Different Domains of Pseudomonas
aeruginosa Exoenzyme S Activate Distinct
TLRs
This information is current as
of June 18, 2017.
Slava Epelman, Danuta Stack, Chris Bell, Erica Wong,
Graham G. Neely, Stephan Krutzik, Kensuke Miyake, Paul
Kubes, Lori D. Zbytnuik, Ling Ling Ma, Xiaobin Xie,
Donald E. Woods and Christopher H. Mody
References
Subscription
Permissions
Email Alerts
This article cites 78 articles, 55 of which you can access for free at:
http://www.jimmunol.org/content/173/3/2031.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2004 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
J Immunol 2004; 173:2031-2040; ;
doi: 10.4049/jimmunol.173.3.2031
http://www.jimmunol.org/content/173/3/2031
The Journal of Immunology
Different Domains of Pseudomonas aeruginosa Exoenzyme
S Activate Distinct TLRs1
Slava Epelman,* Danuta Stack,* Chris Bell,* Erica Wong,* Graham G. Neely,‡
Stephan Krutzik,§ Kensuke Miyake,† Paul Kubes,‡ Lori D. Zbytnuik,‡ Ling Ling Ma,‡
Xiaobin Xie,* Donald E. Woods,* and Christopher H. Mody2*
T
he complex consequences of activation and exploitation
of the innate immune system by single microbial products
are poorly understood. It is known that TLRs recognize
conserved pathogen-associated molecular patterns (PAMPs)3
found within diverse microbial products. To date, eleven human
homologues have been cloned and published (1– 4). TLR4/MD-2
is responsible for transmitting an intracellular signal when a cell is
confronted by Gram-negative LPS (5, 6). TLR2 is responsible for
recognizing products from Gram-positive organisms such as peptidoglycan (PepG), bacterial lipoproteins, lipoteichoic acid (LTA),
and yeast products (7–9). A number of other microbial ligands
activate individual TLR (10 –12); however, only a few examples of
microbial activation of multiple TLR have been described, and the
mechanisms by which this occurs are poorly understood. For example TLR2 governs a response to bacterial lipopeptides and LTA
whereas the combination of TLR2/TLR6 mediates responses to
Departments of *Microbiology and Infectious Diseases, and †Medical Sciences, University of Calgary, Calgary, Alberta, Canada; ‡Division of Dermatology, University
of California School of Medicine, Los Angeles, CA 90095; and §Department of Immunology, Saga Medical School, Saga City, Japan
Received for publication January 13, 2003. Accepted for publication May 20, 2004.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was funded by the Canadian Cystic Fibrosis Foundation. S.E. is supported by a studentship from the Canadian Institutes for Health Research. C.H.M. is
the Jessie Bowden Lloyd Professor of Immunology and a Senior Scholar of the Alberta Heritage Foundation for Medical Research. D.E.W. is a recipient of a Canada
Research chair.
2
Address correspondence and reprint requests to Dr. Christopher H. Mody, Department of Microbiology and Infectious Diseases, University of Calgary, Room 273
Heritage Medical Research Building, 3330 Hospital Drive NW, Calgary, T2N-4N1
Alberta, Canada. E-mail address: [email protected]
3
Abbreviations used in this paper: PAMP, pathogen-associated molecular pattern;
ExoS, exoenzyme S; HSP, heat shock protein; PepG, peptidogylcan; LTA, lipoteichoic acid; IRAK, IL-1-associated kinase.
Copyright © 2004 by The American Association of Immunologists, Inc.
PepG and phenol-soluble modulin from Staphylococcus epidermidis (8, 13, 14). TLR1 has been shown to recognize the bacterial
lipoprotein outer surface protein A (OspA) from Borrelia burgdorferi, and may work in combination with TLR2 (15). Heat shock
protein (HSP)60 and HSP70 are normally intracellular proteins
that become liberated when a cell dies a necrotic death. When
extracellular, they are endogenous ligands of TLR; HSP60 activates TLR2 whereas HSP70 activates both TLR2 and the TLR4/
MD-2 complex (11, 12). The mechanisms by which microbial
products induce activation of multiple receptors are unknown.
Activation of all TLR involves several intracellular adaptors and
kinases. Activation of a TLR recruits the adaptor protein MyD88,
implicated in the activation of downstream signaling pathways for
all TLR (16, 17). MyD88 recruitment is followed by the recruitment of the IL-1-associated kinase (IRAK)1 and IRAK2 (18, 19).
Upon recruitment to MyD88, IRAK1/2 autophosphorylate, recruit
TNFR-associated factor 6 and link TLR stimulation to multiple
downstream signaling pathways and proinflammatory cytokine
production (20 –23). Given that TLR agonists use common signaling pathways, activation by one TLR agonist might influence the
signaling by another TLR agonist.
During infection, it is common that cells are exposed to many
TLR agonists, both simultaneously and sequentially. Repeated exposure to LPS induces a state of self-regulated unresponsiveness or
endotoxin tolerance (24). Tolerance occurs at multiple levels. For
example, both LPS stimulation and bacterial lipoprotein stimulation causes a reduction in the expression of TLR4 and TLR2, respectively, thereby preventing subsequent cellular activation to
products that used those TLRs (25, 26). The observation that tolerance also acts on the level of IRAK1 (both degradation and sequestration), which is shared by the TLR signaling pathways, led
to an understanding of the phenomenon of cross-tolerance between
TLR agonists (27–29).
0022-1767/04/$02.00
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
Some bacterial products possess multiple immunomodulatory effects and thereby complex mechanisms of action. Exogenous
administration of an important Pseudomonas aeruginosa virulence factor, exoenzyme S (ExoS) induces potent monocyte activation
leading to the production of numerous proinflammatory cytokines and chemokines. However, ExoS is also injected directly into
target cells, inducing cell death through its multiple effects on signaling pathways. This study addresses the mechanisms used by
ExoS to induce monocyte activation. Exogenous administration resulted in specific internalization of ExoS via an actin-dependent
mechanism. However, ExoS-mediated cellular activation was not inhibited if internalization was blocked, suggesting an alternate
mechanism of activation. ExoS bound a saturable and specific receptor on the surface of monocytic cells. ExoS, LPS, and peptidoglycan were all able to induce tolerance and cross-tolerance to each other suggesting the involvement of a TLR in ExoSrecognition. ExoS activated monocytic cells via a myeloid differentiation Ag-88 pathway, using both TLR2 and the TLR4/MD2/CD14 complex for cellular activation. Interestingly, the TLR2 activity was localized to the C-terminal domain of ExoS while the
TLR4 activity was localized to the N-terminal domain. This study provides the first example of how different domains of the same
molecule activate two TLRs, and also highlights the possible overlapping pathophysiological processes possessed by microbial
toxins. The Journal of Immunology, 2004, 173: 2031–2040.
2032
Materials and Methods
Cell lines and reagents
The human promonocytic cell line THP-1 (no. TIB-202; American Type
Culture Collection, Rockville, MD) was maintained in RPMI 1640, 10%
heat-inactivated FCS, 2 mM L-glutamate, 100 U/ml penicillin, 100 ␮g/ml
streptomycin, 1 mM sodium pyruvate (all from Invitrogen Life Technologies, Rockville, MD) and grown in 5% CO2. Human embryonic kidney
293 cells (HEK293) and HEK293 cells stably transfected with TLR2 were
obtained from Genentech (South San Francisco, CA) and cultured in a 1/1
mix of F-12 medium and DMEM, 10% heat-inactivated FCS, 100 U/ml
penicillin, 100 ␮g/ml streptomycin in 7% CO2. Ba/F3 cells stably transfected with the NF-␬B-driven luciferase reporter and Ba/F3 cells stably
transfected with NF-␬B/TLR4/MD-2 were cultured in RPMI 1640, 10%
heat-inactivated FCS, 100 U/ml penicillin, 100 ␮g/ml streptomycin, 10%
WEHI-3B conditioned medium as a source of IL-3, and were grown in 7%
CO2 (6). PepG (Sigma-Aldrich, St. Louis, MO), anti-TLR2 (mouse antihuman TLR2, clone TL2.1 IgG2a), anti-TLR4 (mouse anti-human TLR4,
clone HTA125 IgG2a), and the control Ab (mouse IgG2a isotype control),
were all purchased from eBioscience (San Diego, CA).
