The symptoms in anthrax depend on the
type of infection and can take anywhere from 1 day to more than 2 months to
appear. All types of anthrax have the potential, if untreated, to spread
throughout the body and cause severe illness and even death.
Four forms of human anthrax disease are
recognized based on their portal of entry.
•Cutaneous,
the most common form (95%), causes a localized, inflammatory, black, necrotic
lesion.
•Inhalation,
a rare but highly fatal form, is characterized by flu like symptoms, chest
discomfort, diaphoresis, and body aches.
•Gastrointestinal,
a rare but also fatal (causes death to 25%) type, results from ingestion of
spores. Symptoms include: fever and chills, swelling of neck, painful
swallowing, hoarseness, nausea and vomiting (especially bloody vomiting),
diarrhea, flushing and red eyes, and swelling of abdomen.
•Injection,
symptoms are similar to those of cutaneous anthrax, but injection anthrax can
spread throughout the body faster and can be harder to recognize and treat
compared to cutaneous anthrax.
Bacillus
anthracis Has Two Independent Bottlenecks That Are
Dependent on the Portal of Entry in an Intranasal Model of Inhalational
Infection
S.
R. Blanke, Editor
ABSTRACT
INTRODUCTION
Bacillus anthracis is a
Gram-positive sporulating bacterium that is the causative agent of anthrax.
Anthrax can initiate via cutaneous, gastrointestinal, and inhalation
inoculation, with the last two causing significant mortality. The spore is the
infectious particle and is notable for being metabolically dormant and
resistant to host-mediated killing and common disinfectants (1). Once within the
host, spores germinate and become vegetative bacilli, which cause the disease (2–4). The bacilli are
then able to replicate to high levels and disseminate through the host with the
aid of an antiphagocytic capsule and production of lethal toxin and edema toxin
(5–8). The combination of
bacteremia and toxicemia leads to host death. After the disease kills the host,
the bacilli sense the decreasing nutrient levels, sporulate within the carcass,
and return to the soil, where they reside until they infect the next host (1).
The high mortality rate associated with
inhalational anthrax is in part due to the bacterium's success in dissemination
through the host and a lack of diagnostic clinical signs of an infection until
late in infection. Early research with B. anthracis suggested
that inhalational infections disseminate via resident phagocytes acting as a
“Trojan horse” (9, 10). In the Trojan
horse model, inhaled spores are deposited in the alveoli and phagocytosed by
resident lung phagocytes. Either in transit or within the mediastinal lymph
nodes (MLN), the spore germinates, escapes from the phagocyte, and replicates
in the MLN; the bacteria eventually disseminate to the bloodstream via efferent
lymphatics (10–12). However,
experimentally testing this model for early events in pathogenesis is
challenging in animal models due to the fulminant nature of the infection and
its asynchronous dissemination (13, 14).
Bioluminescent imaging (BLI) of B.
anthracis strains with luminescent reporters enabled researchers to
dissect early events in pathogenesis that previously were difficult to study.
These BLI studies identified two novel sites of establishment and dissemination
of infection (15–18). In the upper
respiratory passage, bacteria were first observed to replicate and disseminate
from the nasal mucosa-associated lymphoid tissue (NALT) (13, 15, 16, 19). A second portal of
entry was identified in the lower respiratory passage in mice, where
germination and outgrowth occurred directly in the alveolar lumen (15, 17). The finding of
germination and dissemination from the lung was a surprising result,
since B. anthracis does not generally provoke a classic
bronchopneumonia and there is evidence that the lung is refractory to
germination (10, 14–17, 20). However, induction
of germination in the lungs or preexisting damage to the lungs has been
previously shown to permit spore germination and outgrowth in the lungs (16, 21, 22). While several
studies have sought to understand the Trojan horse model of dissemination,
fewer studies have investigated the dissemination from NALT- and lung-based portals
of entry.
The dissemination patterns of several
bacterial pathogens have been analyzed through clonal analysis to clarify
portals of entry and understand how the host alters the pathogen's population
dynamics (23–27). Clonal analysis
consists of creating a library of isogenic clones that vary only by a signature
tag, e.g., an antibiotic marker or recombinant DNA sequence. The host is then
infected with an equally proportioned mixture of clones that are isolated from
tissues of interest at particular time points to determine if and how the
proportions of clones vary during dissemination. These analyses have been used
to determine the sequence of dissemination as well as to identify population
bottlenecks for several pathogens (23–25, 27). Such bottlenecks
increase the rate of genetic drift and allelic fixation, leading to a drastic
reduction of genomic diversity at the population level, which can be measured
by effective population (28). Effective
population differs from population size because it measures the intensity or
rate of genetic drift on a population, rather than the number of individuals.
While previous research has demonstrated that NALT-based bottlenecks occur as
the bacilli disseminate, it is unknown whether a bottleneck occurs in
lung-based infections (26). If bottlenecks
were to occur in both sites, they could be compared to inform researchers on
how the site of establishment influences the ensuing infection.
The research described here compared dissemination
of B. anthracis from the lungs and dissemination from the NALT
using a library of luminescent DNA-tagged clones of the B. anthracis Sterne
strain. The simultaneous use of BLI with clonal analysis was essential to
dissect how spatial and temporal events alter the bacterial population
dynamics in vivo. Lung-based infections initiate when a small
number of spores germinate in a focal region and spread rapidly through the
host. Since lung infections initiate from a small, distinct focal area and rapidly
disseminate through the host, it is difficult to envision such an event being
captured by classic pathological or diagnostic methods with significant
frequency. Compared to NALT-based infections, lung-based infections undergo
bottlenecks at the site of germination and have a significantly higher
effective population. This suggests that B. anthracis is able
to establish infections in the unique microenvironments of the NALT or lung
lumen within the host respiratory system and that these locations impart
distinct characteristics to the resultant disseminated bacterial population
that reflect how the infection unfolds within the host.
MATERIALS AND
METHODS
Bacterial strains, culture procedures,
and media.
Bacillus anthracis Sterne strain
7702 (pX01+, pX02−) was used for
construction of clonal mutants (see below) and was obtained from BEI (Manassas,
VA). A luminescent B. anthracis 7702 strain, BIG23, was
constructed by integrating the plasmid pIG6-19 into pX01, a kind gift from
Michele Mock (18). The pIG6-19
construct is a derivative of the pAT113 conjugative suicide plasmid, which
allowed for insertion of the plasmid upstream of the endogenous paglocus
without interrupting the gene. This strain contains the luxABCDE operon
from Photorhabdus luminescens under the control of the
protective antigen (PA) promoter, leading to light expression in vivo.
BIG23 was used as the base strain for all mouse infections. B.
anthracis was grown in brain heart infusion (BHI) broth (Becton
Dickson, MD) supplemented with 0.5% glycerol at 37°C unless otherwise noted.
