal escape and intracellular proliferation of Francisella tularensis. Infect Immun 79, 2204-2214, doi:10.1128/IAI.01382-10 (2011).
139 Clemens, D. L., Lee, B. Y. & Horwitz, M. A. Francisella tularensis enters
macrophages via a novel process involving pseudopod loops. Infect Immun 73, 5892-5902, doi:10.1128/IAI.73.9.5892-5902.2005 (2005).
140 Bradburne, C. E. et al. Temporal transcriptional response during infection
of type II alveolar epithelial cells with Francisella tularensis live vaccine strain (LVS) supports a general host suppression and bacterial uptake by macropinocytosis. J Biol Chem 288, 10780-10791, doi:10.1074/jbc.M112.362178 (2013).
141 Chong, A. et al. The early phagosomal stage of Francisella tularensis
determines optimal phagosomal escape and Francisella pathogenicity island protein expression. Infect Immun 76, 5488-5499, doi:10.1128/IAI.00682-08 (2008).
142 Golovliov, I., Baranov, V., Krocova, Z., Kovarova, H. & Sjostedt, A. An
attenuated strain of the facultative intracellular bacterium Francisella tularensis can escape the phagosome of monocytic cells. Infection and immunity 71, 5940-5950 (2003).
143 Clemens, D. L., Lee, B. Y. & Horwitz, M. A. Virulent and avirulent strains
of Francisella tularensis prevent acidification and maturation of their phagosomes and escape into the cytoplasm in human macrophages. Infect Immun 72, 3204-3217, doi:10.1128/IAI.72.6.3204-3217.2004 (2004).
144 Checroun, C., Wehrly, T. D., Fischer, E. R., Hayes, S. F. & Celli, J.
Autophagy-mediated reentry of Francisella tularensis into the endocytic compartment after cytoplasmic replication. Proc Natl Acad Sci U S A 103, 14578-14583, doi:10.1073/pnas.0601838103 (2006).
145 Lai, X. H., Golovliov, I. & Sjostedt, A. Francisella tularensis induces
cytopathogenicity and apoptosis in murine macrophages via a mechanism that requires intracellular bacterial multiplication. Infect Immun 69, 4691-4694, doi:10.1128/IAI.69.7.4691-4694.2001 (2001).
146 Lai, X. H. & Sjostedt, A. Delineation of the molecular mechanisms of
Francisella tularensis-induced apoptosis in murine macrophages. Infect Immun 71, 4642-4646 (2003).
143
147 Schulert, G. S. & Allen, L. A. Differential infection of mononuclear phagocytes by Francisella tularensis: role of the macrophage mannose receptor. J Leukoc Biol 80, 563-571, doi:10.1189/jlb.0306219 (2006).
148 Ben Nasr, A. & Klimpel, G. R. Subversion of complement activation at the
bacterial surface promotes serum resistance and opsonophagocytosis of Francisella tularensis. J Leukoc Biol 84, 77-85, doi:10.1189/jlb.0807526 (2008).
149 Clay, C. D., Soni, S., Gunn, J. S. & Schlesinger, L. S. Evasion of
complement-mediated lysis and complement C3 deposition are regulated by Francisella tularensis lipopolysaccharide O antigen. J Immunol 181, 5568-5578 (2008).
150 Schwartz, J. T. et al. Natural IgM mediates complement-dependent uptake
of Francisella tularensis by human neutrophils via complement receptors 1 and 3 in nonimmune serum. J Immunol 189, 3064-3077, doi:10.4049/jimmunol.1200816 (2012).
151 Dai, S., Rajaram, M. V., Curry, H. M., Leander, R. & Schlesinger, L. S.
Fine tuning inflammation at the front door: macrophage complement receptor 3-mediates phagocytosis and immune suppression for Francisella tularensis. PLoS Pathog 9, e1003114, doi:10.1371/journal.ppat.1003114 (2013).
152 Santic, M., Molmeret, M., Klose, K. E., Jones, S. & Kwaik, Y. A. The
Francisella tularensis pathogenicity island protein IglC and its regulator MglA are essential for modulating phagosome biogenesis and subsequent bacterial escape into the cytoplasm. Cell Microbiol 7, 969-979, doi:10.1111/j.1462-5822.2005.00526.x (2005).