Preparation of recombinant ExoS
Recombinant ExoS, full-length N-terminal ExoS, N⬘-ExoS (amino acids
1–234), and C-terminal Exos, C⬘-ExoS (amino acids 234 – 435), were all
isolated from an Escherichia coli strain BL21(DE3) pLysS bearing a plasmid encoding histidine-tagged ExoS cloned from P. aeruginosa strain 388
(pETrHisExoS, kindly donated by Dr. J. Barbieri, Medical College of Wis-
consin, Milwaukee, WI). ExoS, N⬘-ExoS, and C⬘-ExoS were purified by
Ni2⫹-affinity chromatography (Qiagen, Mississauga, Ontario, Canada)
from cellular lysates and migrated as 52-, 32-, and 28-kDa proteins, respectively (data not shown) (43, 46). Proteins were purified to ⬎99% homogeneity. In experiments in which cellular activation was determined,
polymyxin B was preincubated with the various ExoS preparations for 5
min on ice before being incubated with cells. Although THP-1 cells were
grown in the presence of serum, when they were stimulated with ExoS, the
serum-containing medium was removed and cultured in THP-1 medium in
the absence of serum. Therefore, there would be no soluble CD14 present.
Construction of a GFP-ExoS fusion protein
A flow cytometry-optimized mutant of wild-type GFP was used to construct the GFP-ExoS fusion protein (47). NsiI sites were introduced on both
the 5⬘ and 3⬘ end of gfp in the pGY3 vector by PCR using the following
PCR primers: NsiI up, GGAGATATGTGCATATGAGTAAAGG and NsiI
down, GCTTTGCAATGCATGCCGCCTTTGTATAGTTCATCC. This
product was cloned in frame into the NsiI site downstream of the His tag
and upstream of exoS gene in pET16brHisExoS vector. The GFP-ExoS
fusion protein migrated as an ⬃80-kDa protein, reacted with anti-ExoS
antisera, fluoresced upon excitation at 488 nm and possessed the ability to
induce monocyte cytokine production and T cell activation (data not
shown).
Labeling of ExoS with the fluorescent dye Alexa 488
ExoS was labeled with Alexa 488 according to the manufacturer’s instructions (Molecular Probes, Eugene, OR). Briefly, 0.5 mg of ExoS (in 500 ␮l)
was incubated with the Alexa 488 dye for 1 h at room temperature. The
labeled mixture was layered atop a column containing purification resin,
and the labeled ExoS-Alexa 488 protein was separated from the unlabeled
dye. The progression of the labeled protein was determined by using a UV
light. Fractions containing ExoS-Alexa 488 were collected and pooled. The
protein concentration was determined and amount of dye labeled per mole
of protein was determined by the following equation: moles of Alexa/
moles of ExoS ⫽ [(A494 (of the ExoS-Alexa 488) ⫻ dilution factor)/
(71,000 cm⫺1M⫺1 ⫻ molar protein concentration)]. Typically, labeling
efficiency was ⬃1 mole of Alexa 488 to 1 mole of ExoS.
Incubation of ExoS-Alexa 488 and GFP-ExoS with THP-1 cells
THP-1 cells were incubated with ExoS-Alexa 488 or GFP-ExoS at either
4°C or 37°C for various times. In some experiments, THP-1 cells were
fixed in 1% buffered formalin before incubation with labeled ExoS. Fluorescence of labeled THP-1 cells was analyzed by flow cytometric analysis
(FACScan; BD Biosciences, San Jose, CA). For some experiments, cells
were preincubated with either cytochalasin D (dissolved in DMSO) or
DMSO alone (Calbiochem, La Jolla, CA) for 30 min before activation with
ExoS-Alexa 488. To quench the extracellular ExoS fluorescence, cells
were incubated in 0.1% trypan blue (Calbiochem) for 2 min, washed in
FACS buffer (below), and then analyzed.
Confocal microscopy
Cells were deposited onto poly-L-lysine coated coverslips and 5 ␮l of antifade mounting medium (Molecular Probes) containing 0.001% 4⬘,6⬘-diamidino-2-phenylindole (DAPI) was added. A total of 15–30 sections of 0.5
␮M in length were taken starting at the top of the cell of choice. Sections
were typically taken of cells under either ⫻63 or ⫻100 magnification
(Leica Microsystems, Exton, PA). Sections were then deconvoluted to remove image haze (AutoDeblur7.0; AutoQuant Imaging, Watervliet, NY).
Intracellular cytokine detection and flow cytometric analysis
To determine intracellular cytokine levels, cells were stimulated for 3 h
with ExoS (with 10 ␮g/ml polymyxin B) in the presence of monensin (0.66
␮l/1 ml of culture), collected and washed twice in FACS buffer (PBS, 1%
FBS, 0.l% NaN3). Cells were then fixed in CytoFix-CytoPerm for 20 min
at 4°C and washed in PermWash buffer twice according to the manufacturers instruction (BD Pharmingen, Mississauga, Ontario, Canada). Cells
were then incubated with either anti-TNF-␣ PE, anti-IL-8 PE or isotypematched control Abs (R&D Systems, Minneapolis, MN) for 30 min and
washed twice in PermWash buffer. Cells were resuspended in FACS buffer
and fluorescence was measured by flow cytometric analysis (FACScan; BD
Biosciences). The values obtained from the control Abs were subtracted
from the test group in each experiment.
Electrophoresis and Western blot
Cells were lysed in Laemmeli buffer containing 1 ␮M DTT, heated for 10
min at 37°C and electrophoresed into a 4 –12% Bis-Tris gradient gel for ⬃1
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
In an attempt to understand the complex activation of innate
immunity, we have been studying the role of a microbial product
that contributes to the LPS-independent inflammatory response induced by the Gram-negative pathogen Pseudomonas aeruginosa.
It has been disappointing that despite our knowledge of LPS, studies and clinical trials looking at the effects of neutralizing LPS in
Gram-negative-induced inflammatory responses have had only
limited success, suggesting that other microbial products may be
critical elements in the pathophysiology (30). Exoenzyme S
(ExoS) is a unique P. aeruginosa virulence factor that is produced
by many Pseudomonas isolates from diverse tissue origins and
virtually all pneumonia and cystic fibrosis pulmonary isolates (31).
Increased levels of ExoS correlates with increased pulmonary
damage in both animal models and cystic fibrosis patients (31–34).
ExoS is a complex bifunctional toxin, possessing two distinct
methods of action: extracellular and intracellular. In addition to a
type III (contact)-dependent intracellular delivery system that results in cellular cytotoxicity upon delivery of ExoS to the cytosol,
ExoS can also interact with target cells by acting as a soluble
(extracellular) protein (34 – 40). Extracellular ExoS stimulates both
monocytes and T cells directly (41– 43). Extracellular ExoS induces T cell activation resulting in T cell apoptosis, whereas
monocyte activation results in the production of a wide variety of
proinflammatory cytokines and chemokines (42). Based on these
observations, ExoS likely contributes to inflammatory responses
during P. aeruginosa infection through proinflammatory cytokine
induction (44, 45). However, because of the well-characterized
ability of ExoS to exploit both extracellular and intracellular pathways, there were a number of possible mechanisms by which cellular activation could occur.
