Microbes and Infection 10 (2008) 613e619www.elsevier.com/locate/micinf

Original article

Early interactions between fully virulent Bacillus anthracis andmacrophages that influence the balance between spore

clearance and development of a lethal infection

Christopher K. Cote a, Tracy L. DiMezzo a, David J. Banks b, Bryan France b,Kenneth A. Bradley b, Susan L. Welkos a,*

a Bacteriology Division, United States Army Medical Research Institute of Infectious Diseases (USAMRIID), 1425 Porter Street,

Fort Detrick, Frederick, MD 21702, USAb University of California, Department of Microbiology, Immunology, and Molecular Genetics, 609 Charles E. Young Dr. East,

MSB Room 2825, Los Angeles, CA 90095, USA

Received 29 November 2007; accepted 12 February 2008

Available online 21 February 2008

Abstract

The role of macrophages in the pathogenesis of anthrax is unresolved. Macrophages are believed to support the initiation of infection byBacillus anthracis spores, yet are also sporicidal. Furthermore, it is believed that the anthrax toxins suppress normal macrophage function. How-ever, the significance of toxin effects on macrophages has not been addressed in an in vivo infection model. We used mutant derivatives of mu-rine macrophage RAW264.7 cells that are toxin receptor-negative (R3D) to test the role of toxin-targeting of macrophages during a challengewith spores of the Ames strain of B. anthracis in both in vivo and in vitro models. We found that the R3D cells were able to control challengewith Ames when mice were inoculated with the cells prior to spore challenge. These findings were confirmed in vitro by high dose sporeinfection of macrophages. Interestingly, whereas the R3D cells provided a significantly greater survival advantage against spores than didthe wild type RAW264.7 cells or R3D-complemented cells, the protection afforded the mutant and wild type cells was equivalent against a ba-cillus challenge. The findings appear to be the first specific test of the role of toxin targeting of macrophages during infection with B. anthracisspores.� 2008 Elsevier Masson SAS. All rights reserved.

Keywords: Macrophages; Spores; Bacillus anthracis; Anthrax toxin receptor; Mice

1. Introduction

B. anthracis, a gram-positive, spore-forming bacterium, isthe causative agent of anthrax [1]. Inhalational anthrax, themost severe form of infection, has a high mortality rate andrapid course. It is hypothesized that after inhalation of the in-fectious dormant spores of B. anthracis, the spores are en-gulfed by alveolar macrophages that subsequently migrate tothe regional lymph nodes [2,3]. There, the spores may germi-nate within the phagocytes [3e9] and form vegetative cells

* Corresponding author. Tel.: þ1 301 619 4930; fax: þ1 301 619 2152.

E-mail address: susan.welkos@amedd.army.mil (S.L. Welkos).

1286-4579/$ - see front matter � 2008 Elsevier Masson SAS. All rights reserved.

doi:10.1016/j.micinf.2008.02.006

that are released, multiply, spread systemically, and producetoxins. The anthrax edema and lethal toxins (ET, LT) are re-sponsible for several deleterious effects that ultimately leadto a fatal outcome [10e15].

Several key aspects of the early events in infection remaincontroversial. For instance, in contrast to the evidence for a rel-atively passive role of the macrophages in anthrax pathogene-sis, it was shown that macrophages can also kill spores afterthey germinate intracellularly [8,16,17]. This sporicidal activ-ity is generally studied in vitro at fairly low multiplicities ofinfection (MOI) of �10. It seems feasible that when exposedto a large multiplicity of spores, although the macrophage killsmost of the germinated organisms, a low percentage survivesthe antimicrobial environment and ultimately escapes the

614 C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

macrophage to begin the extracellular infection [18]. Mecha-nisms utilized by B. anthracis to escape from the macrophageare not well characterized, but a role for anthrax toxin hasbeen proposed and recently shown [19]. Following phagocyto-sis, B. anthracis spores appear to traffic to phagolysosomesthat display membrane-bound anthrax toxin receptor(ANTXR) [19e22]. Co-localization of receptor and germinat-ing, toxin-expressing spores would allow for binding of thetoxin and translocation of the toxin directly from the phagoly-sosome into the cytosol. In addition to the unresolved nature ofthe macrophage e spore interactions, recent studies haveshown that macrophages are not absolutely required to estab-lish infection by B. anthracis, as demonstrated with the Amesstrain in mouse peritoneal and inhalational models of anthrax[23,24]. In this paper, we examined further the interactions be-tween B. anthracis and macrophages early in the infection thatdetermine the balance between spore clearance and develop-ment of a lethal infection.