153 Nauseef, W. M. Assembly of the phagocyte NADPH oxidase. Histochem
Cell Biol 122, 277-291, doi:10.1007/s00418-004-0679-8 (2004). 154 Allen, L. A. Mechanisms of pathogenesis: evasion of killing by
polymorphonuclear leukocytes. Microbes Infect 5, 1329-1335 (2003). 155 Rada, B. K., Geiszt, M., Kaldi, K., Timar, C. & Ligeti, E. Dual role of
phagocytic NADPH oxidase in bacterial killing. Blood 104, 2947-2953, doi:10.1182/blood-2004-03-1005 (2004).
156 McCaffrey, R. L. et al. Multiple mechanisms of NADPH oxidase inhibition
by type A and type B Francisella tularensis. J Leukoc Biol 88, 791-805, doi:10.1189/jlb.1209811 (2010).
144
157 KuoLee, R., Harris, G., Conlan, J. W. & Chen, W. Role of neutrophils and NADPH phagocyte oxidase in host defense against respiratory infection with virulent Francisella tularensis in mice. Microbes Infect 13, 447-456, doi:10.1016/j.micinf.2011.01.010 (2011).
158 Schulert, G. S. et al. Francisella tularensis genes required for inhibition of
the neutrophil respiratory burst and intramacrophage growth identified by random transposon mutagenesis of strain LVS. Infect Immun 77, 1324-1336, doi:10.1128/IAI.01318-08 (2009).
159 Santic, M., Molmeret, M. & Abu Kwaik, Y. Modulation of biogenesis of the
Francisella tularensis subsp. novicida-containing phagosome in quiescent human macrophages and its maturation into a phagolysosome upon activation by IFN-gamma. Cell Microbiol 7, 957-967, doi:10.1111/j.1462-5822.2005.00529.x (2005).
160 Clemens, D. L., Lee, B. Y. & Horwitz, M. A. Francisella tularensis
phagosomal escape does not require acidification of the phagosome. Infect Immun 77, 1757-1773, doi:10.1128/IAI.01485-08 (2009).
161 Fortier, A. H. et al. Growth of Francisella tularensis LVS in macrophages:
the acidic intracellular compartment provides essential iron required for growth. Infect Immun 63, 1478-1483 (1995).
162 Lindgren, H. et al. Factors affecting the escape of Francisella tularensis
from the phagolysosome. J Med Microbiol 53, 953-958, doi:10.1099/jmm.0.45685-0 (2004).
163 Barel, M., Meibom, K., Dubail, I., Botella, J. & Charbit, A. Francisella
tularensis regulates the expression of the amino acid transporter SLC1A5 in infected THP-1 human monocytes. Cell Microbiol 14, 1769-1783, doi:10.1111/j.1462-5822.2012.01837.x (2012).
164 Marohn, M. E. et al. Members of the Francisella tularensis phagosomal
transporter subfamily of major facilitator superfamily transporters are critical for pathogenesis. Infect Immun 80, 2390-2401, doi:10.1128/IAI.00144-12 (2012).
165 Gesbert, G. et al. Asparagine assimilation is critical for intracellular
replication and dissemination of Francisella. Cell Microbiol 16, 434-449, doi:10.1111/cmi.12227 (2014).
166 Gesbert, G. et al. Importance of branched-chain amino acid utilization in
Francisella intracellular adaptation. Infect Immun 83, 173-183, doi:10.1128/IAI.02579-14 (2015).
145
167 Jang, S. & Imlay, J. A. Micromolar intracellular hydrogen peroxide disrupts metabolism by damaging iron-sulfur enzymes. The Journal of biological chemistry 282, 929-937, doi:10.1074/jbc.M607646200 (2007).
168 Chong, A. & Celli, J. The francisella intracellular life cycle: toward
molecular mechanisms of intracellular survival and proliferation. Frontiers in microbiology 1, 138, doi:10.3389/fmicb.2010.00138 (2010).
169 Mariathasan, S., Weiss, D. S., Dixit, V. M. & Monack, D. M. Innate
immunity against Francisella tularensis is dependent on the ASC/caspase-1 axis. J Exp Med 202, 1043-1049, doi:10.1084/jem.20050977 (2005).