Spores were produced by plating on nutrient broth-yeast extract (NBY) medium
with antibiotics when necessary, incubated at 30°C for 10 days, and then
harvested and purified on an Omnipaq gradient (350 mg of iodine/ml) (GE
Healthcare, NJ), similar to previously described methods (16, 18, 29). Escherichia
colistrains were either α-select (Bioline, MA) or dcm-
and dam-deficient GM119 (dcm-6 dam-3) (30) and were grown in
LB broth with appropriate antibiotics when necessary (Becton Dickson, MD). All
strains and plasmids are described in Table
1.
Strains
and plasmids used in this study
Construction of B. anthracis clonal
mutants.
Unique DNA-tagged clones were
constructed by inserting recombinant DNA from a progressive 100-bp DNA step
ladder (Promega, WI) into the chromosomal B. anthracis eag gene.
DNA tags were inserted into B. anthracis by allelic exchange
with pSE02 and pCZ03, which are modified from the pHY304 temperature-sensitive
allelic exchange vector, kindly provided by S. Leppla (Table
1)
(31). To facilitate
allelic exchange in B. anthracis, a 2-kb section of the eag gene
was constructed by splice-by-overlap (SOE) PCR using primers FIG4, RIG4, FIG5,
and RIG5, which also introduced a BamHI site (see Table S1 in the supplemental
material). The modified eagfragment was then inserted into pHY304
to generate pIG13. pIG13 then had two different multiple-cloning sites (MCS)
inserted into the recombinant BamHI site in eag to allow for
greater flexibility for cloning in DNA tags. The pSE02 vector contains a short
DNA linker with an EcoRV site and a SalI site flanked by BamHI sites, whereas
pCZ03 has the linker flanked by EcoRI, SpeI, and NotI sites. The pSE02 vector
was constructed by SOE PCR where the overlap region contained an EcoRV, SpeI,
SalI, and BamHI sites within the eag gene. The two SOE regions
were amplified from pIG13 using RDL22 and F SOE linker and FIG8 and R SOE
linker and were spliced together by PCR using the FIG8 and RDL22 primers. The
pCZ03 vector was generated by inserting the linker into the MCS of pGEM-T Easy
(Promega, WI) and then isolating the MCS containing pGEM-T Easy plasmid by EagI
digestion and blunting using the NEB Quick Blunt kit (NEB, MA). Next, pIG13 was
digested with BamHI blunted using the NEB Quick Blunt kit, and the MCS was
inserted into pIG13. The fragments from the DNA ladder were blunted with the
NEB Quick Blunt kit and ligated into the EcoRV site in the eag region
of homology of the MCS of either pCZ03 or pSE02 using T4 ligase (NEB, MA).
After pSE02 and pCZ03 were ligated with
fragments from a progressive 100-bp step ladder, they were transformed into
α-select chemically competent E. coli (Bioline, MA).
Transformants were then screened with M13 primers to ensure that the fragments
were ligated into the plasmid. The verified plasmid was then passaged through
GM119 to remove DNA methylation and purified before electroporation into 7702
as previously described (32). Mutants were
passaged at 42°C twice to select for merodiploid insertion of the
temperature-sensitive plasmid carrying erythromycin (Erm) resistance into eag and
then repeatedly passaged at 37°C until bacteria became Erm sensitive,
indicating that allelic exchange had occurred. PCR was used to identify
bacteria with the DNA ladder insert and confirm the presence of pX01. Lastly,
to make the tagged strains luminescent, pIG6-19 was conjugated into tagged 7702
strains and selected on BHI plates supplemented with 5 μg/ml Erm and 60 μg/ml
polymyxin B (33). Strains were
sporulated and purified as described above.
Determination of growth rates of B.
anthracis clones.
Strains were grown overnight in BHI
medium at 37°C and back diluted to an optical density at 600 nm of ∼0.005
in 50 ml of fresh BHI medium. The optical density was then measured every hour
for 8 h and then at 12 h and 24 h. This was repeated three times independently.
Infection and dissection of mice.
Mouse manipulations and husbandry were
performed in accordance with the guidelines set forth in the Guide for
the Care and Use of Laboratory Animals of the National Research
Council (34). All techniques and
animal husbandry were approved by the University of Virginia Animal Care and Use
Committee (protocol 3671) and were designed to minimize distress and pain for
the animals. All mice were 6- to 12-week-old female A/J mice (Jackson
Laboratory, ME) and were housed and/or bred using specific-pathogen-free
rearing procedures at the University of Virginia. Infections were performed by
anesthetizing mice with methoxyflurane (Metafane; Matrix Scientific, SC) in a
nose cone and then delivering spores by intranasal instillation as previously
described (16). The spore inoculum
varied from 5 × 106 CFU/mouse to 5
× 107 CFU/mouse. In
preliminary studies, some mice would be infected with 1.4 × 108 CFU/mouse to
develop lung infections; however, lung infections were found to occur at lower
doses. The infections were monitored by anesthetizing mice with a 2.5%
(vol/vol) isoflurane-oxygen mix with an XGI-8 gas anesthesia system
(Perkin-Elmer, MA), and light production was followed in the host with either
the IVIS100 or the IVIS Spectrum (Perkin-Elmer, MA) until light was detected in
the NALT or ventral thoracic cavity after 1 min of exposure with large binning.
Light production was measured and quantified using LivingImage software
(version 3.2; Perkin-Elmer, MA). Any mouse with light in the kidneys was
considered to have a disseminated infection and was euthanized.
After euthanasia, the NALT, cervical
lymph nodes (CLN), mediastinal lymph nodes, lungs, and kidney were removed and
homogenized on ice using Dounce homogenizers and ice-cold 1× phosphate-buffered
saline (PBS) as previously described (35). The homogenates
were divided in half, with one half heated to 65°C for 20 min to kill
vegetative bacilli such that the remaining CFU represent spores. Both fractions
were serially diluted and spread on LB plates with 5 μg/ml Erm to enumerate
CFU. A portion of each sample was also used to inoculate 3 ml BHI broth with 5
μg/ml Erm and grown overnight at 37°C. If any bacterial growth occurred in
overnight cultures, an aliquot was brought to 10% (vol/vol) glycerol and stored
as frozen samples in order to confirm clonal proportions or increase the number
of colonies analyzed at a later date, if necessary.
Determination of clonal identity in
organs.
Clonal proportions were established for
each organ that contained at least 30 CFU. Organs that contained fewer than 30
CFU were not used for further study, as small numbers of bacteria are subject
to greater skewing of clonal proportions due to chance. For each organ, 48
colonies were individually lysed using the HotSHOT technique to extract DNA (36). Using 5 μl of the
lysed colony DNA sample, clonal identity was established by PCR amplification
of the eag region flanking the DNA tag using Agilent Paq5000
polymerase (Agilent, CA) and primers FDL26 and RDL26 to yield unique fragment
sizes for each clone.
Probability modeling to determine the
size of the clonal library.
The probability of two exact clones
passing through a bottleneck was determined by using the hypergeometric
formula h(x; N, n, k) =
(kCx)(N−kCn−x)/(NCn), where N is
the inoculum delivered in CFU, x is the number of identical
clones found in the kidney, k is the number of CFU from a
particular clone in the inoculum, and n is the number of
clones that pass through the bottleneck (bottleneck size). The probability was
then calculated for a given bottleneck size for a given number of clones in the
library.