2. Materials and methods

2.1. In vitro spore challenge assays

Macrophages were cultured in 24-well trays with coverslipsfor in vitro phagocytosis and macrophage viability assays thatwere performed using a modification of previously described

Phagocytosis of Ames spores by macrophages as deter

immunofluorescence microscopy

Infection Total no. No. phagocytosedCells MOI CDa MΦs counted spores/MΦb

RAW 62.5 + 162 0.0562.5 121 4.51

0 - 300 0R3D-c 43.5 168 0.04

43.5 141 5.84300 0

R3D 67.8 260 0.0267.8 73 10.01

-

+-

0 -+-

0 - 300 0aCytochalasin D (CD) was included during phagocytosis period, (+) or nbMean number intracellular spores as determined by counting spores la with FITC-conjugated goat anti-rabbit secondary Ab in permeabilized McMean total no. extracellular and intracellular spores labeled by TRITC-conjugated secondary Abs before, and FITC-conjugated Abs after,permeabilization, respectively. The mean nos. phagocytosed spores/Mthe MΦs infected in the presence or absence of CD were not significan different for all three cell lines (p >0.13, determined by ANOVA and t-te

Fig. 1. Phagocytosis of Ames spores by macrophages as determined by immunofluo

representative of five, are shown. Macrophages were incubated with Ames spores

lustrate the appearance of wild and toxin receptor mutant macrophages infected wit

and TRITC-labeled Abs, respectively, as described in the text and previously [17].

against whole killed spores of Ames that recognized ungerminated and germinated

ondary IgG Abs conjugated to red (TRITC) or green (FITC) fluorescent dyes and wh

A e RAW264.7 cells infected with Ames spores in the presence of CD; panel B

methods [8,9,19]. The macrophage cell line RAW264.7, a lethaltoxin (LT)-resistant mutant of it that is unable to express theANTXR2/CMG-2 anthrax toxin receptor 2 (R3D) [21,22], anda variant of the mutant that was genetically complementedwith the receptor (R3D-c) were infected with heat-activated, un-germinated spores of the Ames strain of Bacillus anthracis pre-pared as described previously [8,9,19]. The spores were addedat a concentration of 50e100 cfu per cell in supplemented Dul-becco’s minimum essential medium with 10% fetal bovine se-rum (DF10) [8,9]. The infected culture trays were centrifugedat 1150 rpm (300 � g) for 30 min at 30 �C; and then incubated30 min at 37 �C in 5% CO2 [19]. After the wells were washed10 times in phosphate buffered saline (PBS), the coverslipswere removed and washed again in a set of beakers [8,9]. Thecoverslips were transferred to DF10 containing 5 mg/ml cytocha-lasin D and 2.5 mg/ml gentamicin and incubated at 37 �C for30 min. The wells were then washed five times with PBS andDF10 with 5 mg/ml cytochalasin D and anti-PA antibody at a finalconcentration of 10 mg/ml was added per well. They were incu-bated at 37 �C for 4 h, washed and given fresh medium, andreincubated for a total of 8 h, accompanied by hourly washes[19]. These procedures minimize residual unphagocytosedspores that could germinate and produce toxin and/or infect thecells. At the end of the 8 h incubation, the coverslips were pro-cessed to determine macrophage viability as assessed by propi-dium iodide (PI) staining [19]. Additionally the extent of spore

mined by

Total no.spore/MΦc

0.085.04

00.086.13

00.07

10.040

ot (-).

beledΦs.

Φ fortlysts).

A

B

rescence microscopy. Numerical results and micrographs from one experiment,

at the multiplicity of infection (MOI) shown in the table. The micrographs il-

h Ames spores as determined by inside/outside staining with FITC-labeled Abs

Specifically the MF samples were incubated with a rabbit antibody prepared

spores as well as bacilli. They were then incubated with goat anti-rabbit sec-

ich were added either before or after MF permeabilization, respectively. Panel

e R3D cells infected with Ames spores in the absence of CD.