170 Henry, T., Brotcke, A., Weiss, D. S., Thompson, L. J. & Monack, D. M.
Type I interferon signaling is required for activation of the inflammasome during Francisella infection. J Exp Med 204, 987-994, doi:10.1084/jem.20062665 (2007).
171 Jones, J. W. et al. Absent in melanoma 2 is required for innate immune
recognition of Francisella tularensis. Proc Natl Acad Sci U S A 107, 9771-9776, doi:10.1073/pnas.1003738107 (2010).
172 Dotson, R. J. et al. Repression of inflammasome by Francisella tularensis
during early stages of infection. J Biol Chem 288, 23844-23857, doi:10.1074/jbc.M113.490086 (2013).
173 Wickstrum, J. R. et al. Francisella tularensis induces extensive caspase-3
activation and apoptotic cell death in the tissues of infected mice. Infect Immun 77, 4827-4836, doi:10.1128/IAI.00246-09 (2009).
174 Parmely, M. J., Fischer, J. L. & Pinson, D. M. Programmed cell death and
the pathogenesis of tissue injury induced by type A Francisella tularensis. FEMS Microbiol Lett 301, 1-11, doi:10.1111/j.1574-6968.2009.01791.x (2009).
175 Kurtz, S. L., Foreman, O., Bosio, C. M., Anver, M. R. & Elkins, K. L.
Interleukin-6 is essential for primary resistance to Francisella tularensis live vaccine strain infection. Infection and immunity 81, 585-597, doi:10.1128/IAI.01249-12 (2013).
176 Laws, T. R., Clark, G. & D'Elia, R. V. Differential role for interleukin-6
during Francisella tularensis infection with virulent and vaccine strains. Infection and immunity 81, 3055-3056, doi:10.1128/IAI.00234-13 (2013).
177 Ray, H. J. et al. Oral live vaccine strain-induced protective immunity
against pulmonary Francisella tularensis challenge is mediated by CD4+ T
146
cells and antibodies, including immunoglobulin A. Clinical and vaccine immunology : CVI 16, 444-452, doi:10.1128/CVI.00405-08 (2009).
178 Conlan, J. W. Tularemia vaccines: recent developments and remaining
hurdles. Future microbiology 6, 391-405, doi:10.2217/fmb.11.22 (2011). 179 Eneslatt, K. et al. Signatures of T cells as correlates of immunity to
Francisella tularensis. PloS one 7, e32367, doi:10.1371/journal.pone.0032367 (2012).
180 Eneslatt, K. et al. Persistence of cell-mediated immunity three decades
after vaccination with the live vaccine strain of Francisella tularensis. Eur J Immunol 41, 974-980, doi:10.1002/eji.201040923 (2011).
181 Foshay, L. A comparative study of the treatment of tularemia with immune
serum, hyperimmune serum and streptomycin. The American journal of medicine 1, 180-188 (1946).
182 Kirimanjeswara, G. S., Olmos, S., Bakshi, C. S. & Metzger, D. W. Humoral
and cell-mediated immunity to the intracellular pathogen Francisella tularensis. Immunological reviews 225, 244-255, doi:10.1111/j.1600-065X.2008.00689.x (2008).
183 Crane, D. D., Warner, S. L. & Bosio, C. M. A novel role for plasmin-
mediated degradation of opsonizing antibody in the evasion of host immunity by virulent, but not attenuated, Francisella tularensis. Journal of immunology 183, 4593-4600, doi:10.4049/jimmunol.0901655 (2009).
184 Crane, D. D., Scott, D. P. & Bosio, C. M. Generation of a convalescent
model of virulent Francisella tularensis infection for assessment of host requirements for survival of tularemia. PloS one 7, e33349, doi:10.1371/journal.pone.0033349 (2012).
185 Cole, L. E. et al. Antigen-specific B-1a antibodies induced by Francisella
tularensis LPS provide long-term protection against F. tularensis LVS challenge. Proceedings of the National Academy of Sciences of the United States of America 106, 4343-4348, doi:10.1073/pnas.0813411106 (2009).
186 Crane, D. D., Griffin, A. J., Wehrly, T. D. & Bosio, C. M. B1a cells enhance
susceptibility to infection with virulent Francisella tularensis via modulation of NK/NKT cell responses. Journal of immunology 190, 2756-2766, doi:10.4049/jimmunol.1202697 (2013).