Statistical analysis.
Student t tests,
one-way and two-way analyses of variance (ANOVAs), Sidak's multiple-comparison
posttest, and Tukey multiple-comparison posttests were performed using GraphPad
Prism (version 5; GraphPad Software, CA). Growth rates were statistically
compared by nonlinear regression and comparing F values from
each strain. Fisher's exact tests were performed using the R project for
statistical programming version 2.15.0 (The R Foundation for Statistical
Programming, Vienna, Austria).
RESULTS
Bacterial burdens are similar
when B. anthracis disseminates from NALT and lung portals of
entry.
BLI analysis of infections identified
the NALT and the lung lumen as sites of initial germination and growth
for B. anthracis (15–17). Because BLI
analysis indicated that the bacteria had different dissemination patterns in
NALT- and lung-based infections, it was hypothesized that these routes of
dissemination may have different levels of bacterial burden in similar lymph
nodes. To test this hypothesis, A/J mice were infected intranasally with BIG23,
a luminescent derivative of Sterne 7702. BIG23 has the genes necessary for
light production under the control of the protective antigen promoter and
produces light when vegetative bacteria reside under the toxin-inducing
conditions found throughout the host (18). This construct is
capable of having luminescence detected both in vitro and in
vivo as early as 2 h after germination and has no differences in its
ability to produce PA relative to that of the parental 7702 (see Fig. S1 in the
supplemental material) (16, 17). Intranasally
infected mice developed either NALT- or lung-based infections (Fig.
1).
Concurrent dual-portal infections were observed in only 2 mice out of all 30 of
our infections; therefore, mice were analyzed only if there was luminescence
exclusively in the NALT or the lungs to avoid confounding each location's
contribution to the infectious process. After either a NALT- or lung-based
infection was detected, the cervical lymph nodes (CLN), lungs, MLN, and kidneys
were homogenized and the total bacteria and spore populations were analyzed.
Since spores are more resistant to heat than bacilli, any CFU obtained from the
heated homogenate is considered to have been derived from an ungerminated
spore. Thus, the proportion of the bacterial population consisting of spores
can be isolated by heating an aliquot of the tissue homogenate and comparing it
to an unheated aliquot.
Inhalational B.
anthracis infections cause similar bacterial burdens in most organs.
(A) Dorsal and ventral black-and-white photographs of an A/J mouse with
pseudocolor overlay of luminescence intensity. The mouse was infected with 4.6
× 106 spores ...
A NALT infection was indicated when the
mouse had measurable luminescence above background solely in the nasal passages
(Fig.
1A).
Luminescence from NALT-based infections were detected at between 2 and 7 days
postinfection (dpi), with the median at 3 dpi (see Fig. S2A in the supplemental
material). In agreement with previous publications, these mice had luminescence
detected in the NALT, in the CLN, and, if the infections were allowed to
progress, in the kidney (data not shown). The kidneys were chosen as a proxy to
indicate systemically disseminated bacilli since the kidneys reach a high
bacterial burden, lack spores, and are the first major organs to become
luminescent after dissemination (15). Mice that had
infections disseminate from the NALT did not show luminescent foci in their
lungs when dissected (data not shown) (15, 17). Bacteria in the
lymph nodes and kidneys were almost exclusively heat sensitive, as all CFU from
heat-resistant spores were below the limit of detection (Fig.
1B).
Enumeration of CFU in the lungs demonstrated that spores were present in the
lungs even when a NALT-based infection occurred. However, the heat-resistant
bacteria comprised only ∼10% of the bacterial population when
mice with luminescence in the CLN were dissected (1.1 × 106 CFU total
bacteria [standard deviation {SD}, ±9.9 × 105 CFU] versus 1.1 × 105 heat-resistant
CFU [SD, ±1.6 × 105 CFU]).
A lung infection was defined as when
light was detected solely through the ventral thorax, as previously described (Fig.
1C)
(16, 17). Luminescence from
lung-based infections were detected at between 3 and 8 days dpi, with the
median at 5 dpi (see Fig. S2A in the supplemental material). A correlation
between the CFU in the organs and the luminescence in NALT- and lung-based
infection indicated that a higher bacterial burden in the lungs was needed to
be detected (see Fig. S2B in the supplemental material). Upon closer
examination after dissection, the luminescence in lung infections was found to
be confined to a single focal region in the lungs and was regularly detected in
the draining MLN. If the lung infection was allowed to progress to later
stages, the kidneys also displayed measurable luminescence, in agreement with
previous publications (data not shown) (15–17). Similarly, in
lung-based infections the MLN and the kidneys contained high numbers of
heat-sensitive CFU, and all heat-resistant CFU were below the limit of detection
(Fig.
1D).
Additionally, the lungs in lung-based infections contained 12-fold more
heat-sensitive bacteria than heat-resistant bacteria. The NALT was not analyzed,
since luminescence was not detected when a lung infection occurred, suggesting
that it lacked vegetative bacilli. The CLN, however, typically had very small
amounts of CFU present. In some cases, there was bacterial growth in the LB
broth that was inoculated with the CLN homogenates. This suggested that
bacteria were present but were too few to enumerate directly from the tissue.
The lung-based bottleneck is distinct
from the NALT-based bottleneck.
B. anthracis undergoes a
bottleneck as it disseminates from the NALT to the draining CLN (26). A population
bottleneck occurs when there is a dramatic population decrease that leads to a
nonbiased reduction in genetic variability relative to that of the original
population. BLI analysis has revealed that B. anthracis has
two portals of entry for inhalational anthrax (15). Given that the
respiratory mucosa in the alveoli is distinct from the NALT and that there were
differences in the bacterial burden in the CLN, we tested to see if lung-based
infections also passed through a bottleneck. Probability modeling determined
that 5 clones was the optimum number of DNA-tagged clones to allow for the
greatest resolution of the bottleneck without causing unnecessary experimental
burden (see Fig. S3 in the supplemental material). DNA-tagged strains were then
tested for growth rate defects and in vivo virulence to ensure
that B. anthracis was undergoing a bottleneck and not
selection against a less virulent strain. A comparison of the in vitro growth
rates showed no significant growth defects between any of the tagged strains
and the parent BIG23 strain (see Fig. S4A in the supplemental material)
(nonlinear regression and comparison of F values, P =
0.9991). Additionally, there were no statistically significant differences
between the frequencies of clones in the kidneys and those delivered to the
mouse, supporting the conclusion that no single clone had a competitive
advantage over the others and that passage through a bottleneck is random (see
Fig. S4B in the supplemental material) (Fisher's exact test, P =
0.521).
In agreement with previous
publications, NALT-based infections were found to have no difference in the
clonal numbers or proportions between the NALT and the initial inoculum early
in infection (see Fig. S5 in the supplemental material). However, NALT-based
infections undergo a significant reduction in the number of clones between the
NALT and CLN, as all organs other than the NALT contained an average of 1 clone
(Fig.
2A)
(one-way ANOVA with Tukey multiple-comparison posttest, P <
0.0001) (26). Mice with
lung-based infections had a less pronounced decrease in the number of clones
represented as the bacilli disseminated from the lung lumen (Fig.