The viability of wild type and toxin receptor mutant macrophages

infected with B. anthracis Ames spores

CellsNo. fields Total no. PI+ Mean no. PI+

Treatmenta counted macrophages macrophages/fieldb

RAW +/+ 15 130 8.7 (1.0)c

RAW +/- 7 652 93.1 (22)RAW -/- 13 74 5.7 (0.6)R3D-c +/+ 15 62 4.1 (0.4)R3D-c +/- 6 764 127.3 (30.3)R3D-c -/- 15 98 6.5 (1.8)R3D +/+ 15 126 8.4 (1.4)R3D +/- 15 68 4.5 (0.8)R3D -/- 20 154 7.7 (1.2)

a+/+, MΦs infected in the presence of CD; +/- MΦs infected in the absence ofCD; -/-, MΦs mock-infected in the absence of CD.bThe number of PI+ cells/microscopic field, as determined by counting allPI+ spores in fields that were uniformly approximately 80% confluent.cMean (SEM) shown.The mean no. PI+ cells/field for cells infected in the absence of CD was significantly greater than that for cells infected in thepresence of CD for the RAW and R3D-c MΦs (p = <0.01) but not for R3D MΦs.

A B

C D

Fig. 2. The viability of wild type and toxin receptor mutant macrophages infected with Ames spores. Macrophages were incubated with Ames spores at the mul-

tiplicity of infection (MOI) indicated in the chart in Fig. 2. The number of PIþ cells/microscopic field are shown, as determined by counting all PIþ macrophages in

fields that were uniformly approximately 80% confluent for all three cell lines. Counting was done using the 20� objective (table). The micrographs illustrate the

appearance of wild and toxin receptor mutant macrophages infected with Ames spores as determined by staining with PI. Panel A: RAW264.7, þ/�; panel B:

RAW264.7 (or R3D or R3D-c), þ/þ; panel C: R3D, þ/�; and panel D: R3D-c, þ/�. The data from one experiment, representative of five, are shown.

615C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

616 C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

phagocytosis and of residual unphagocytosed organisms weredetermined microscopically by inside/outside double-label fluo-rescent immunostaining, as previously described [8,9,17].

2.2. In vivo spore challenge assays

Female BALB/c mice (NCI, Frederick, MD) were inocu-lated with lethal doses of B. anthracis strain Ames spores afterreceiving a single intraperitoneal (i.p.) dose of wild type ormutant macrophages, as described previously [19,23,24] andshown (Fig. 3). 1 LD50 equivalent for BALB/c mice usingAmes strain spores is approximately 500 spores injected i.p.[25]. In some experiments, mice were challenged with bacilli.To prepare the bacillus challenge, 10 ml of L-broth were inoc-ulated with approximately 4e5 � 106 spores and incubatedwith slow shaking (100 rpm) for 2.5 h at 37 �C to induce ger-mination. The culture was centrifuged, washed in PBS, diluted1:10 in PBS, and 200 ml volumes were injected i.p. into mice.The resulting dose was approximately 4900 cfu (all heat-sensitive), as determined by plate count; each cfu representeda mean of 4.5 � 2.5 (SD) bacilli per chain. Time course exper-iments to determine the kinetics of infection with spores wereperformed as described previously [24].

2.3. Statistics

Survival rates were compared between each treatmentgroup and control group by Fisher exact tests with permutationadjustment for multiple comparisons. KaplaneMeier/product-limit estimation was used to construct survival curves and tocompute mean survival times. Survival curves were comparedbetween each treatment group and control group by log ranktests with Hochberg adjustment for multiple comparisons.Mean times-to-death were compared between each treatmentgroup and control group by t-tests with permutation adjust-ment for multiple comparisons. The above analyses were con-ducted using SAS Version 8.2 (SAS Institute Inc., SASOnlineDoc, Version 8, Cary, NC 2000).

3. Results

3.1. R3D mutant macrophages are protected from aninfection with fully virulent B. anthracis Ames strainspores in vitro

The R3D mutant derivative of RAW264.7 macrophages isboth deficient in expression of a receptor for the PA componentof anthrax toxin and resistant to killing by LT. Banks et al. [19]showed that although these mutant cells could phagocytosespores of B. anthracis vaccine strain Sterne, they survivedhigh dose spore challenge, unlike the wild type macrophages.Thus it was proposed that spores use the macrophage-encodedanthrax toxin receptors to promote toxin-mediated killing ofthe macrophage from inside the cell. In this study, we providedevidence to support this hypothesis and also the concept thatmacrophages play a dual role in the pathogenesis of and protec-tion against anthrax. The results of fluorescent antibody staining