187 Metzger, D. W., Salmon, S. L. & Kirimanjeswara, G. Differing effects of
interleukin-10 on cutaneous and pulmonary Francisella tularensis live vaccine strain infection. Infection and immunity 81, 2022-2027, doi:10.1128/IAI.00024-13 (2013).
147
188 Hunt, D. et al. Francisella tularensis elicits IL-10 via a PGE(2)-inducible
factor, to drive macrophage MARCH1 expression and class II down-regulation. PloS one 7, e37330, doi:10.1371/journal.pone.0037330 (2012).
189 Wilson, J. E., Katkere, B. & Drake, J. R. Francisella tularensis induces
ubiquitin-dependent major histocompatibility complex class II degradation in activated macrophages. Infection and immunity 77, 4953-4965, doi:10.1128/IAI.00844-09 (2009).
190 Woolard, M. D. et al. Francisella tularensis-infected macrophages release
prostaglandin E2 that blocks T cell proliferation and promotes a Th2-like response. J Immunol 178, 2065-2074 (2007).
191 Zaslona, Z., Serezani, C. H., Okunishi, K., Aronoff, D. M. & Peters-Golden,
M. Prostaglandin E2 restrains macrophage maturation via E prostanoid receptor 2/protein kinase A signaling. Blood 119, 2358-2367, doi:10.1182/blood-2011-08-374207 (2012).
192 Asakrah, S. et al. Post-exposure therapeutic efficacy of COX-2 inhibition
against Burkholderia pseudomallei. PLoS neglected tropical diseases 7, e2212, doi:10.1371/journal.pntd.0002212 (2013).
193 Cowley, S. C. & Elkins, K. L. Immunity to francisella. Frontiers in
microbiology 2, 26, doi:10.3389/fmicb.2011.00026 (2011). 194 Skyberg, J. A. Immunotherapy for tularemia. Virulence 4, 859-870,
doi:10.4161/viru.25454 (2013). 195 Parrow, N. L., Fleming, R. E. & Minnick, M. F. Sequestration and
scavenging of iron in infection. Infection and immunity 81, 3503-3514, doi:10.1128/IAI.00602-13 (2013).
196 Griffin, A. J., Crane, D. D., Wehrly, T. D. & Bosio, C. M. Successful
protection against tularemia in C57BL/6 mice is correlated with expansion of Francisella tularensis-specific effector T cells. Clin Vaccine Immunol 22, 119-128, doi:10.1128/CVI.00648-14 (2015).
197 Su, J. et al. Genome-wide identification of Francisella tularensis virulence
determinants. Infection and immunity 75, 3089-3101, doi:10.1128/IAI.01865-06 (2007).
198 Singh, R. & Eltis, L. D. The multihued palette of dye-decolorizing
peroxidases. Archives of biochemistry and biophysics 574, 56-65, doi:10.1016/j.abb.2015.01.014 (2015).
148
199 Zubay, G., Morse, D. E., Schrenk, W. J. & Miller, J. H. Detection and isolation of the repressor protein for the tryptophan operon of Escherichia coli. Proceedings of the National Academy of Sciences of the United States of America 69, 1100-1103 (1972).
200 Ashkenazy, H., Erez, E., Martz, E., Pupko, T. & Ben-Tal, N. ConSurf 2010:
calculating evolutionary conservation in sequence and structure of proteins and nucleic acids. Nucleic acids research 38, W529-533, doi:10.1093/nar/gkq399 (2010).
201 Landau, M. et al. ConSurf 2005: the projection of evolutionary
conservation scores of residues on protein structures. Nucleic acids research 33, W299-302, doi:10.1093/nar/gki370 (2005).
202 Troxell, B. & Hassan, H. M. Transcriptional regulation by Ferric Uptake
Regulator (Fur) in pathogenic bacteria. Frontiers in cellular and infection microbiology 3, 59, doi:10.3389/fcimb.2013.00059 (2013).
203 Kammler, M., Schon, C. & Hantke, K. Characterization of the ferrous iron
uptake system of Escherichia coli. Journal of bacteriology 175, 6212-6219 (1993).