The goal of this study was to determine the mechanisms by
which extracellular ExoS induces proinflammatory cytokine production from monocytes. To determine whether extracellular ExoS
was internalized or bound a specific cell surface receptor, flow
cytometric analysis, confocal microscopy, and competition studies
were used. Tolerization experiments and dominant negative inhibitors were used to identify the class of receptors required for ExoS
signaling. Genetic complementation studies were used to identify
the specific receptors activated by ExoS and to determine which
ExoS domains were involved in cellular activation.
ExoS ACTIVATES TLRs
The Journal of Immunology
2033
h, 200 V (Invitrogen Life Technologies, Burlington, Ontario, Canada). Proteins were transferred onto 0.4 ␮M pore sized nitrocellulose (Bio-Rad,
Mississauga, Ontario, Canada) for 1 h, 30 mV using the Novagen X-Cell
Sure lock system (Invitrogen Life Technologies). Membranes were allowed to dry for 10 min and were blocked for 1 h in 1–5% BSA (SigmaAldrich) in TBST (500 mM Tris base, 137 mM NaCl, 0.1% Tween 20,
pH ⫽ 7.6). Blots were washed two times in TBST and incubated overnight
in TBST containing 1–5% BSA with the primary Ab of choice antiphospho-Erk1/2 (1/500; Cell Signaling Technology, Beverly, MA) or antiactin (1/1000; Santa Cruz Biotechnology, Santa Cruz, CA). The next day,
the membranes were washed two times in TBST and incubated for 1 h at
room temperature in TBST with 5% skim milk powder with sheep antimouse F(ab⬘)2-HRP (1/5000). Blots were washed four times in TBST,
washed once in sterile water, and then incubated in substrate for 5 min
(Supersignal Pico; Pierce, Rockford, IL) and developed by autoradiography (Amersham Biosciences, Uppsala, Sweden).
Luciferase measurements
LPS repurification
Escherichia coli O111:B4 was purchased from Sigma-Aldrich and repurified by the protocol of Hirschfeld et al. (48). Repurified LPS did not activate TLR2 expressing cells (data not shown). LPS was used at concentrations between 1 ng/ml and 10 ␮g/ml.
Results
ExoS was specifically internalized via an actin-dependent
process
It was hypothesized that extracellular ExoS might contact the cell
surface, become internalized and induce cytokine production via
its well-characterized ability to modify signaling pathways. Specifically, ExoS possesses N-terminal ADP-ribosyl transferase activity and C-terminal GTPase activating activity for Rho family
GTPases (40, 49). If internalization of extracellular ExoS were
responsible for cellular activation, it would be through one of two
mechanisms. Following internalization, the enzymatic or catalytic
activities of intracytoplasmic ExoS could activate cells. Alternatively, the internalization process itself, during which numerous
kinases are activated, could trigger cytokine production; this process would be similar to activation of TLR9 by microbial unmethylated CpG DNA or epidermal growth-factor receptor-mediated
activation (50, 51).
THP-1 cells, an ExoS-responsive cell line (42) were used to
determine the mechanism by which ExoS activates monocytic
cells. THP-1 cells were incubated with ExoS-Alexa 488 and using
confocal microscopy the cellular distribution of ExoS-Alexa 488
(within a cytoplasmic plane or extracellular plane) was determined. ExoS-Alexa 488 accumulated within THP-1 cells, as was
seen in a cytoplasmic/intracellular section, indicating that ExoS
was internalized (Fig. 1A). Internalized ExoS-Alexa 488 exhibited
a punctate distribution, which was initially located close to the cell
surface (15 min incubation), but subsequently ExoS-Alexa 488
was found in greater amounts and dispersed throughout the intra-
FIGURE 1. ExoS was internalized. A, THP-1 cells were incubated with
ExoS-Alexa 488 (5 ␮M) for 1 h at 4°C, and then for 15 or 60 min at 37°C.
The cells were resuspended in anti-fade mounting medium (DAPI 0.1 ␮g/
ml) and cytospun onto poly-L-lysine coated slides. Fluorescence images
were captured using 30 ⫻ 0.5 ␮M sections at ⫻100 magnifications. The
optical sections were deconvoluted and a plane corresponding to the center
of the cell is shown. B, THP-1 cells were incubated at 37°C with various
concentrations of ExoS-Alexa 488 (5–90 min) or a goat anti-rabbit IgGAlexa 488 (Con). The amount of IgG-Alexa 488 was adjusted to be equifluorescent to 1 ␮M ExoS-Alexa 488. Fluorescence was determined by
flow cytometric analysis and is displayed as the mean fluorescence intensity (n ⫽ 2–3).
cellular space. Internalization was both dose- and time-dependent,
but also specific, as the accumulation of ExoS-Alexa 488 was
greater than an equifluorescent control protein (goat anti-rabbit
IgG-Alexa 488) (Fig. 1B). To determine the mechanism by which
ExoS was internalized, THP-1 cells were preincubated with cytochalasin D, an inhibitor of actin polymerization before the addition of ExoS-Alexa 488. Pretreatment with cytochalasin D inhibited accumulation of ExoS-Alexa in a dose-dependent manner
(Fig. 2A). To demonstrate that ExoS was internalized and that
cytochalasin D inhibited internalization, THP-1 cells were incubated with ExoS-Alexa 488 at 4°C (to permit only surface association) and 37°C (to allow internalization). ExoS-Alexa 488 was
quenched with trypan blue more efficiently in cells incubated at
4°C vs those incubated at 37°C (66.55 ⫾ 6.17% vs 14.41 ⫾
7.57%, p ⫽ 0.003, respectively, paired one-tailed Student’s t test).
When the THP-1 cells were pretreated with cytochalasin D, trypan
blue was able to quench the fluorescence suggesting that cytochalasin D blocked the internalization and ExoS-Alexa 488 was maintained on the cell surface (Fig. 2B).
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
A total of 2.5 ⫻ 105 HEK293 or HEK293-TLR2 cells were plated into
one-well in a six-well plate (in 3 ml) overnight. pNF-␬B-luciferase (1 ␮g;
firefly luciferase, Promega, Madison, WI) was mixed with 6 ␮l of Superfect (Qiagen) in 100 ␮l of medium (1/1 mix of DMEM and F12 medium;
Invitrogen Life Technologies) for 10 min. The mix was diluted to 700 ␮l
with complete HEK medium (1/1 mix of DMEM and F12 medium, 10%
FCS and 1% penicillin/streptomycin). The medium was removed from the
HEK cells and 700 ␮l of the Superfect/plasmid DNA mix was added, and
incubated for 2.5 h at 37°C. The cells were washed once in PBS and
resuspended in completed medium and plated in 100 ␮l samples in 96-well
plates, and rested for either 3– 4 h or overnight. In 100 ␮l, the HEK293
cells or the Ba/F3 cells were stimulated with various stimuli for 24 h in
triplicate. Each sample was collected, washed in PBS and the cell pellet
was lysed in 50 ␮l of passive lysis buffer (Promega). The lysates were
centrifuged and 20 ␮l of the soluble material lysate was assayed by addition of firefly luciferase substrate (100 ␮l). Samples were read in triplicate
in a luminometer.
2034
ExoS ACTIVATES TLRs
ExoS demonstrated cross-tolerization with LPS and PepG
FIGURE 2. ExoS was internalized by an actin-dependent process, although internalization was not required for cytokine production. A, THP-1
cells were preincubated for 30 min with different concentrations of cytochalasin D (Cyto D) or a vehicle control (DMSO), and then incubated
with 0.333 mM ExoS-Alexa 488 for various times (n ⫽ 3). B, THP-1 cells
were incubated with 1.0 ␮M ExoS-Alexa 488 for 3 h at either 4°C or 37°C
(cells incubated at 37°C were first pretreated with either DMSO or 10 ␮M
cytochalasin D for 30 min before ExoS-Alexa 488 addition). THP-1 cells
were washed and extracellular fluorescence was quenched by incubation
with 0.1% trypan blue for 2 min. Cells were analyzed by flow cytometric
analysis and the data are expressed as the percentage of ExoS-Alexa 488positive cells (n ⫽ 3). C, THP-1 cells were preincubated for 30 min with
different concentrations of cytochalasin D or DMSO, and then the cells
were stimulated with 1 ␮g/ml ExoS for 3 h in the presence of monensin.