to determine the extent of spore phagocytosis by the cells as wellthe internal or external location of the spores, indicated that theparent, mutant, and complemented mutant cells were all simi-larly capable of phagocytosing Ames spores, as shown inFig. 1. The near absence of fluorescent antibody-labeledextracellular spores (red/orange spores in panel B of Fig. 1) con-firmed the efficiency of the washing procedure in removing un-phagocytosed spores. This conclusion was reiterated by the lowspore burden of macrophages treated with cytochalasin D (CD)before spore exposure (Fig. 1 chart). Thus, any differences in theviability of infected mutant and wild type macrophages wouldreflect activity of the intracellular organisms and not reducedmacrophage phagocytic activity. Also, spore stains of infectedmacrophages confirmed that spores germinated in R3D cellsas well as they did in the RAW264.7 cells; and thus LT-resis-tance of the mutant does not result from an inability of the sporesto germinate intracellularly ([19] and data not shown).

Both the wild type and R3D macrophages complementedwith the gene encoding the CMG2 receptor (R3D-c) exhibitedextensive staining with PI and loss of viability after in vitrospore challenge (Fig. 2, panels A and D). In contrast, the in-fected, LT-resistant R3D macrophage samples were not killedto an extent that was appreciably greater than that of the unin-fected cells or cells infected in the presence of CD (Fig. 2, ta-ble and panels B and C). Previously, it has been reported thatthe anthrax toxin genes encoding PA and LF are expressedvery early during spore germination in vitro. LF can be de-tected early after germination [5] and PA can be detected onthe germinating spore surface but is then rapidly released[17]. These findings support the proposed model of macro-phage susceptibility to killing by toxin produced within thephagosome by germinating B. anthracis spores [19].

3.2. The impact of exogenous macrophages on a mouseinfection model

Mice that have been depleted of macrophages are markedlyincreased in susceptibility to aerosol and parenteral challengewith the virulent Ames strain of B. anthracis [23,24], a findingthat suggests macrophages are not required for in vivo germina-tion of B. anthracis and that macrophages play a significant pro-tective role in anthrax pathogenesis. Resistance of mice toinfection with Ames can be increased by supplementing the an-imal’s native macrophage population with exogenousRAW264.7 macrophages [23,24]. In the current study, to fur-ther examine the role of macrophages in anthrax pathogenesisand host response, we evaluated the effect of treating micewith toxin-resistant mutant macrophages on their resistance toa spore challenge. Mice were supplemented with wild typeRAW264.7 or mutant macrophages and exposed to a lethali.p. challenge with Ames spores. There were no significant dif-ferences observed in the mean time-to-death (TTD) of any ofthe groups; however, the mice augmented with the R3D cellswere statistically better protected than were the animals that re-ceived the RAW264.7 cells when survival curves were exam-ined (Fig. 3A). Survival analysis, which takes into accountthe kinetics of infection and shape of the survival curves,

0

1

2

3

4

5

6

7

8

9

10

0 2 4 6 8 10 12 14Days

# su

rvivin

g

0

1

2

3

4

5

6

7

8

9

10

120 2 4 6 8 10 14Days

# su

rvivin

g

A

B

Fig. 3. Anthrax toxin receptor mutant cells protect mice from spore challenge

better than do wild type or complemented mutant macrophages. A) Native

peritoneal macrophages were augmented with approximately 1 � 107 of either

RAW264.7 (-) or R3D (,) cells (determined to be 95% and 94% viable, re-

spectively), or with the HBSS buffer alone (:). Mice (n ¼ 10) were then

challenged i.p. with approximately 2200 B. anthracis Ames strain spores

and observed for morbidity and mortality for 14 days. B) Native peritoneal

macrophages were augmented with approximately 1 � 107 of either R3D-c

(C) or R3D (,) cells (determined to be 86% and 84% viable, respectively),

or with the HBSS buffer alone (:). Mice (n ¼ 10) were then challenged i.p.

with approximately 2200 B. anthracis Ames strain spores and observed for

morbidity and mortality for 14 days. The results of one experiment, represen-

tative of three for figure A and two for figure B are shown.