204 Soares, M. P. & Weiss, G. The Iron age of host-microbe interactions.
EMBO Rep 16, 1482-1500, doi:10.15252/embr.201540558 (2015). 205 Radtke, A. L. & O'Riordan, M. X. Intracellular innate resistance to bacterial
pathogens. Cell Microbiol 8, 1720-1729, doi:10.1111/j.1462-5822.2006.00795.x (2006).
206 Paradkar, P. N. et al. Iron depletion limits intracellular bacterial growth in
macrophages. Blood 112, 866-874, doi:10.1182/blood-2007-12-126854 (2008).
207 Chlosta, S. et al. The iron efflux protein ferroportin regulates the
intracellular growth of Salmonella enterica. Infect Immun 74, 3065-3067, doi:10.1128/IAI.74.5.3065-3067.2006 (2006).
208 Pan, X., Tamilselvam, B., Hansen, E. J. & Daefler, S. Modulation of iron
homeostasis in macrophages by bacterial intracellular pathogens. BMC Microbiol 10, 64, doi:10.1186/1471-2180-10-64 (2010).
209 Hartley, G. et al. Grey variants of the live vaccine strain of Francisella
tularensis lack lipopolysaccharide O-antigen, show reduced ability to survive in macrophages and do not induce protective immunity in mice. Vaccine 24, 989-996, doi:10.1016/j.vaccine.2005.08.075 (2006).
149
210 Soni, S. et al. Francisella tularensis blue-gray phase variation involves structural modifications of lipopolysaccharide o-antigen, core and lipid a and affects intramacrophage survival and vaccine efficacy. Frontiers in microbiology 1, 129, doi:10.3389/fmicb.2010.00129 (2010).
211 Lindgren, H. et al. Resistance of Francisella tularensis strains against
reactive nitrogen and oxygen species with special reference to the role of KatG. Infection and immunity 75, 1303-1309, doi:10.1128/IAI.01717-06 (2007).
212 Ma, Z. et al. EmrA1 membrane fusion protein of Francisella tularensis LVS
is required for resistance to oxidative stress, intramacrophage survival and virulence in mice. Molecular microbiology 91, 976-995, doi:10.1111/mmi.12509 (2014).
213 Crane, D. D., Bauler, T. J., Wehrly, T. D. & Bosio, C. M. Mitochondrial
ROS potentiates indirect activation of the AIM2 inflammasome. Front Microbiol 5, 438, doi:10.3389/fmicb.2014.00438 (2014).
214 Bou-Abdallah, F., Lewin, A. C., Le Brun, N. E., Moore, G. R. & Chasteen,
N. D. Iron detoxification properties of Escherichia coli bacterioferritin. Attenuation of oxyradical chemistry. The Journal of biological chemistry 277, 37064-37069, doi:10.1074/jbc.M205712200 (2002).
215 Honarmand Ebrahimi, K., Hagedoorn, P. L. & Hagen, W. R. Unity in the
biochemistry of the iron-storage proteins ferritin and bacterioferritin. Chem Rev 115, 295-326, doi:10.1021/cr5004908 (2015).
216 Andrews, S. C. The Ferritin-like superfamily: Evolution of the biological
iron storeman from a rubrerythrin-like ancestor. Biochim Biophys Acta 1800, 691-705, doi:10.1016/j.bbagen.2010.05.010 (2010).
217 Weeratunga, S. K. et al. Binding of Pseudomonas aeruginosa
apobacterioferritin-associated ferredoxin to bacterioferritin B promotes heme mediation of electron delivery and mobilization of core mineral iron. Biochemistry 48, 7420-7431, doi:10.1021/bi900561a (2009).
218 Chen, C. Y. & Morse, S. A. Neisseria gonorrhoeae bacterioferritin:
structural heterogeneity, involvement in iron storage and protection against oxidative stress. Microbiology 145 ( Pt 10), 2967-2975, doi:10.1099/00221287-145-10-2967 (1999).
219 Mey, A. R., Wyckoff, E. E., Kanukurthy, V., Fisher, C. R. & Payne, S. M.
Iron and fur regulation in Vibrio cholerae and the role of fur in virulence. Infection and immunity 73, 8167-8178, doi:10.1128/IAI.73.12.8167-8178.2005 (2005).