Intracellular TNF-␣ production was detected by permeabilization and incubation with a fluorescently conjugated anti-TNF-␣ Ab. The data were
expressed as the percentage of TNF-␣⫹ cells above unstimulated cells
based on a 100% value for ExoS alone (n ⫽ 3).
Given that the process of internalization of ExoS or the enzymatic activity of ExoS could modify intracellular signaling pathways, it was necessary to determine whether internalization was
required for proinflammatory cytokine production. Surprisingly, at
concentrations that inhibited internalization, cytochalasin D was
not able to inhibit ExoS-induced TNF-␣ production, suggesting
that internalization or intracellular delivery of ExoS was not required to induce cytokine production (Fig. 2C).
ExoS binding to monocytic cells was saturable and specific
Because internalization was not required for ExoS-induced activation, an alternate mechanism of cellular activation was investigated. Experiments were performed to determine whether ExoS
activated cells through a cell surface receptor. THP-1 cells were
incubated with ExoS-Alexa 488 or an equifluorescent amount of a
control protein (goat anti-rabbit-Alexa 488) and binding was assessed by flow cytometric analysis. ExoS-Alexa 488 bound to the
Having demonstrated binding characteristics consistent with a receptor, it became of real interest to determine the receptor identity.
Based on the ability of ExoS to directly activate monocytes, it was
hypothesized that TLR were involved in ExoS recognition (42,
FIGURE 3. ExoS binds a specific and saturable receptor on monocytic
cells. A, THP-1 cells were fixed, and either incubated with PBS (shaded
area), ExoS-Alexa 488 (0.96 ␮M, solid line) or equifluorescent goat antirabbit-Alexa 488 (dashed line) for 20 min at 4°C (n ⫽ 3). Fluorescence was
measured by flow cytometric analysis. B, Saturable binding of ExoS to
THP-1 cells was demonstrated by incubating different concentrations of
ExoS-Alexa 488 with fixed THP-1 (20 min, 4°C) followed by thorough
washing. The R2 value was generated in Microsoft Excel, as was the curve
that describes the distribution of data. The data points are the average of at
least two independent experiments. Specific binding was demonstrated by
preincubating THP-1 cells with unlabeled ExoS for 20 min at 4°C before
incubation with either ExoS-Alexa 488 (300 nM, n ⫽ 2) (C) or GFP-ExoS,
(900 nM, n ⫽ 3) (D) for an additional 20 min. The cells were thoroughly
washed and fluorescence determined as above.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
THP-1 cells as indicated by an increase in fluorescence, whereas
there was no increase in fluorescence when the control protein was
used. There was a shift in the fluorescence of the entire population,
suggesting that ExoS bound to the entire THP-1 cell population
(Fig. 3A).
The two hallmark characteristics of receptor binding kinetics are
saturability and specificity. To test the first characteristic, THP-1
cells were incubated with various amounts of ExoS-Alexa 488. At
lower ExoS-Alexa 488 concentrations a rapid rise in binding occurred that was saturated as greater concentrations were used (Fig.
3B). The curve fit an exponential equation that was consistent with
saturable binding, with half-maximal binding occurring at ⬃100
nM of ExoS-Alexa 488. Unlabeled ExoS competed with the binding of labeled ExoS-Alexa 488 (to a maximum of ⬃40%) (Fig.
3C). The inability of unlabeled ExoS to completely block binding
of ExoS-Alexa 488 may be due to the propensity of ExoS to selfaggregate (46). To ensure that the Alexa 488 labeling procedure
did not affect the binding of ExoS, an additional approach was
taken. An N-terminal GFP-ExoS fusion protein was constructed.
GFP-ExoS binding could also be similarly outcompeted by unlabeled ExoS (Fig. 3D). Together, these observations suggested the
presence of a saturable and specific ExoS receptor on the surface
of THP-1 cells.
The Journal of Immunology
FIGURE 4. ExoS and other microbial products induced cross-tolerization to each other. A, THP-1 cells were pretreated with either medium (⫺),
1 ␮g/ml ExoS (E), 0.5 ␮g/ml peptidoglycan (P), or 10 ␮g/ml LPS (L) for
24 h in the absence of FCS. THP-1 cells were washed and the stimulated
with 1 ␮g/ml ExoS (⫹) for an additional 3 h in the presence of monensin.
B, THP-1 cells were pretreated as above and stimulated for an additional
3 h with either 0.5 ␮g/ml peptidoglycan (P) or 1 ␮g/ml LPS (L) in the
presence of monensin. Intracellular TNF-␣ was assessed by flow cytometric analysis (n ⫽ 3– 4). ⴱ, p ⬍ 0.05: paired, one-tailed Student’s t test.
Activation by ExoS was inhibited by a dominant negative of
MyD88
All TLR have been demonstrated to use the intracellular adaptor
molecule MyD88 as one of the most proximal intracellular signaling molecules, linking TLR to proinflammatory cytokine production (17, 19, 23, 54). To determine whether a TLR was involved in
ExoS induced cellular activation, or some other receptor that did
not use MyD88, THP-1 cells were transfected with a dominant
negative version of MyD88 (MyD88 152–296, MyD88DN) (18).
THP-1 cells were cotransfected with pGFP and either pCDNA3
(control vector) or pMyD88DN. The cells were rested overnight,
stimulated with ExoS and TNF-␣ production assessed in GFP⫹
cells by flow cytometric analysis. MyD88DN inhibited ExoS-induced TNF-␣ production in a dose-dependent manner (Fig. 5).
This result suggested that an MyD88-dependent pathway was involved in the activation of monocytes by ExoS and is consistent
with a TLR being involved in ExoS recognition.
ExoS activated cells expressing either TLR2 or
TLR4/MD-2/CD14
Because TLR2 and TLR4/MD-2 agonists (PepG and LPS) induced
a state of tolerance to further stimulation by ExoS, experiments
were performed to investigate whether TLR2 and/or TLR4/MD-2
were involved in the recognition of ExoS. To determine whether
the TLR4/MD-2 complex was involved in ExoS signaling, a human IL-3-dependent B cell line Ba/F3 that expressed TLR4,
MD-2, and an NF-␬B-luciferase reporter was compared with a cell
line that expressed only the reporter (6). ExoS induced dose-dependent induction in the cell line expressing TLR4/MD-2, but did
not activate cells lacking TLR4/MD-2 (Fig. 6A). Similarly, LPS
induced activation of Ba/F3 cells only if they expressed the TLR4/
FIGURE 5. ExoS induced TNF-␣ production via a MyD88-dependent
pathway. THP-1 cells were electroporated with pGFP (1 ␮g/1 ⫻ 106 cells)
and either pCDNA3 or pMyD88DN at various concentrations. THP-1 cells
were washed and rested. The following day, THP-1 cells were stimulated
with either 10 ␮g/ml polymyxin B alone, or 1 ␮g/ml ExoS in the presence
of 10 ␮g/ml polymyxin B for 3 h and intracellular TNF-␣ levels determined by FACS analysis. For each group, unstimulated (polymyxin B
alone) TNF-␣ mean fluorescence intensity was subtracted from that of the
stimulated group to determine the net TNF-␣ induction. The data are presented as a percentage of net TNF-␣ production based on the control plasmid (pCDNA3) for each of the plasmid concentrations. TNF-␣ production
was assessed by first gating on GFP⫹ cells and determining TNF-␣ production within that subset. This data presented are the average of two
independent experiments ⫾ SEM.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
44). To explore this possibility, experiments exploited the observation that stimulation of monocytes by LPS, followed by a secondary stimulation with LPS induces endotoxin tolerance, or an
inability of a cell type to be activated by a secondary exposure to
LPS (24). The phenomenon also holds for other TLR ligands, as
activation with one TLR ligand leads to tolerization of other family
members (52). Tolerance is not a result of general cellular activation, as LPS pretreatment does not inhibit cellular activation in
response to phorbol ester stimulation (53). Because the hypothesis
was that ExoS was operating through a TLR, the possibility that
there could be some cross-tolerance induced between ExoS and
TLR agonists was explored.