Table 1

The dose-related protection of mice by R3D macrophages

Treatment* No. survivors/total no. (%) P value** MST*** P value**

HBSS 3/10 (30) e 6.1 e

R3D, 104 5/10 (50) 0.32 9.2 0.21

R3D, 105 8/10 (80) 0.07 12.1 0.03

R3D, 106 9/10 (90) 0.03 13.4 0.01

*Dose e no. cells delivered i.p. before spore challenge (2400 spores i.p.). **In

comparison to HBSS-treated group. ***Mean Survival Time, days.

617C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

revealed that the mice receiving the R3D cells survived signif-icantly better than the RAW264.7-supplemented mice(P ¼ 0.041) and the HBSS control mice (P ¼ 0.0005). Also,the R3D macrophage-treated mice had a significantly highersurvival rate as compared to the HBSS-treated controls(P ¼ 0.0036). Finally, when survival curves were analyzed,the mice receiving the RAW264.7 cells were more resistant tothe spore challenge than the HBSS control mice (P ¼ 0.0180).

In addition, the effect of R3D-complemented cells was alsoexamined in vivo. These cells were less protective (P ¼ 0.1)

for the mice than the R3D cells (P ¼ 0.0007 for survivalrate and 0.0015 for survival curve) when compared to HBSStreatment (Fig. 3B). The results obtained in this experimentwere similar to those seen when wild type RAW 264.7 cellswere used (Fig. 3A), suggesting that the complementation ofthe anthrax toxin receptor mutation resulted in a phenotypesimilar to the wild type cells.

Dose-response experiments were done to further examinethe role of toxin-resistant R3D cells in protection of mice. Al-though doses of R3D cells exhibited an apparent dose-responsein protection of mice against lethal infection with spores (with2400 cfu), only the group pretreated with the highest dose, 106

cells, showed a significantly greater rate of survival (90%) com-pared to the buffer-pretreated control mice (30% survival),P ¼ 0.03 (Table 1). The groups given 105 and 106 cells, butnot 104 cells, exhibited significantly longer mean survival timescompared to the control group (P ¼ 0.03 and 0.01, respectively)(Table 1). The results of logistic regression analysis of the effectof log10 dose of R3D cells on probability of survival were sug-gestive but not conclusive of a relationship between dose andsurvival outcome (P ¼ 0.0577).

Time course experiments were performed to determine theimpact of macrophage augmentation on bacterial survival andreplication. Mice were pretreated with wild type RAW264.7cells, R3D cells, or buffer alone (HBSS) and then challengedwith a high dose of spores (1.7 � 106 cfu, approximately 3400LD50 equivalents). Differences in response to infection wereobserved at later times post-inoculation (as described below).These mice had a shorter course of disease (predicted to suc-cumb by 24 h) and at various times after inoculation (0, 5, and18 h) peritoneal lavage fluids were collected and plated to de-termine the recovery of viable B. anthracis. The mean dose ofbacteria recovered from mice immediately after inoculation(t0) was comparable for all three groups (P > 0.05). The con-trol HBSS-treated mice had 8.0 � 104 cfu/ml lavage fluid, SD4.8 � 103 (n ¼ 4), the mice pretreated with RAW264.7 cellshad 9.3 � 104 cfu/ml, SD 2.2 � 104 (n ¼ 4) in the collectedperitoneal samples, and the R3D-pretreated mice yielded9.0 � 104 cfu/ml, SD 2.4 � 104 (n ¼ 4) at t0. The results ofthe 5 h samples suggested that, while there was mouse-to-mouse variability, supplementation of the mice with exoge-nous macrophages was not associated with a more extensivedecline in bacterial recovery in the initial hours post-challengecompared to mice with only native macrophage populations(data not shown). However, by 18 h, it appeared the mice re-ceiving the R3D cells were able to control the infectionmore efficiently than the control mice (Fig. 4), which was

0

1.0×104

2.0×104

3.0×104

4.0×104

5.0×104

6.0×104

HBSS RAW264.7 R3D

CF

Us/m

l o

f p

erito

neal lavag

e flu

id

18h

p

ost-in

fectio

n

Fig. 4. The kinetics of infection in mice pretreated with exogenous macro-

phages. Mice pretreated by i.p. injection with wild type RAW264.7 cells,

R3D cells, or buffer alone (HBSS) were challenged i.p. with spores. At inter-

vals post-inoculation, peritoneal lavage fluids were collected from several

mice and cultured quantitatively for total viable counts. Shown are the com-

bined 18 h sample culture results of two experiments. Three mice from each

treatment group were cultured in experiment 1 and 4 mice per treatment group

were cultured in experiment 2. In experiment 1, the mice received 5 � 105

cells per mouse (R3D or RAW264.7 macrophages determined to be 91%

and 94% viable, respectively) and were challenged with 1.7 � 106 spores. In

experiment #2, the mice received 9 � 105 cells (R3D or RAW264.7 macro-

phages both determined to be 94% viable) and were challenged with

1.4 � 106 spores. Most of the samples having 6.0 � 104 per ml actually had

a higher concentration of bacteria as 6 � 104 per ml was the highest concen-

tration that could be counted. Peritoneal bacterial burdens for each of seven

mice are shown.