2B).
Organs were dissected, and clonal identities were determined when light was
detected in the ventral thoracic cavity but before luminescence was detected in
the kidneys. At this stage of infection, there were too few CFU in the CLN for
clonal analysis, and the CLN therefore were not analyzed. This allowed for the
determination of the average number of clones in each organ before the bacteria
returned to the lungs in large numbers from the circulatory system, which would
confound analysis. The number of clones represented in the whole lungs was not
significantly different from the number of clones delivered in the initial
inoculum (Fig.
2B)
(one-way ANOVA with Tukey multiple-comparison posttest, P >
0.05). The average number of clones in the MLN, which drain bacteria from the
lungs, ranged from 1 to 4 clones, with a significantly lower average of 2
clones represented, compared to the initial inoculum's 5 clones (Fig.
2B)
(one-way ANOVA with Tukey multiple-comparison posttest, P <
0.0001). In areas where heat-resistant spores were never found, i.e., the
kidneys, there was an average of 2 different clones, which was significantly
lower than in the inoculum (Fig.
2B)
(one-way ANOVA with Tukey multiple-comparison posttest, P <
0.0001).
Mice
with lung-based bottlenecks have a higher number of clones in the lungs and
MLN. (A) Number of different clones represented in each organ of mice dissected
after a NALT-based infection had progressed to the draining CLN or kidneys.
Each symbol represents ...
To test if there were differences between
the bottlenecks in NALT-based infections and lung-based infections, the numbers
of different clones found in the lungs, MLN, and kidneys were compared. Mice
with lung infections did not have bacteria present in their CLN until they were
moribund and were thus not compared. In contrast to those with NALT-based
infections, mice with lung-based infections had a significantly higher number
of clones represented in the lungs and MLN (Fig.
2C)
(two-way ANOVA with Sidak's multiple-comparison test, P <
0.01 and P < 0.0001).
Lung-based infections have disseminated
clonal populations that do not match populations in the MLN and the whole lung,
whereas NALT-based infections have disseminated populations that resemble those
in the CLN.
In order to determine whether the clone
or clones that pass through the bottleneck are the founders of the disseminated
population in the host, the clonal proportions in each organ were analyzed. BLI
was used to synchronize the infections to when light was first detected in the
kidneys. Organs were then removed from the infected mice, and the proportions
and identities of clones were compared between the organs to determine
dissemination patterns. NALT-based infections were found to be almost
completely comprised of 1 clone in all organs, and the clone found in the CLN
had the same identity as that in the kidneys (Fig.
3A)
(Fisher's exact test, P > 0.05). The lungs in mouse 1 had
all clones present; however, the lung homogenates were further tested to
determine if this population represented ungerminated spores in the lungs. The
percent germination was calculated by comparing heated and unheated lung tissue
homogenates. There was a small difference in CFU between the heated (spores
only) and unheated (bacilli and spores) lung homogenates (1.2 × 105 spore CFU/lung
and 1.8 × 105 total bacterial
CFU/lung), implying that most bacteria in the lungs remained as spores.
Additionally, there was no statistical difference between the clonal
proportions from the heat-resistant bacteria in the lung homogenates and total
bacterial lung homogenates for mouse 1, further suggesting that the lungs
contained only spores (Fig.
3A)
(Fisher's exact test, P > 0.05). This suggests that the
increased number of different clones in the lungs in mouse 1 is due to it being
dissected before the disseminated bacteria returned to the lungs after
dissemination from the NALT and CLN.
Clonal
proportions in NALT-based infections are similar between organs, whereas
lung-based infections show greater diversity. (A) Clonal proportions in the
CLN, lungs, and kidneys of mice that had a NALT infection that disseminated at
least to the draining ...
In agreement with the data in Fig.
2,
the clonal proportions in lungs and MLN present in lung-based infections were
more diverse than those in NALT-based infections. Out of six mice, only mice E
and F had a single clone constitute the majority of the whole lung, i.e., 82%
and 52%, respectively (Fig.
3B)
(Fisher's exact test, P > 0.05). Importantly, mice E and F
were dissected later than the other four mice, meaning that the bacilli had a
longer time to replicate and disseminate through the mouse, which could allow
recolonization of the lung from the periphery. The other 4 mice had no more
than 35% of the clonal proportions containing a single clone. Therefore, if a
bottleneck occurred, clonal analysis was not sufficient to detect lung
bottlenecks when the whole lungs were analyzed in a single homogenate.
Similarly, only one mouse out of six had no statistical difference between the
kidneys and MLN, potentially suggesting the bacilli disseminated without using
the lung lymphatics (Fig.
3C)
(Fisher's exact test, P < 0.0001).
The focal regions of bacterial growth
arise from a few clones that replicate in a small area, and these founders go
on to disseminate through the host.
The population structure of the
disseminated bacteria from lung-based infections suggested that a bottleneck
occurred by the time the bacteria were systemically disseminated, but
whole-lung analysis was potentially hindered by the high spore burden in the
lungs (Fig.
1D and and2B).2B). Since previous publications have
noted that lung germination and outgrowth occur in distinct focal regions in
the lung, it was investigated whether these regions contained a higher
proportion of bacilli than spores (15, 17). Moreover, the high
degree of luminescence within this region suggests a higher proportion of
bacilli, which would reduce the likelihood of sampling spores and allow for a
greater ability to determine if a bottleneck occurred in the lungs. After
luminescence was detected in the ventral thoracic cavity, the lungs were
removed and separated into luminescent and nonluminescent regions (Fig.
4A).
The focal luminescent region had a greater percentage of bacilli than the
nonluminescent regions as determined by heat sensitivity (Fig.
4B)
(Student's t test, P < 0.0001).
Germination in the nonluminescent region averaged 39% (SD, 15.9%), indicating
that the majority of the bacteria were still in spore form. However, an average
of 86% (SD, 18.9%) of the bacteria in the luminescent focal region were heat
sensitive in addition to expressing a luminescent reporter that produces light
only in metabolically active vegetative bacilli. This is in contrast to the
NALT, where luminescence is detected throughout the anterior portion of the
palate (see Fig. S6 in the supplemental material). If a bottleneck occurred in
the focal region, this could suggest that a few spores in the focal region
germinated, replicated, and eventually disseminated. Alternatively, it is
possible that all 5 clones in the library germinated in a small area, but only
the first few clones that reached the draining lymph node or bloodstream were
the founders of the disseminated population. To differentiate between these two
scenarios, the clonal diversities of the focal regions were analyzed separately
from the rest of the lungs and compared to the clonal diversity of a similarly
sized nonluminescent area of the lung. Mice were first dissected when there was
clearly luminescence present in the thoracic cavity, which corresponded to
>3.5 × 106 CFU/g in the
focal region. In these mice, all 5 mice had significant differences in clonal
diversity between the initial inoculum and the focal region (Fig.