When THP-1 cells were pretreated with ExoS, PepG or LPS, a
secondary stimulation by ExoS was inhibited (60 –90% inhibition)
(Fig. 4A). Additionally, pretreatment with ExoS inhibited secondary stimulation by both PepG and LPS (Fig. 4B). In fact, all three
ligands induced tolerance and cross-tolerance to themselves and to
each other. These results suggested that a TLR was involved in the
recognition of ExoS or that there were common elements in the
activation pathways between PepG, LPS, and ExoS.
2035
2036
ExoS ACTIVATES TLRs
FIGURE 6. ExoS activated cells that express TLR4/CD14/MD-2. A,
Ba/F3 that stably expressed either an NF-␬B promoter driven luciferase
vector (NF-␬B) or Ba/F3 cells that stably expressed NF-␬B-luciferase/
TLR4/MD-2 were stimulated with different concentrations of ExoS (in
micrograms per milliliter) in the presence of 10 ␮g/ml polymyxin B for
24 h. Triplicate samples were lysed and luciferase levels were determined.
Data are presented as the mean of the triplicate samples (n ⫽ 2– 4) for all
groups. B, Ba/F3 that stably expressed either NF-␬B-luciferase/TLR4/
MD-2 or NF-␬B-luciferase/TLR4/MD-2/CD14 were stimulated with different concentrations of ExoS (␮g/ml) in the presence of 10 ␮g/ml polymyxin B or LPS for 24 h. The samples were analyzed in quadruplicate. One
of four representative experiments. ⴱ, p ⬍ 0.05 when compared with Ba/F3
cells stably transfected NF-␬B/TLR4/MD-2.
MD-2 complex (data not shown). To determine whether CD14
enhanced cellular activation, Ba/F3 cells that expressed TLR4/
MD-2/NF-␬B-luciferase were compared with cells expressing
TLR4/CD-2/CD14/NF-␬B-luciferase. Both cells lines were stimulated with LPS and ExoS using conditions that maximized the
difference between the cell lines. The presence of CD14 enhanced
the responsiveness of Ba/F3 cells to both LPS and ExoS (Fig. 6B).
Previous studies had demonstrated that LPS contamination was
not responsible for ExoS-induced proinflammatory cytokine production (42, 44). In addition, the amount of LPS activity in the
ExoS preparation was calculated using titration experiments
(4.3 ⫾ 1.3 ng of LPS activity per 1.0 ␮g of ExoS). The amount of
polymyxin B (10 ␮g/ml) present was sufficient to neutralize at
FIGURE 7. ExoS activated cells that express TLR2. A, HEK293 cells or
HEK293-TLR2 cells were stimulated for various times with either 5 ␮g/ml
peptidoglycan (P) for 5 min or 5 ␮g/ml ExoS (E) for 30 min in the presence
of 10 ␮g/ml polymyxin B. Cells were lysed and a Western blot was performed for phospho-Erk1/2 and actin. One of two representative experiments is shown. B, HEK293 cells or HEK293-TLR2 cells were transiently
transfected with an NK-␬B-luciferase reporter plasmid and then stimulated
were stimulated with ExoS or peptidoglycan (PepG) for 6 h, in the presence
of 10 ␮g/ml polymyxin B. The cells were then lysed, luciferase production
determined, and the data are presented as fold induction over unstimulated
cells in triplicate (n ⫽ 3). ⴱ, p ⬍ 0.05 when compared with HEK293 cells
alone.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
least 1 ␮g of LPS, so that LPS contamination was not responsible
for the TLR4-mediated cellular activation by ExoS. To further
assess whether any other contaminating molecules were present in
the ExoS preparation we performed the following experiment. E.
coli BL21 (DE3) cells containing the plasmid encoding ExoS were
grown for 2 h. The cells were split into two groups-ExoS production was induced with isopropyl ␤-D-thiogalactoside (IPTG) in one
group, whereas the other group received only PBS as a control for
an additional 2 h. Both groups were then subject to the identical
ExoS purification protocol. The eluted fractions from each group
were used to stimulate THP-1 cells and intracellular TNF-␣ production was assessed. Only the eluted fractions from the IPTG
induced group were capable of inducing TNF-␣ from THP-1 cells.
The IPTG-induced eluent plus polymyxin B induced TNF-␣ on
8.2 ⫾ 1.2% THP-1 cells vs 0.1 ⫾ 0.3% of THP-1 cells when
stimulated by the uninduced eluent, n ⫽ 3, p ⫽ 0.015). The data
indicated that ExoS was responsible for the cytokine induction,
and not any other molecules that were copurified with ExoS. Experiments were also performed to determine whether ExoS activated cells expressing TLR2. We used HEK293 cells that expressed TLR2 (11). These cells were stimulated with ExoS and the
phosphorylation of Erk1/2, a critical factor in the induction of
TNF-␣ was determined (21, 55). Both ExoS, and the previously
characterized TLR2 agonist PepG, induced Erk1/2 phosphorylation only in cells that expressed TLR2 (Fig. 7A). HEK-TLR2 cells
were also transfected with an NF-␬B reporter plasmid, and they
The Journal of Immunology
2037
too were responsive to stimulation by ExoS, the 19-kDa lipoprotein from Mycobacterium tuberculosis, but not highly purified
LPS, although the magnitude of activation was not as great as seen
with Erk1/2 phosphorylation (Fig. 7B and data not shown).
Additional experiments were performed to ensure that TLR2
and TLR4 activity was indeed specific. For this purpose, neutralizing Abs to TLR2 and TLR4 were used to block ExoS-induced
TNF-␣ production from THP-1 monocytes. Blocking TLR2 resulted in a modest but statistically significant reduction (19.0 ⫾
8.6%) in the expression of TNF-␣ (Fig. 8). Blocking TLR4 resulted in a substantial reduction (57.2 ⫾ 7.8%) in the expression of
TNF-␣. Together, the blocking Abs resulted in an additive effect
(71.4 ⫾ 9.9% inhibition). The specificity of the Abs was demonstrated by blocking the response to PepG and LPS. The anti-TLR2
Ab inhibited PepG induced TNF-␣ production (70.4 ⫾ 4.9% inhibition, n ⫽ 5, p ⬍ 0.05) and the anti-TLR4 Ab inhibited highly
purified LPS induced TNF-␣ production (83.3 ⫾ 4.6% inhibition,
n ⫽ 3, p ⬍ 0.05).