0

1

2

3

4

5

6

7

8

9

10

0 2 4 6 8 10 12 14Days

# S

urvivin

g

Fig. 5. Anthrax toxin receptor mutant cells do not confer a survival advantage

compared to wild type macrophages in mice challenged with B. anthracis

bacilli. Native peritoneal macrophages were augmented with approximately

5 � 105 of either RAW264.7 (-) or R3D (,) cells (determined to be 94%

and 91% viable, respectively), or with the HBSS buffer alone (:). Mice

(n ¼ 10) were then challenged i.p. with approximately 4900 B. anthracis

Ames strain vegetative cfu and observed for morbidity and mortality for

14 days.

618 C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

associated with their significant survival advantage comparedto RAW264.7- or R3Dc-supplemented mice (Fig. 3, panelsA and B, respectively).

To examine possible differential effects of the presence oftoxin-resistant macrophages on the rate and extent of survivalin response to spore compared to vegetative cell challenge,mice were challenged with early germinated vegetative bacilli(approximately 4900 cfu). The kinetics of infection in thesemice challenged with bacilli supported the data above that sug-gest that macrophages have a protective role against lethal infec-tion. Whereas 100% of the control mice injected with HBSSbuffer alone succumbed by day 3 post-inoculation, 30% of theR3D- and RAW264.7-supplemented mice survived the durationof the two week experiment. These differences were not statis-tically significant (P > 0.05). In addition, the data did not distin-guish any enhanced protection afforded by toxin-resistantcompared to wild type RAW264.7 macrophages (Fig. 5), a resultthat perhaps suggests a more important role for the macrophagetoxin receptor in protection against toxin from ungerminatedspores than from germinated bacilli. Conceivably, the presenceof newly-formed anti-phagocytic capsule on the bacilli mightsupplant the importance of targeting the toxin receptor in pro-tecting spores against antimicrobial macrophage activities. Itis not known whether the vegetative organisms proliferatedand produced the toxins extracellularly or were phagocytosed.

4. Discussion

As proposed previously [8] and suggested by the data fromdifferent laboratories, macrophages appear to be capable of

both impeding or clearing the infection by B. anthracis sporesand facilitating intracellular survival and release of the germi-nated organisms [5e9,16e19,23,24]. The particular macro-phage function manifested at a given time is likelydependent on the macrophage spore burden, but also on fac-tors such as virulence of the challenge strain, resistance ofthe host, route of infection, and source of the respondingmacrophages.

The importance of anthrax toxin activities on specific targetcell types during infection has not previously been reported.The results of in vitro studies suggest that anthrax lethal andedema toxins can suppress functions of a variety of immunecells, including macrophages, dendritic cells, neutrophils,and T- and B-lymphocytes (reviewed by Banks et al. [26]and Baldari et al. [27]). Evidence resulting from these studiessupports a role for anthrax toxins in inhibiting cytokine pro-duction in a number of cell types [14,15,19,25e29]. Despitethese findings, increased cytokine levels are detected duringin vivo infections with spores using a murine model [26,30].This apparent discrepancy between the effects of spores andthe toxins on the cellular inflammatory responses might be ex-plained by the different times during the infection when theactivating and inhibitory effects of spores and the toxins, re-spectively, are exhibited [27,28].