150
220 Velayudhan, J., Castor, M., Richardson, A., Main-Hester, K. L. & Fang, F.
C. The role of ferritins in the physiology of Salmonella enterica sv. Typhimurium: a unique role for ferritin B in iron-sulphur cluster repair and virulence. Molecular microbiology 63, 1495-1507, doi:10.1111/j.1365-2958.2007.05600.x (2007).
221 Smith, J. L. The physiological role of ferritin-like compounds in bacteria.
Crit Rev Microbiol 30, 173-185, doi:10.1080/10408410490435151 (2004). 222 Ma, J. F. et al. Bacterioferritin A modulates catalase A (KatA) activity and
resistance to hydrogen peroxide in Pseudomonas aeruginosa. J Bacteriol 181, 3730-3742 (1999).
223 Reddy, P. V., Puri, R. V., Khera, A. & Tyagi, A. K. Iron storage proteins
are essential for the survival and pathogenesis of Mycobacterium tuberculosis in THP-1 macrophages and the guinea pig model of infection. Journal of bacteriology 194, 567-575, doi:10.1128/JB.05553-11 (2012).
224 Pandey, R. & Rodriguez, G. M. A ferritin mutant of Mycobacterium
tuberculosis is highly susceptible to killing by antibiotics and is unable to establish a chronic infection in mice. Infection and immunity 80, 3650-3659, doi:10.1128/IAI.00229-12 (2012).
225 Bakshi, C. S. et al. An improved vaccine for prevention of respiratory
tularemia caused by Francisella tularensis SchuS4 strain. Vaccine 26, 5276-5288, doi:10.1016/j.vaccine.2008.07.051 (2008).
226 Lee, B. Y., Horwitz, M. A. & Clemens, D. L. Identification, recombinant
expression, immunolocalization in macrophages, and T-cell responsiveness of the major extracellular proteins of Francisella tularensis. Infection and immunity 74, 4002-4013, doi:10.1128/IAI.00257-06 (2006).
227 Zaide, G. et al. Identification and characterization of novel and potent
transcription promoters of Francisella tularensis. Appl Environ Microbiol 77, 1608-1618, doi:10.1128/AEM.01862-10 (2011).
228 Mishra, S. & Imlay, J. Why do bacteria use so many enzymes to scavenge
hydrogen peroxide? Archives of biochemistry and biophysics 525, 145-160, doi:10.1016/j.abb.2012.04.014 (2012).
229 Postic, G. et al. Identification of small RNAs in Francisella tularensis. BMC
genomics 11, 625, doi:10.1186/1471-2164-11-625 (2010).
151
230 Bradley, J. M., Le Brun, N. E. & Moore, G. R. Ferritins: furnishing proteins with iron. J Biol Inorg Chem 21, 13-28, doi:10.1007/s00775-016-1336-0 (2016).
231 Savitt, A. G., Mena-Taboada, P., Monsalve, G. & Benach, J. L. Francisella
tularensis infection-derived monoclonal antibodies provide detection, protection, and therapy. Clinical and vaccine immunology : CVI 16, 414-422, doi:10.1128/CVI.00362-08 (2009).
232 Rotem, S. et al. Protective immunity against lethal F. tularensis holarctica
LVS provided by vaccination with selected novel CD8+ T cell epitopes. PloS one 9, e85215, doi:10.1371/journal.pone.0085215 (2014).
233 Bauler, T. J., Chase, J. C., Wehrly, T. D. & Bosio, C. M. Virulent
Francisella tularensis destabilize host mRNA to rapidly suppress inflammation. J Innate Immun 6, 793-805, doi:10.1159/000363243 (2014).
234 Navratil, A. R., Brummett, A. M., Bryan, J. D. & Woolard, M. D. Francisella
tularensis LVS induction of prostaglandin biosynthesis by infected macrophages requires specific host phospholipases and lipid phosphatases. Infect Immun 82, 3299-3311, doi:10.1128/IAI.02060-14 (2014).
235 Rockx-Brouwer, D. et al. Low dose vaccination with attenuated Francisella
tularensis strain SchuS4 mutants protects against tularemia independent of the route of vaccination. PloS one 7, e37752, doi:10.1371/journal.pone.0037752 (2012).
236 Richer, M. J., Nolz, J. C. & Harty, J. T. Pathogen-specific inflammatory
milieux tune the antigen sensitivity of CD8(+) T cells by enhancing T cell receptor signaling. Immunity 38, 140-152, doi:10.1016/j.immuni.2012.09.017 (2013).