Given that ExoS activated both TLR2 and the TLR4/MD-2 complex, we hypothesized that the same domain of ExoS was responsible for both activities. Recombinant protein preparations of ExoS
were produced that contained full-length ExoS (amino acid
1– 435), N terminus (N⬘-ExoS amino acid 1–234), or C terminus
(C⬘-ExoS amino acid 232– 435). It has previously been shown that
these domains are biologically active when they are expressed in
the truncated form (46, 56). Full-length ExoS, N⬘-ExoS, and C⬘ExoS were used to stimulate both TLR4/MD-2 and TLR2-expressing cell lines. N⬘-ExoS induced activation of Ba/F3 cells that ex-
FIGURE 8. Neutralization of both TLR2 and TLR4 blocks ExoS-induce TNF-␣ production. THP-1 cells were pretreated with medium (ExoS),
10 ␮g/ml control Ab (ExoS ⫹ Con), 10 ␮g/ml anti-TLR2 (Exo ⫹ anti-2),
10 ␮g/ml anti-TLR4 (Exo ⫹ anti-4), 20 ␮g/ml control Ab (ExoS ⫹ Con*)
or 10 ␮g/ml anti-TLR2 ⫹ 10 ␮g/ml anti-TLR4 (Exo ⫹ anti-2/4) for 45 min
at 37°C. Cells were then stimulated with either 10 ␮g/ml polymyxin B
alone, or 500 ng/ml ExoS in the presence of 10 ␮g/ml polymyxin B for 4 h
at 37°C and intracellular TNF-␣ levels were determined by FACS analysis.
The percentage TNF-␣⫹ cells induced by ExoS was determined and the
values were normalized to ExoS alone. The data were presented as the
average ⫾ SEM (n ⫽ 4 –5 for all groups). ⴱ, p ⬍ 0.05 as compared with
10 ␮g/ml control Ab. #, p ⬍ 0.05 as compared with 20 ␮g/ml control Ab.
FIGURE 9. The N-terminal domain of ExoS was responsible for TLR4mediated activation whereas the C-terminal domain was responsible for
TLR2-mediated activation. A, Ba/F3 stably transfected with either an
ELAM-1 promoter driven luciferase vector (NF-␬B) or Ba/F3 cells stably
transfected with NF-␬B-luciferase/TLR4/MD-2 were stimulated with different concentrations of either the N-terminal ExoS (N⬘) or C-terminal
ExoS (C⬘) in the presence of 10 ␮g/ml polymyxin B. Triplicate samples
were lysed and luciferase levels were determined. Data are presented as the
mean of the triplicate samples (n ⫽ 3). B, HEK293 cells or HEK293-TLR2
cells were stimulated for 5 min with various concentrations of either N⬘ExoS or C⬘-ExoS. Cells were lysed and a Western blot analysis was performed for phospho-Erk1/2 and actin. One of two representative experiments is shown.
pressed TLR4/MD-2, whereas it did not activate control Ba/F3
cells (Fig. 9A). The degree of activation was almost as great as
with full-length ExoS. By contrast, the C⬘-ExoS domain was incapable of activating Ba/F3 cells that expressed TLR4/MD-2 (Fig.
9A). Surprisingly, the C⬘-ExoS domain possessed the ability to
induce TLR2-dependent Erk1/2 phosphorylation (Fig. 9B),
whereas the N terminus was much less capable of the same activity. Thus, the data indicated that the N⬘-ExoS activated TLR4 and
the C⬘-ExoS activated TLR2, and therefore that two distinct domains of ExoS were responsible for the TLR2- and TLR4/MD-2dependent activity.
Discussion
In this report we have made six observations: 1) ExoS was internalized by monocytic cells, although internalization was not required for cellular activation; 2) ExoS bound a specific and saturable receptor; 3) primary stimulation with either ExoS, LPS or
PepG induced a state of cross-tolerance to one another; 4) ExoSinduced TNF-␣ production was MyD88-dependent; 5) ExoS activated cells by both TLR2 and TLR4/MD-2/CD14-dependent pathways; and 6) the ability to activate cells expressing TLR2 was
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
TLR2 and TLR4 activity was dependent on different domains
of ExoS
2038
burgdorferi OspA can act directly on T cells and provide costimulatory signals for T cell activation (67). Taken together, these observations suggest that ExoS may use a TLR to activate
monocytes.
One of the important characteristics of TLR is that they display
cross-tolerization. That is, ligands that bind and activate cells via
one TLR cause decreased sensitivity to different ligands that bind
the same, but also different TLR (52). Experiments were performed to determine whether stimulation with ExoS would decrease the sensitivity to ligands known to activate cells via TLR2
or TLR4 and vice versa. Pretreatment of THP-1 cells with ExoS
resulted in tolerance to subsequent stimulation with ExoS, and
both PepG and LPS. Conversely, pretreatment with PepG or LPS
resulted in tolerance to ExoS. In fact, cross-tolerance was demonstrated between all the products, suggesting use of similar signaling molecules involved in the TLR pathway. This process is similar to that observed with LPS or LTA, which render monocytes
tolerant to further stimulation with either one of the two stimuli
(52). Tolerization is known to occur at multiple levels, including at
the level of receptor down-regulation and degradation IRAK1, a
common signaling intermediate. However, an unlikely possibility
existed that the tolerance observed was not due to TLR activation,
but a secondary effect due to secretion of cytokines during the
primary stimulation period, such as TNF-␣ and IL-1. ExoS induces
both TNF-␣ and IL-1 (42, 44) and both cytokines have been shown
to tolerize monocytic cells to secondary stimulation by LPS (68).
Although this possibility was unlikely (52), additional experiments
were required to definitely demonstrate that TLR were the class of
receptors involved in ExoS recognition.
ExoS induced MyD88-dependent TNF-␣ production. MyD88 is
required for NF-␬B activation and subsequent proinflammatory
cytokine production downstream of all TLR, which provided
strong evidence that a TLR receptor was involved in ExoS recognition (10, 17, 20, 54, 69). Because TNF-␣ production was assessed in the presence of monensin, a secretion inhibitor, the effects of the MyD88DN was a primary event, and not due to the
release of a soluble factor. Together with the ability to tolerize the
cells, the data implicated a TLR in the recognition of ExoS. As
PepG and LPS induced a state of tolerance to further ExoS stimulation, we tested ability of TLR4/MD-2 and/or TLR2 to transform
ExoS nonresponsive cell lines into ExoS-responsive lines.
Studies with stably transfected cell lines and Ab neutralization
experiments indicated that ExoS activated cells through both
TLR2 and the TLR4/MD-2/CD14 complexes. This response is
similar to the endogenous TLR agonist HSP70 (11), which has also
been shown to work through these two receptors, and indicates that
a single microbial protein can also activate multiple TLR (although
the active domains of HSP70 have not been established). In either
the case of HSP70 or ExoS, it is not known whether these products
directly bind a TLR, or whether they act similarly to LPS and bind
a high affinity receptor (CD14), in which a complex of molecules
then interacts with TLR4 (70, 71). The presence of CD14 enhanced the responsiveness of cells to ExoS, but was not required
indicating that CD14 was not essential in ExoS-mediated cellular
activation. Both ExoS and HSP70 do however bind specific receptors. In the case of ExoS, the receptor binding studies indicated
that ExoS did bind specifically to a receptor, although whether this
was to a TLR, CD14, or to a molecule analogous to CD14 has yet
to be determined. HSP70 binds specifically to CD91, and it is
possible that a complex between CD91 and a TLR forms (72).
Additionally, it was possible that ExoS bound to two different high
affinity receptors, one that interacted with TLR2 and one that interacted with TLR4/MD-2. However, when examining the results
of ExoS binding, it did not appear that it followed second order
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
attributed to the C terminus of ExoS and the ability to activate the
TLR4/MD-2 complex was attributed to the N terminus of ExoS.
It is known that ExoS is secreted by a bacterial type III secretion
system through both contact-dependent (intracellular delivery) and
contact-independent pathways (extracellular action). In addition to
the contact-dependent activity (attributed to ExoS injection into
target cells) (35), contact-independent secretion results in the release of soluble extracellular ExoS, which has previously been
shown to possess potent immunoinflammatory activity. Extracellular ExoS activates both monocytes and T cells directly, resulting
in monocyte proinflammatory cytokine production, T cell proliferation and T cell apoptosis (41, 42, 44, 45, 57, 58). However, the
mechanism by which extracellular ExoS activated these immune
cells was not known, nor was it known whether activation was
dependent on the enzymatic/catalytic domains possessed by ExoS.