5A)
(Fisher's exact test, P < 0.0001). Two mice, mouse 4 and
mouse 5, had the same clonal population comprise a majority in the focal region
as well as in the nonluminescent region. These mice had among the greatest
amount of luminescence and bacterial burden (>9.5 × 106 CFU/g) of the
group. A second group of mice was dissected at the earliest sign of
luminescence, which corresponded to <3.5 × 106 CFU/g, to
determine how early in infection the bottleneck could be detected. In this
group, there were no significant differences between the initial inoculum and
the focal region in 3 out of 4 mice (Fig.
5B).
Most nonluminescent regions were significantly different from the focal region,
but no single subpopulation was greater than 42%. This suggests that slight
shifts in clonal proportions occur as spores transit from the nasal passage to
the lungs. Germination varied in the focal regions of these mice, with mouse 6
having only 38% germination, mouse 9 having 91% germination, and mouse 7 and
mouse 8 having >95% germination. To ensure that the clonality in the focal
region is not an artifact due to the intranasal delivery of the spores to the
lungs, the heated homogenates from the focal regions were analyzed. All clonal
populations were present in the focal regions as spores, but most mice showed
statistically significant shifts from the focal regions (Fig.
5C).
However, averaging the clonal proportions in all heated homogenates of the
focal regions showed no difference from the initial inoculum.
Lung-based
infections occur in distinct focal regions that are composed mainly of
vegetative bacilli. (A) Photograph of a dissected mouse lung with luminescent
focal region with a pseudocolor overlay representing luminescence intensity as
indicated on ...
The
luminescent focal region is the site of the bottleneck in lung-based B.
anthracis infections. (A and B) The percentages of clones in the focal
region were compared to the initial inoculum and a nonluminescent region
outside the focal region when there ...
The dominant clone in the lung focal
region is the founder of the disseminated population.
To test whether the dominant clone in
the focal region goes on to disseminate through the host, the clones in the
kidneys were compared to the clones in the focal region. A comparison between
the subpopulations in the focal regions and kidneys in five out of eight mice
did not show a significant difference, suggesting that the dominant population
in the lung went on to disseminate through the mouse (Fig.
6A)
(Fisher's exact test, P > 0.05). The focal regions in mice
6 and 7 had a significant difference compared to the kidneys, but these were
dissected early after luminescence was detected in the lungs. A bottleneck,
however, occurred between the focal region and the kidneys. Mouse 3 had two
populations comprise 76% of the focal region and the same two populations
comprise 98% of the kidney, but the two populations were found to be
significantly different from each other (Fig.
6A).
The MLN, however, had unique clonal proportions compared to those in the
kidneys and focal regions of five out of six mice, suggesting that these
populations did not contribute to the disseminated population (Fig.
6B).
Clonal
proportions in the lung focal regions resemble those in the kidney but not the
MLN. The percentages of clones in the focal region (A) or the MLN (B) were
compared to the percentages of clones in the kidneys. Each bar represents the
indicated organ ...
The lung-based bottleneck allows more
clones to pass through and has a lower effective population.
In order to determine if the lung-based
bottleneck allowed more clones to pass through the bottleneck, additional mice
were infected to increase the number of lung infections analyzed. The average
number of clones represented in the disseminated population was measured by
analyzing the average number of different clones in the kidneys from NALT-based
and lung-based infections. The average number of different clones in the kidneys
was significantly greater in lung-based infections than in NALT-based
infections (Fig.
7)
(Student's t test, P = 0.014). The effective
population (Ne) could then be calculated from the
number of clones in the disseminated population and used to compare the sizes
of the bottlenecks that the bacilli pass through from different portals of
entries. The effective population (Ne) size is an
idealized population that has the same rate of genetic drift as the actual
population size and allows for a quantitative comparison of bottlenecks between
portals of entry. One method of determining Ne is to analyze
the number of offspring between nonoverlapping generations. If there is a
decrease in population due to a bottleneck, then the effective population will
decrease in proportion to the stringency of the bottleneck. In a B.
anthracis infection, the generations occur between the inoculation
with spores and the death of the host. As such, one can calculate the effective
population by determining the number of clones in the disseminated population,
i.e., the kidney clonal population. The effective population was significantly
higher in lung-based infections (Ne, 1.65) than in NALT-based
infections (Ne, 0.72) (Student's t test, P =
0.014) (the Ne for no
bottleneck was 5).
Mice
with lung-based bottlenecks have a higher number of clones in the kidneys. The
mean numbers of different clones represented in the kidneys of mice with
NALT-based (closed circles; n = 12) and lung-based (closed
squares; n = 16) were compared. Each ...
DISCUSSION
Our work provides insight on the
ability of B. anthracis to colonize and disseminate from
unique microenvironments in the host respiratory tract by combining BLI
technology with clonal analysis. Utilization of the luminescent reporter under
the control of the PA promoter allowed for detection of the vegetative bacilli
early and throughout infection (15, 16). Potential
drawbacks to this reporter are that (i) the bacteria would not be detected if
there were areas where toxin production does not occur and (ii) there is a
slight delay of approximately 2 h between luminescent detection and germination
(16). All mice had B.
anthracis pass through a bottleneck regardless of where the infection
initiated. Clonal analysis of lung-based infections suggested that a bottleneck
occurred in the lumen of the lung, possibly as a few bacteria germinated in the
lungs. These bacteria then rapidly disseminate from the lungs to the
bloodstream and other organs. Further, the clone population in most MLN was
statistically different from those in the respective kidneys and focal growth
in the lung. NALT-based infections, however, passed through a bottleneck
between the NALT and the CLN. In contrast to the lung-based infections, the CLN
always had the same tagged clone as the NALT and kidneys. Analysis of the
kidneys from mice with lung-based infections showed a greater mean number of
clones and a higher effective population than in mice with NALT-based
infections. Overall, these data suggest that B. anthracis disseminates
differently from the two inhalational portals of entry and that some routes of
inhalational dissemination allow greater numbers of bacilli to disseminate; thus,
the portal of entry defines the characteristics of the subsequent disseminated
infection.
The experiments performed in this study
were done using the BSL-2 Sterne strain of B. anthracis in the
A/J mouse model, which has advantages and disadvantages. Since the Sterne
strain lacks the pX02 plasmid encoding the antiphagocytic capsule, the
experiments can be performed without the difficulties of biosafety level 3
(BSL3) facilities, providing a more economical model of study. Additionally,
several studies have noted key similarities between the A/J mouse model of
anthrax and larger animal models (13, 15, 37–39). These include a
similar dissemination pattern and a high level of bacteremia within the host (13, 15, 37). As such, mice are
useful models to analyze colonization and early steps in pathogenesis (37, 40). A disadvantage,
however, is the lack of capsule, which is thought to account for pathological
differences found during the bacteremic stages of infection. Acapsular strains
have limited splenic colonization and do not cause vasculitis and mediastinitis
(40). Therefore,
comparisons between the Sterne-A/J model and human anthrax must be considered
carefully. It is important to note that the focus of this study was to
investigate the dissemination of B. anthracis in inhalational
infections, not to model the effects of late-stage anthrax.