Previously, it was demonstrated that macrophages are ableto limit B. anthracis infection and thus are an important com-ponent of the immune response to this pathogen [23,24]. Here,we test the model that anthrax lethal and/or edema toxin target

619C.K. Cote et al. / Microbes and Infection 10 (2008) 613e619

macrophages during infection, thereby crippling the ability ofthese cells to respond to spores and allowing for bacterial out-growth. Using mutant R3D macrophages that are insensitive toboth lethal and edema toxins we demonstrate that toxin target-ing of macrophages is an important function of anthrax toxins.Importantly, these experiments were performed using mice inwhich all other cell types retain their native level of toxin-sensitivity. Thus, increased resistance to B. anthracis sporechallenge in mice supplemented with R3D compared toR3D-c or wild type RAW264.7 cells (Fig. 3) is due to macro-phage-specific toxin effects. Our results support the conclusionthat anthrax toxin targeting of macrophages by the early ger-minating spore represents an important virulence mechanismfor B. anthracis.

Acknowledgements

We thank Carol Chapman for her invaluable technical assis-tance and Sarah Norris for her expert statistical analysis.

Opinions, interpretations, conclusions, and recommenda-tions are those of the authors and are not necessarily endorsedby the U.S. Army.

Research was conducted in compliance with the Animal Wel-fare Act and other federal statutes and regulations relating to an-imals and experiments involving animals and adheres to theprinciples stated in the Guide for the Care and Use of LaboratoryAnimals, National Research Council, 1996. The facility wherethis research was conducted is fully accredited by the Associa-tion for Assessment and Accreditation of Laboratory Animal.

The research described herein was sponsored by the Medi-cal Biological Defense Research Program, JSTO-CBD/DTRAprojects A.1X001-04-RDB and 1.1A0010-07-RDB (S.L.W.),and by National Institutes of Health (NIH) award AI-057870(K.A.B.).

References

[1] M. Mock, A. Fouet, Anthrax. Annu. Rev. Microbiol. 55 (2001) 647e671.

[2] J.M. Barnes, The development of anthrax following the administration of

spores by inhalation, Brit. J. Exp. Path 28 (1947) 385e393.

[3] J.M. Ross, The pathogenesis of anthrax following the administration of

spores by the respiratory route, J. Path. Bacteriol. 73 (1957) 485e494.

[4] T.C. Dixon, A.A. Fadl, T.M. Koehler, J.A. Swanson, P.C. Hanna, Early

Bacillus anthracis-macrophage interactions: intracellular survival and

escape, Cell. Microbiol. 2 (2000) 453e463.

[5] C. Guidi-Rontani, M. Weber-Levy, E. Labruyere, M. Mock, Germination

of Bacillus anthracis spores within alveolar macrophages, Mol. Micro-

biol. 31 (1999) 9e17.

[6] C. Guidi-Rontani, M. Weber-Levy, E. Labruyere, M. Mock, Fate of

germinated Bacillus anthracis spores in primary murine macrophages,

Mol. Microbiol. 42 (2001) 931e938.

[7] P.C. Hanna, J.A. Ireland, Understanding Bacillus anthracis pathogenesis,

Trends Microbiol. 7 (1999) 180e182.

[8] S. Welkos, A. Friedlander, S. Weeks, S. Little, I. Mendelson, In-vitro

characterisation of the phagocytosis and fate of anthrax spores in

macrophages and the effects of anti-PA antibody, J. Med. Microbiol.

51 (2002) 821e831.

[9] S. Welkos, S. Little, A. Friedlander, D. Fritz, P. Fellows, The role of

antibodies to Bacillus anthracis and anthrax toxin components in

inhibiting the early stages of infection by anthrax spores, Microbiology

147 (2001) 1677e1685.

[10] P. Hanna, Lethal toxin actions and their consequences, J. Appl.

Microbiol. 87 (1999) 285e287.

[11] P.C. Hanna, D. Acosta, R.J. Collier, On the role of macrophages in

anthrax, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 10198e10201.

[12] J.E. Kirby, Anthrax lethal toxin induces human endothelial cell

apoptosis, Infect. Immun. 72 (2004) 430e439.

[13] J. O’Brien, A. Friedlander, T. Dreier, J. Ezzell, S. Leppla, Effects of

anthrax toxin components on human neutrophils, Infect. Immun. 47

(1985) 306e310.

[14] S.G. Popov, R. Villasmil, J. Bernardi, E. Grene, J. Cardwell, A. Wu,

D. Alibek, C. Bailey, K. Alibek, Lethal toxin of Bacillus anthracis causes

apoptosis of macrophages, Biochem. Biophys. Res. Commun. 293 (2002)

349e355.

[15] S.G. Popov, R. Villasmil, J. Bernardi, E. Grene, J. Cardwell, T. Popova,

A. Wu, D. Alibek, C. Bailey, K. Alibek, Effect of Bacillus anthracis

lethal toxin on human peripheral blood mononuclear cells, FEBS Lett.