Extracellular ExoS was internalized into monocytic cells, whereas
a control protein was not internalized, suggesting that the process
was specific. Internalization was dependent on actin polymerization and led to the accumulation of ExoS within monocytic cells.
Internalized ExoS labeled with a punctate pattern within the cell,
which suggested that ExoS remained in vesicles rather than being
diffusely localized through the cytosol. Additionally, when THP-1
cells were allowed to internalize ExoS-Alexa 488 (incubation at
37°C), the fluorescent signal was protected from quenching by
trypan blue, confirming that ExoS was in fact intracellular.
We felt that there were a number of possible mechanisms by
which internalized ExoS may have been involved in cellular activation. Internalization may have been a prerequisite step before the
entry of ExoS into the cytosol, a compartment in which the enzymatic activities of ExoS could play a role. Alternatively, the internalization process, during which numerous kinases are activated, could trigger cytokine production; this process would be
similar to the internalization of membrane-bound vesicles containing TLR9 during CpG DNA-induce activation, or could follow
from internalization of the epidermal growth-factor ligand/receptor
complex (50, 51, 59). However, blockade of internalization did not
inhibit ExoS-induced cytokine production indicating that although
ExoS was specifically internalized, neither internalization nor the
process of internalization was required for ExoS-induced proinflammatory cytokine production.
This lead to an alternate hypothesis whereby ExoS might bind a
specific cell surface receptor that initiated a signal transduction
cascade leading to proinflammatory cytokine production. ExoSAlexa 488 bound THP-1 cells in specific and saturable manner.
However, only 40% of the binding could be outcompeted by unlabeled ExoS, which may be due to either ExoS binding nonspecifically to the cell membrane and/or may be related to the ability
of ExoS to self-aggregate (46). It was clear that ExoS bound a
saturable and specific receptor and the preferential internalization
of ExoS may be due to receptor ligation, rather than simply part of
the natural endocytic process of monocytic cells.
Because microbial products can activate cells via pattern recognition receptors, and TLR are a major class of these molecules, the
possibility that ExoS activated a TLR was considered. TLR bear
the responsibility for recognizing a wide variety of stimuli including LPS, LTA, lipoproteins, and PepG (60, 61). There are a number of similarities in the inflammatory response to these molecules
and ExoS. LTA, PepG, and lipoproteins are best known for inducing proinflammatory cytokine production from monocytes, and the
cytokine profile is similar to that induced by ExoS (44, 62– 64). In
addition, LTA, PepG, and lipoproteins have been shown to have
direct effects on T cells. LTA has been shown to bind directly to
T cells and act as a T cell mitogen, similarly to ExoS (65). PepG
was also shown to be mitogenic for human T cells (66) and B.
ExoS ACTIVATES TLRs
The Journal of Immunology
Acknowledgments
We thank Dr. Joseph Barbieri for kindly providing the full-length pETrHisExoS vector as well as the vectors encoding the N-terminal and C-terminal fragments. We thank Paul Godowski (Genentech, South San Francisco,
CA) for TLR2 expressing HEK cells, Jessica Boyd and Julie Ethier for help
in the construction of the GFP-ExoS fusion protein, and Robert Modlin for
help with experiments involving TLR-transfected cells. He also thank Tularik (South San Francisco, CA) for providing the dominant negative
MyD88 construct.
References
1. Zhang, D., G. Zhang, M. S. Hayden, M. B. Greenblatt, C. Bussey, R. A. Flavell,
and S. Ghosh. 2004. A Toll-like receptor that prevents infection by uropathogenic
bacteria. Science 303:1522.
2. Muzio, M., N. Polentarutti, D. Bosisio, M. K. Prahladan, and A. Mantovani.
2000. Toll-like receptors: a growing family of immune receptors that are differentially expressed and regulated by different leukocytes. J. Leukocyte Biol.
67:450.
3. Rock, F. L., G. Hardiman, J. C. Timans, R. A. Kastelein, and J. F. Bazan. 1998.
A family of human receptors structurally related to Drosophila Toll. Proc. Natl.
Acad. Sci. USA 95:588.
4. Takeuchi, O., T. Kawai, H. Sanjo, N. G. Copeland, D. J. Gilbert, N. A. Jenkins,
K. Takeda, and S. Akira. 1999. TLR6: a novel member of an expanding Toll-like
receptor family. Gene 231:59.
5. Chow, J. C., D. W. Young, D. T. Golenbock, W. J. Christ, and F. Gusovsky.
1999. Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J. Biol. Chem. 274:10689.
6. Shimazu, R., S. Akashi, H. Ogata, Y. Nagai, K. Fukudome, K. Miyake, and
M. Kimoto. 1999. MD-2, a molecule that confers lipopolysaccharide responsiveness on Toll-like receptor 4. J. Exp. Med. 189:1777.
7. Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda,
and S. Akira. 1999. Differential roles of TLR2 and TLR4 in recognition of Gramnegative and Gram-positive bacterial cell wall components. Immunity 11:443.
8. Schwandner, R., R. Dziarski, H. Wesche, M. Rothe, and C. J. Kirschning. 1999.
Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by Tolllike receptor 2. J. Biol. Chem. 274:17406.
9. Underhill, D. M., A. Ozinsky, A. M. Hajjar, A. Stevens, C. B. Wilson,
M. Bassetti, and A. Aderem. 1999. The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature 401:811.
10. Hayashi, F., K. D. Smith, A. Ozinsky, T. R. Hawn, E. C. Yi, D. R. Goodlett,
J. K. Eng, S. Akira, D. M. Underhill, and A. Aderem. 2001. The innate immune
response to bacterial flagellin is mediated by Toll-like receptor 5. Nature
410:1099.
11. Vabulas, R. M., P. Ahmad-Nejad, S. Ghose, C. J. Kirschning, R. D. Issels, and
H. Wagner. 2002. HSP70 as endogenous stimulus of the Toll/interleukin-1 receptor signal pathway. J. Biol. Chem. 277:15107.
12. Ohashi, K., V. Burkart, S. Flohe, and H. Kolb. 2000. Cutting edge: heat shock
protein 60 is a putative endogenous ligand of the Toll-like receptor-4 complex.
J. Immunol. 164:558.
13. Ozinsky, A., D. M. Underhill, J. D. Fontenot, A. M. Hajjar, K. D. Smith,
C. B. Wilson, L. Schroeder, and A. Aderem. 2000. The repertoire for pattern
recognition of pathogens by the innate immune system is defined by cooperation
between Toll-like receptors. Proc. Natl. Acad. Sci. USA 97:13766.
14. Hajjar, A. M., D. S. O’Mahony, A. Ozinsky, D. M. Underhill, A. Aderem,
S. J. Klebanoff, and C. B. Wilson. 2001. Cutting edge: functional interactions
between Toll-like receptor (TLR) 2 and TLR1 or TLR6 in response to phenolsoluble modulin. J. Immunol. 166:15.
15. Alexopoulou, L., V. Thomas, M. Schnare, Y. Lobet, J. Anguita, R. T. Schoen,
R. Medzhitov, E. Fikrig, and R. A. Flavell. 2002. Hyporesponsiveness to vaccination with Borrelia burgdorferi OspA in humans and in TLR1- and TLR2deficient mice. Nat. Med. 8:878.
16. Kaisho, T., O. Takeuchi, T. Kawai, K. Hoshino, and S. Akira. 2001. Endotoxininduced maturation of MyD88-deficient dendritic cells. J. Immunol. 166:5688.
17. Kawai, T., O. Adachi, T. Ogawa, K. Takeda, and S. Akira. 1999. Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity 11:115.