In our studies, A/J mice were given
spores by intranasal instillation to cause inhalational infections. A concern
with the intranasal model is effective delivery of spores to the lungs (38). In our hands,
mouse lungs would contain about 5% of the inoculum at the time of dissection.
However, when the clonal proportions were analyzed both within the focal region
and in the nonluminescent regions, all clonal populations were present, and
when averaged they were not significantly different from the inoculum. This
suggests that while our mice had fewer spores reach the lungs than previously
published numbers reported in studies using aerosol models, all clonal
identities were capable of reaching the focal regions (37). As such, the
bottlenecks detected in the focal region are unlikely to occur due to uneven
delivery of the spores to the lungs due to intranasal instillation.
The high spore burden relative to
vegetative bacteria within the lung from the initial inoculum was likely the
factor that obscured the detection of a bottleneck at early time points. If an
organ has a large and random population of spores, there is a greater chance of
sampling ungerminated spores of a dissimilar clonal subpopulation. While heat
treatment of homogenates can differentiate spores from total bacteria, there is
no current methodology for isolating only the vegetative bacilli for clonal
analysis, which, if possible, would permit earlier analysis of the bottleneck.
To circumvent the high spore burden in
the lungs, luminescent focal regions and nonluminescent regions were separated
to (i) attempt to isolate a region with a higher proportion of bacilli to
spores and (ii) determine the extent to which the bacilli can spread through
lung tissue without hematogenous dissemination. The luminescent focal areas
were found to comprise mainly vegetative bacilli, whereas the rest of the lung
tissue had the majority of bacteria in spore form (Fig.
4A and andB).B). While previous studies found that
germination and outgrowth occur in the lungs, it is interesting to note that
the bacilli do not migrate far from the focal region without prior systemic
dissemination through the host. This suggests that the site of lung germination
as previously reported by Sanz et al. may represent the earliest moments of the
focal infection formation in the lung (17).
The luminescent focal regions were
examined at various levels of bacterial burdens to determine when the
bottlenecks could be detected. When focal regions were analyzed with ≥3.5 × 106 CFU/g, there
were large increases in only 1 or 2 clones. However, a bottleneck could not be
detected when mice were dissected at the earliest observation of luminescence.
This suggests that detection of the lung bottleneck is dependent on both germination
and bacterial replication and that the disseminated population arises from a
small number of spores. Interestingly, bacilli could be found in the kidneys
before the bottleneck could be detected in the lungs; this suggests rapid
dissemination from the lungs after germination. Mice were also dissected when
there was a high level of luminescence (>9.5 × 106 CFU/g in the
focal regions) and were found to have increased amounts of the disseminated
population in the nonluminescent regions. This suggests that the nonluminescent
areas of the lung have the disseminated population return secondarily, rather
than primarily spreading through the lung.
In lung-based infections, bacterial
dissemination from the portal of entry to the draining lymph nodes is
sufficient but not necessary for dissemination. A comparison between the MLN
and the kidneys found that six out of 12 mice had the greatest clonal
proportion in the MLN in common with the kidney. Lethal toxin is also capable
of increasing epithelial and endothelial permeability, from which one could
speculate that bacilli are able to bypass the lymphatic routes in lung-based
infection and access the vasculature directly (41, 42). However, these
experiments alone do not provide sufficient evidence to definitively support
this hypothesis, since the earliest events in the lung-based infections cannot
be analyzed by these methods. It is also possible that the MLN comprise a
population of bacteria distinct from that in the focal region. If this is the
case, either these bacteria do not disseminate through the host or they
contribute to the disseminated population to a lesser extent, since the focal
regions with dominant clones do not statistically vary from those in the
kidneys. Interestingly, other routes of dissemination in murine anthrax, such
as subcutaneous and gastrointestinal infections, have also shown dissemination
through the lymph node to be required (16, 18). We speculate that
such a route could be possible, given that both exotoxin components, edema
toxin and lethal toxin, are capable of increasing vascular and alveolar
epithelial permeability (41, 42). However, future
studies will need to be performed in order to determine if and to what degree
the mouse lung epithelium becomes permeabilized. Epithelial layers have also
been identified as being capable of spore uptake and to allow transcytosis of
spores (43). Therefore, another
possibility is that spores are directly internalized by the lung epithelium to
then transit directly to the circulatory system and thereby bypass the lymph
nodes. In contrast to the case for mice, pulmonary infections of rabbits have
shown that germination occurs not in the lungs but in the lung-associated lymph
nodes (44). The difference in
germination and dissemination may be due to differences in the animal models,
method of spore instillation, and/or bacterial strains. Given the paucity of
human cases of inhalation anthrax and the difficulty of early detection of
infection, it is not clear how early events in these animal models resemble
anthrax infection in humans. Nonetheless, it is clear that the site of
germination and replication plays a role in how the bacteria disseminate.
Germination and outgrowth from the
lungs is not without controversy. The potential of lungs to act as primary
sites of infection was first suggested by preliminary work which found
bacterial replication and damage in the lung and bronchi (45, 46). It is worth noting
that several of these studies used experimental animals in which the lungs were
compromised by an unrelated previous infection or experimental manipulation (22, 47, 48). Similarly, a
majority of male victims in the Sverdlosk anthrax outbreak were moderate to
heavy smokers and their most common occupation was welder, factors which
increase the risk of lung damage and disease (49). Therefore,
previous damage to the lungs may skew infections to germinate and disseminate
from the lungs rather than to disseminate from the NALT or lymphatics.
Conversely, there is support for lymphatic dissemination from the lungs, since
large-scale germination does not occur in the lungs and there is no sign of
bronchopneumonia (9, 10, 14, 20, 37, 50). A central caveat
to previous pathological analyses and determination of bacterial burden is that
the infection was analyzed either at defined time points or after death and in
particular organs. Mice in the current study had similar bacterial burdens in
most organs at the time of dissection regardless of where the infection
initiated. This suggests that determining dissemination patterns by quantifying
bacterial loads in organs may not be conclusive by itself. In a number of
animal models, lung lesions have been observed upon postmortem analysis, but
there were no real-time assays to track the infection in the host (22, 47, 48, 51–55). Therefore, caution
should be used when concluding where an infection began, given that there is
still little known about the early events in pathogenesis in vivo or
how differences in delivery, host, and environment could alter the portal of
entry.
Our results provide a possible
reconciliation between conflicting reports on B. anthracis infections
that initiate and later disseminate from the lung and previous observations of
initiation and dissemination from the MLN. This is not to say that MLN-based
infections do not occur; rather, it is to argue that lung-based infections can
occur and may not be easily detected. Clonal analysis suggests that a small
number of spores germinate in a small, discrete location in the lung. The lungs
of these mice contained bacteria in spore form throughout most of the lungs,
except for the focal area where a small number of spores germinated and
replicated. Rather than spreading primarily through the lungs, the bacilli
rapidly spread either through the lymphatics or directly to the bloodstream.