237 Hubalek, M. et al. Comparative proteome analysis of cellular proteins
extracted from highly virulent Francisella tularensis ssp. tularensis and less virulent F. tularensis ssp. holarctica and F. tularensis ssp. mediaasiatica. Proteomics 4, 3048-3060, doi:10.1002/pmic.200400939 (2004).
238 Pavkova, I., Hubalek, M., Zechovska, J., Lenco, J. & Stulik, J. Francisella
tularensis live vaccine strain: proteomic analysis of membrane proteins enriched fraction. Proteomics 5, 2460-2467, doi:10.1002/pmic.200401213 (2005).
239 Peng, K., Broz, P., Jones, J., Joubert, L. M. & Monack, D. Elevated AIM2-
mediated pyroptosis triggered by hypercytotoxic Francisella mutant strains
152
is attributed to increased intracellular bacteriolysis. Cell Microbiol 13, 1586-1600, doi:10.1111/j.1462-5822.2011.01643.x (2011).
240 Hazlett, K. R. et al. Adaptation of Francisella tularensis to the mammalian
environment is governed by cues which can be mimicked in vitro. Infect Immun 76, 4479-4488, doi:10.1128/IAI.00610-08 (2008).
241 Singh, A. et al. Discordant results obtained with Francisella tularensis
during in vitro and in vivo immunological studies are attributable to compromised bacterial structural integrity. PLoS One 8, e58513, doi:10.1371/journal.pone.0058513 (2013).
242 Bocharov, G., Argilaguet, J. & Meyerhans, A. Understanding Experimental
LCMV Infection of Mice: The Role of Mathematical Models. J Immunol Res 2015, 739706, doi:10.1155/2015/739706 (2015).
243 Wehrly, T. D. et al. Intracellular biology and virulence determinants of
Francisella tularensis revealed by transcriptional profiling inside macrophages. Cell Microbiol 11, 1128-1150, doi:10.1111/j.1462-5822.2009.01316.x (2009).
244 Valentino, M. D. et al. A broadly applicable approach to T cell epitope
identification: application to improving tumor associated epitopes and identifying epitopes in complex pathogens. J Immunol Methods 373, 111-126, doi:10.1016/j.jim.2011.08.007 (2011).
245 Badovinac, V. P., Porter, B. B. & Harty, J. T. Programmed contraction of
CD8(+) T cells after infection. Nat Immunol 3, 619-626, doi:10.1038/ni804 (2002).
246 Pearce, E. L. & Shen, H. Making sense of inflammation, epigenetics, and
memory CD8+ T-cell differentiation in the context of infection. Immunol Rev 211, 197-202, doi:10.1111/j.0105-2896.2006.00399.x (2006).
247 Liu, X. et al. OxyR Is a Key Regulator in Response to Oxidative Stress in
Streptomyces avermitilis. Microbiology, doi:10.1099/mic.0.000251 (2016). 248 Ishiga, Y. & Ichinose, Y. Pseudomonas syringae pv. tomato OxyR Is
Required for Virulence in Tomato and Arabidopsis. Mol Plant Microbe Interact 29, 119-131, doi:10.1094/MPMI-09-15-0204-R (2016).
249 Chung, C. H., Fen, S. Y., Yu, S. C. & Wong, H. C. Influence of oxyR on
Growth, Biofilm Formation, and Mobility of Vibrio parahaemolyticus. Appl Environ Microbiol 82, 788-796, doi:10.1128/AEM.02818-15 (2015).
153
250 Christman, M. F., Storz, G. & Ames, B. N. OxyR, a positive regulator of hydrogen peroxide-inducible genes in Escherichia coli and Salmonella typhimurium, is homologous to a family of bacterial regulatory proteins. Proc Natl Acad Sci U S A 86, 3484-3488 (1989).
251 Schell, M. A. Molecular biology of the LysR family of transcriptional
regulators. Annu Rev Microbiol 47, 597-626, doi:10.1146/annurev.mi.47.100193.003121 (1993).
252 Imlay, J. A. Transcription Factors That Defend Bacteria Against Reactive
Oxygen Species. Annu Rev Microbiol 69, 93-108, doi:10.1146/annurev-micro-091014-104322 (2015).