527 (2002) 211e215.

[16] J.A. Bozue, N. Parthasarathy, L.R. Phillips, C.K. Cote, P.F. Fellows,

I. Mendelson, A. Shafferman, A.M. Friedlander, Construction of a rham-

nose mutation in Bacillus anthracis affects adherence to macrophages but

not virulence in guinea pigs, Microb. Pathog. 38 (2005) 1e12.

[17] C.K. Cote, C.A. Rossi, A.S. Kang, P.R. Morrow, J.S. Lee, S.L. Welkos,

The detection of protective antigen (PA) associated with spores of Bacil-lus anthracis and the effects of anti-PA antibodies on spore germination

and macrophage interactions, Microb. Pathog. 38 (2005) 209e225.

[18] W.J. Ribot, R.G. Panchal, K.C. Brittingham, G. Ruthel, T.A. Kenny,

D. Lane, B. Curry, T.A. Hoover, A.M. Friedlander, S. Bavari, Anthrax le-

thal toxin impairs innate immune functions of alveolar macrophages and

facilitates Bacillus anthracis survival, Infect. Immun. 74 (2006) 29e5034.

[19] D.J. Banks, M. Barnajian, F.J. Maldonado-Arocho, A.M. Sanchez,

K.A. Bradley, Anthrax toxin receptor 2 mediates Bacillus anthracis

killing of macrophages following spore challenge, Cell. Microbiol. 7

(2005) 1173e1185.

[20] K.A. Bradley, J. Mogridge, M. Mourez, R.J. Collier, J.A. Young,

Identification of the cellular receptor for anthrax toxin, Nature 414

(2001) 225e229.

[21] K.A. Bradley, J.A. Young, Anthrax toxin receptor proteins, Biochem.

Pharmacol. 65 (2003) 309e314.

[22] H.M. Scobie, G.J. Rainey, K.A. Bradley, J.A. Young, Human capillary

morphogenesis protein 2 functions as an anthrax toxin receptor, Proc.

Natl. Acad. Sci. U.S.A. 100 (2003) 5170e5174.

[23] C.K. Cote, K.M. Rea, S.L. Norris, N. van Rooijen, S.L. Welkos, The use

of a model of in vivo macrophage depletion to study the role of

macrophages during infection with Bacillus anthracis spores, Microb.

Pathog. 37 (2004) 169e175.

[24] C.K. Cote, N. van Rooijen, S.L. Welkos, The roles of macrophages and

neutrophils in the early host response to Bacillus anthracis spores using

a mouse model of infection, Infect. Immun. 74 (2006) 469e480.

[25] S.G. Popov, T.G. Popova, E. Grene, F. Klotz, J. Cardwell, C. Bradburne,

Y. Jama, M. Maland, J. Wells, A. Nalca, T. Voss, C. Bailey, K. Alibek,

Systemic cytokine response in murine anthrax, Cell. Microbiol. 6

(2004) 225e233.

[26] D.J. Banks, S.C. Ward, K.A. Bradley, New insights into the functions of

anthrax toxin, Expert Rev. Mol. Med. 8 (2006) 1e18.

[27] C.T. Baldari, F. Tonello, S.R. Paccani, C. Montecucco, Anthrax toxins:

a paradigm of bacterial immune suppression, Trends Immunol. 27

(2006) 434e440.

[28] K.C. Brittingham, G. Ruthel, R.G. Panchal, C.L. Fuller, W.J. Ribot,

T.A. Hoover, H.A. Young, A.O. Anderson, S. Bavari, Dendritic cells

endocytose Bacillus anthracis spores: implications for anthrax pathogen-

esis, J. Immunol. 174 (2005) 5545e5552.

[29] J.L. Erwin, L.M. DaSilva, S. Bavari, S.F. Little, A.M. Friedlander,

T.C. Chanh, Macrophage-derived cell lines do not express proinflamma-

tory cytokines after exposure to Bacillus anthracis lethal toxin, Infect.

Immun. 69 (2001) 1175e1177.

[30] A.K. Pickering, M. Osorio, G.M. Lee, V.K. Grippe, M. Bray, T.J. Merkel,

Cytokine response to infection with Bacillus anthracis spores, Infect.

Immun. 72 (2004) 6382e6389.