Therefore, even in our mice with a lung infection, the lungs on the whole were
mostly refractory to germination, and overt lung damage was limited to a
portion of a lobe that could have easily gone undetected in past studies that
did not utilize BLI. Therefore, when the lungs were studied as a complete unit,
rather than focusing on an important subregion that contained the growth as we
have achieved, it would have been difficult to detect vegetative growth from
this small area above the background of CFU from ungerminated spores within the
rest of the lung tissues. Moreover, a postmortem dissection of the animal
without the BLI data would have greater difficulty discerning if the lung
lesions occurred prior to dissemination.
SUPPLEMENTARY
MATERIAL
Supplemental
material:
ACKNOWLEDGMENTS
We acknowledge Doug Taylor for
thoughtful discussions and comments on population genetics. Kevin Janes
assisted with concepts related to probability modeling. We also thank Heath
Damron, Zach P. Weiner, and Margot K. Williams for thoughtful discussion and
comments on the manuscript.
This work was supported by
grant 5R03AI090205 from the NIAID, and David Lowe was supported by
grant 5T35AI060528.
FOOTNOTES
Published
ahead of print 16
September 2013
REFERENCES
2. Ireland
JA, Hanna PC. 2002. Amino acid- and purine ribonucleoside-induced
germination of Bacillus anthracis DeltaSterne endospores: gerS mediates
responses to aromatic ring structures. J. Bacteriol.184:1296–1303 [PMC free article] [PubMed]
3. Titball
RW, Manchee RJ. 1987. Factors affecting the germination of spores of
Bacillus anthracis. J. Appl. Bacteriol. 62:269–273 [PubMed]
4. Corre
JP, Piris-Gimenez A, Moya-Nilges M, Jouvion G, Glomski IJ, Mock M, Sirard JC,
Goossens PL. 2013. In vivo germination of Bacillus anthracis spores during
murine cutaneous infection. J. Infect. Dis.207:450–457 [PubMed]
5. Green
BD, Battisti L, Koehler TM, Thorne CB, Ivins BE. 1985. Demonstration of a
capsule plasmid in Bacillus anthracis. Infect.
Immun. 49:291–297 [PMC free article] [PubMed]
6. Makino
S, Uchida I, Terakado N, Sasakawa C, Yoshikawa M. 1989. Molecular
characterization and protein analysis of the cap region, which is essential for
encapsulation in Bacillus anthracis. J. Bacteriol.171:722–730 [PMC free article] [PubMed]
7. Smith
H, Stanley JL. 1962. Purification of the third factor of anthrax
toxin. J. Gen. Microbiol. 29:517–521 [PubMed]
8. Stanley
JL, Sargeant K, Smith H. 1960. Purification of factors I and II of the
anthrax toxin produced in vivo. J. Gen. Microbiol. 22:206–218 [PubMed]
9. Barnes
JM. 1947. The development of anthrax following the administration of
spores by inhalation. Br. J. Exp. Pathol. 28:385–394
10. Ross
JM. 1957. The pathogenesis of anthrax following the administration of
spores by the respiratory route. J. Pathol. Bacteriol. 73:485–494
11. Guidi-Rontani
C. 2002. The alveolar macrophage: the Trojan horse of Bacillus
anthracis. Trends Microbiol. 10:405–409 [PubMed]
12. Lincoln
RE, Hodges DR, Klein F, Mahlandt BG, Jones WI, Jr, Haines BW, Rhian MA, Walker
JS. 1965. Role of the lymphatics in the pathogenesis of anthrax. J.
Infect. Dis. 115:481–494 [PubMed]
13. Loving
CL, Khurana T, Osorio M, Lee GM, Kelly VK, Stibitz S, Merkel TJ.
2009. Role of anthrax toxins in dissemination, disease progression, and
induction of protective adaptive immunity in the mouse aerosol challenge
model. Infect. Immun. 77:255–265 [PMC free article] [PubMed]
14. Swartz
MN. 2001. Recognition and management of anthrax—an update. N. Engl.
J. Med. 345:1621–1626 [PubMed]
15. Glomski
IJ, Dumetz F, Jouvion G, Huerre MR, Mock M, Goossens PL. 2008. Inhaled
non-capsulated Bacillus anthracis in A/J. mice: nasopharynx and alveolar space
as dual portals of entry, delayed dissemination, and specific organ
targeting. Microbes Infect. 10:1398–1404 [PubMed]
16. Glomski
IJ, Piris-Gimenez A, Huerre M, Mock M, Goossens PL. 2007. Primary
involvement of pharynx and peyer's patch in inhalational and intestinal
anthrax. PLoS Pathog.3:e76.10.1371/journal.ppat.0030076 [PMC free article] [PubMed] [Cross Ref]
17. Sanz
P, Teel LD, Alem F, Carvalho HM, Darnell SC, O'Brien AD. 2008. Detection
of Bacillus anthracis spore germination in vivo by bioluminescence
imaging. Infect. Immun. 76:1036–1047 [PMC free article][PubMed]
18. Weiner
ZP, Boyer AE, Gallegos-Candela M, Cardani AN, Barr JR, Glomski IJ.
2012. Debridement increases survival in a mouse model of subcutaneous
anthrax. PLoS One7:e30201.10.1371/journal.pone.0030201 [PMC free article] [PubMed] [Cross Ref]
19. Dumetz
F, Jouvion G, Khun H, Glomski IJ, Corre JP, Rougeaux C, Tang WJ, Mock M, Huerre
M, Goossens PL. 2011. Noninvasive imaging technologies reveal edema toxin
as a key virulence factor in anthrax. Am. J.
Pathol. 178:2523–2535 [PMC free article] [PubMed]
20. Guidi-Rontani
C, Weber-Levy M, Labruyere E, Mock M. 1999. Germination of Bacillus
anthracis spores within alveolar macrophages. Mol.
Microbiol. 31:9–17 [PubMed]
21. Cote
CK, Bozue J, Twenhafel N, Welkos SL. 2009. Effects of altering the
germination potential of Bacillus anthracis spores by exogenous means in a
mouse model. J. Med. Microbiol. 58:816–825[PubMed]
22. Gleiser
CA, Berdjis CC, Hartman HA, Gochenour WS. 1963. Pathology of experimental
respiratory anthrax in Macaca mulatta. Br. J. Exp.
Pathol. 44:416–426 [PMC free article] [PubMed]
23. Bakardjiev
AI, Stacy BA, Portnoy DA. 2005. Growth of Listeria monocytogenes in the
guinea pig placenta and role of cell-to-cell spread in fetal infection. J.
Infect. Dis. 191:1889–1897 [PubMed]
24. Bakardjiev
AI, Theriot JA, Portnoy DA. 2006. Listeria monocytogenes traffics from
maternal organs to the placenta and back. PLoS
Pathog. 2:e66.10.1371/journal.ppat.0020066 [PMC free article] [PubMed][Cross Ref]
25. Barnes
PD, Bergman MA, Mecsas J, Isberg RR. 2006. Yersinia pseudotuberculosis
disseminates directly from a replicating bacterial pool in the
intestine. J. Exp. Med. 203:1591–1601 [PMC free article][PubMed]
26. Plaut
RD, Kelly VK, Lee GM, Stibitz S, Merkel TJ. 2012. Dissemination bottleneck
in a murine model of inhalational anthrax. Infect.
Immun. 80:3189–3193 [PMC free article] [PubMed]
27. Schwartz
DJ, Chen SL, Hultgren SJ, Seed PC. 2011. Population dynamics and niche
distribution of uropathogenic Escherichia coli during acute and chronic urinary
tract infection. Infect. Immun. 79:4250–4259 [PMC free article] [PubMed]
28. Charlesworth
B. 2009. Fundamental concepts in genetics: effective population size and
patterns of molecular evolution and variation. Nat. Rev.
Genet. 10:195–205 [PubMed]
29. Sylvestre
P, Couture-Tosi E, Mock M. 2005. Contribution of ExsFA and ExsFB proteins
to the localization of BclA on the spore surface and to the stability of the
bacillus anthracis exosporium. J. Bacteriol. 187:5122–5128 [PMC free article] [PubMed]
30. Arraj
JA, Marinus MG. 1983. Phenotypic reversal in dam mutants of Escherichia
coli K-12 by a recombinant plasmid containing the dam+ gene. J.
Bacteriol. 153:562–565 [PMC free article] [PubMed]
31. Chaffin
DO, Beres SB, Yim HH, Rubens CE. 2000. The serotype of type Ia and III
group B streptococci is determined by the polymerase gene within the
polycistronic capsule operon. J. Bacteriol.182:4466–4477 [PMC free article] [PubMed]
32. Koehler
TM, Dai Z, Kaufman-Yarbray M. 1994. Regulation of the Bacillus anthracis
protective antigen gene: CO2 and a trans-acting element
activate transcription from one of two promoters. J. Bacteriol.176:586–595 [PMC free article] [PubMed]
33. Janes
BK, Stibitz S. 2006. Routine markerless gene replacement in Bacillus
anthracis. Infect. Immun.74:1949–1953 [PMC free article] [PubMed]
34. National
Research Council 2011. Guide for the care and use of laboratory animals,
8th ed. National Academies Press, Washington, DC
35. Cote
CK, Van Rooijen N, Welkos SL. 2006. Roles of macrophages and neutrophils
in the early host response to Bacillus anthracis spores in a mouse model of
infection. Infect. Immun. 74:469–480 [PMC free article] [PubMed]
36. Truett
GE, Heeger P, Mynatt RL, Truett AA, Walker JA, Warman ML.
2000. Preparation of PCR-quality mouse genomic DNA with hot sodium
hydroxide and tris (HotSHOT). Biotechniques 29:52–54[PubMed]
37. Loving
CL, Kennett M, Lee GM, Grippe VK, Merkel TJ. 2007. Murine aerosol
challenge model of anthrax. Infect. Immun. 75:2689–2698 [PMC free article] [PubMed]
38. Lyons
CR, Lovchik J, Hutt J, Lipscomb MF, Wang E, Heninger S, Berliba L, Garrison K.
2004. Murine model of pulmonary anthrax: kinetics of dissemination,
histopathology, and mouse strain susceptibility. Infect.
Immun. 72:4801–4809 [PMC free article] [PubMed]
39. Welkos
SL, Keener TJ, Gibbs PH. 1986. Differences in susceptibility of inbred
mice to Bacillus anthracis. Infect. Immun. 51:795–800 [PMC free article] [PubMed]
40. Twenhafel
NA. 2010. Pathology of inhalational anthrax animal models. Vet.
Pathol. 47:819–830[PubMed]
41. Langer
M, Duggan ES, Booth JL, Patel VI, Zander RA, Silasi-Mansat R, Ramani V, Veres
TZ, Prenzler F, Sewald K, Williams DM, Coggeshall KM, Awasthi S, Lupu F, Burian
D, Ballard JD, Braun A, Metcalf JP. 2012. Bacillus anthracis lethal toxin
reduces human alveolar epithelial barrier function. Infect. Immun.80:4374–4387 [PMC free article] [PubMed]
42. Warfel
JM, Steele AD, D'Agnillo F. 2005. Anthrax lethal toxin induces endothelial
barrier dysfunction. Am. J. Pathol. 166:1871–1881 [PMC free article] [PubMed]
43. Russell
BH, Vasan R, Keene DR, Koehler TM, Xu Y. 2008. Potential dissemination of
Bacillus anthracis utilizing human lung epithelial cells. Cell.
Microbiol. 10:945–957 [PubMed]
44. Lovchik
J, Drysdale M, Koehler TM, Hutt JA, Lyons CR. 2012. Expression of either
lethal toxin or edema toxin by Bacillus anthracis is sufficient for virulence
in a rabbit model of inhalational anthrax. Infect.
Immun. 80:2414–2425 [PMC free article] [PubMed]
45. Buchner
H. 1888. Special conditions for the penetration of infective pathogens
through the intact pulmonary surface. Arch. Hyg. 8:217–245
46. Fraenkel
E. 1925. Inhalation anthrax. Virchow's Archiv. 254:363–378
47. Velu
H, Soulie P, Bellocq B. 1943. Pulmonary anthrax and chlorine poisoning:
minimum infectious dose of spores. C. R. Soc. Biol. 137:160–161
48. Velu
H, Soulie P, Bellocq B. 1943. Experimental pulmonary anthrax of mice:
initial bacteremia or septicemia. C. R. Soc. Biol. 137:159–160
49. Meselson
M, Guillemin J, Hugh-Jones M, Langmuir A, Popova I, Shelokov A, Yampolskaya O.
1994.The Sverdlovsk anthrax outbreak of 1979. Science 266:1202–1208 [PubMed]
50. Young
GA, Jr, Zelle MR, Lincoln RE. 1946. Respiratory pathogenicity of Bacillus
anthracis spores; methods of study and observations on pathogenesis. J.
Infect. Dis. 79:233–246 [PubMed]
51. Twenhafel
NA, Leffel E, Pitt ML. 2007. Pathology of inhalational anthrax infection
in the African green monkey. Vet. Pathol. 44:716–721 [PubMed]
52. Gleiser
CA, Gochenour WS, Jr, Ward MK. 1968. Pulmonary lesions in dogs and pigs
exposed to a cloud of anthrax spores. J. Comp.
Pathol. 78:445–448 [PubMed]
53. Zaucha
GM, Pitt LM, Estep J, Ivins BE, Friedlander AM. 1998. The pathology of
experimental anthrax in rabbits exposed by inhalation and subcutaneous
inoculation. Arch. Pathol. Lab. Med. 122:982–992[PubMed]
54. Fritz
DL, Jaax NK, Lawrence WB, Davis KJ, Pitt ML, Ezzell JW, Friedlander AM.
1995. Pathology of experimental inhalation anthrax in the rhesus
monkey. Lab. Invest. 73:691–702 [PubMed]
55. Abramova
FA, Grinberg LM, Yampolskaya OV, Walker DH. 1993. Pathology of
inhalational anthrax in 42 cases from the Sverdlovsk outbreak of
1979. Proc. Natl. Acad. Sci. U. S. A. 90:2291–2294 [PMC free article] [PubMed]
No comments:
Post a Comment