18. Wesche, H., W. J. Henzel, W. Shillinglaw, S. Li, and Z. Cao. 1997. MyD88: an
adapter that recruits IRAK to the IL-1 receptor complex. Immunity 7:837.
19. Muzio, M., J. Ni, P. Feng, and V. M. Dixit. 1997. IRAK (Pelle) family member
IRAK-2 and MyD88 as proximal mediators of IL-1 signaling. Science 278:1612.
20. Wang, Q., R. Dziarski, C. J. Kirschning, M. Muzio, and D. Gupta. 2001. Micrococci
and peptidoglycan activate TLR23 MyD883 IRAK3 TRAF3 NIK3 IKK3
NF-␬B signal transduction pathway that induces transcription of interleukin-8. Infect.
Immun. 69:2270.
21. Dziarski, R., Y. P. Jin, and D. Gupta. 1996. Differential activation of extracellular
signal-regulated kinase (ERK) 1, ERK2, p38, and c-Jun NH2-terminal kinase
mitogen-activated protein kinases by bacterial peptidoglycan. J. Infect. Dis.
174:777.
22. Guha, M., M. A. O’Connell, R. Pawlinski, A. Hollis, P. McGovern, S. F. Yan,
D. Stern, and N. Mackman. 2001. Lipopolysaccharide activation of the MEKERK1/2 pathway in human monocytic cells mediates tissue factor and tumor
necrosis factor ␣ expression by inducing Elk-1 phosphorylation and Egr-1 expression. Blood 98:1429.
23. Zhang, F. X., C. J. Kirschning, R. Mancinelli, X. P. Xu, Y. Jin, E. Faure,
A. Mantovani, M. Rothe, M. Muzio, and M. Arditi. 1999. Bacterial lipopolysaccharide activates nuclear factor-␬B through interleukin-1 signaling mediators in
cultured human dermal endothelial cells and mononuclear phagocytes. J. Biol.
Chem. 274:7611.
24. Greisman, S. E., E. J. Young, and W. E. Woodward. 1966. Mechanisms of endotoxin tolerance. IV. Specificity of the pyrogenic refractory state during continuous intravenous infusions of endotoxin. J. Exp. Med. 124:983.
25. Wang, J. H., M. Doyle, B. J. Manning, Q. D. Wu, S. Blankson, and
H. P. Redmond. 2002. Induction of bacterial lipoprotein tolerance is associated
with suppression of Toll-like receptor 2 expression. J. Biol. Chem. 277:36068.
26. Nomura, F., S. Akashi, Y. Sakao, S. Sato, T. Kawai, M. Matsumoto,
K. Nakanishi, M. Kimoto, K. Miyake, K. Takeda, and S. Akira. 2000. Cutting
edge: endotoxin tolerance in mouse peritoneal macrophages correlates with
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
kinetics, which would have suggested a second receptor. Other
TLR agonists interact with several different cell surface proteins.
For example, PepG binds directly to TLR2, although PepG also
binds to a host of transmembrane PepG binding proteins and also
CD14 (73, 74). In addition to the well-characterized interactions
between LPS and CD14, LPS responsiveness is enhanced by the
CD11/CD18 complex, acting in an analogous fashion to CD14
(75). Therefore, it is possible that multiple proteins/receptors may
interact with ExoS to facilitate TLR-dependent activation (76).
Interestingly, the TLR2 and TLR4 activity of ExoS could be
localized to different ExoS domains. ExoS possesses a C-terminal
ADP-ribosyl transferase domain and an N-terminal GAP domain
for Rho family GTPases. When translocated intracellularly by a
bacterial type III secretion system, intracellular ExoS inhibits
DNA synthesis and actin polymerization via these two domains,
respectively (35, 40, 56, 77). Extracellularly, it also appears that
these two domains play different roles, although this is not likely
to be due to their enzymatic activity. N⬘-ExoS possessed the ability
to activate cells expressing TLR4, whereas C⬘-ExoS activated cells
expressing TLR2. Amino acids 36 – 68 of the N⬘-ExoS, when expressed intracellularly, target ExoS to the cytoplasmic leaflet of the
plasma membrane (56). Whether this intracellular targeting domain is involved during extracellularly delivery is not yet known.
To our knowledge, this is first report indicating that two different domains of the same molecule activate two different TLR.
Simultaneous stimulation of TLR2 and TLR4/MD-2/CD14 induces additive activation, which may explain why such an extensive array of proinflammatory cytokines and chemokines is induced by ExoS (76). However, this process may be different from
coligation/coaggregation of TLR2 and TLR4/MD-2/CD14 together (42). Unlike simultaneous activation of TLR2 and TLR4/
MD-2/CD14, in which both signal transduction pathways would be
activated in parallel, but spatially could be segregated; coligation
would bring the signaling pathways into close proximity with one
another, which could qualitatively change the outcome. For example, inhibition of Fc⑀RI signaling occurs only when the Fc␥RII
receptor is cross-linked to the Fc⑀RI by the same multivalent Ag,
and if both Fc␥RII and Fc⑀RI are ligated, but not cross-linked,
Fc⑀RI signaling remains normal (78). The idea of coligation is an
interesting, and as of yet totally unexamined concept in the field of
TLR biology. This is particularly relevant given the likelihood of
TLR being simultaneously stimulated and/or coligated during infection with organisms that produces numerous TLR agonists, or
as in the case of ExoS, different TLR being coligated by the same
molecule.
The importance of such PAMP is an important link between
innate and adaptive immunity. The current work has identified an
important ligand from a Gram-negative bacteria that has both intracellular, and specific TLR-mediated extracellular activity. There
is much to be learned about PAMP and the current study brings us
one step closer by identifying a mechanism by which a single
ligand can stimulate multiple TLR.
2039
2040
27.
28.
29.
30.
31.
32.
33.
34.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
polysaccharide signal transduction pathway through mitogen-activated protein
kinase and nuclear factor-␬B cascades, but not lipopolysaccharide-incorporation
steps. Biochim. Biophys. Acta 1450:130.
Takeuchi, O., K. Takeda, K. Hoshino, O. Adachi, T. Ogawa, and S. Akira. 2000.
Cellular responses to bacterial cell wall components are mediated through
MyD88-dependent signaling cascades. Int. Immunol. 12:113.
van der Bruggen, T., S. Nijenhuis, E. van Raaij, J. Verhoef, and B. S. van Asbeck.
1999. Lipopolysaccharide-induced tumor necrosis factor ␣ production by human
monocytes involves the raf-1/MEK1-MEK2/ERK1-ERK2 pathway. Infect. Immun. 67:3824.
Pederson, K. J., S. Pal, A. J. Vallis, D. W. Frank, and J. T. Barbieri. 2000.
Intracellular localization and processing of Pseudomonas aeruginosa ExoS in
eukaryotic cells. Mol. Microbiol. 37:287.
Barclay, N. G., J. C. Spurrell, T. F. Bruno, D. G. Storey, D. E. Woods, and
C. H. Mody. 1999. Pseudomonas aeruginosa exoenzyme S stimulates murine
lymphocyte proliferation in vitro. Infect. Immun. 67:4613.
Mody, C. H., D. E. Buser, R. M. Syme, and D. E. Woods. 1995. Pseudomonas
aeruginosa exoenzyme S induces proliferation of human T lymphocytes. Infect.
Immun. 63:1800.
Ahmad-Nejad, P., H. Hacker, M. Rutz, S. Bauer, R. M. Vabulas, and H. Wagner.
2002. Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors at
distinct cellular compartments. Eur. J. Immunol. 32:1958.
Aliprantis, A. O., R. B. Yang, M. R. Mark, S. Suggett, B. Devaux, J. D. Radolf,
G. R. Klimpel, P. Godowski, and A. Zychlinsky. 1999. Cell activation and apoptosis by bacterial lipoproteins through Toll-like receptor-2. Science 285:736.
Muzio, M., and A. Mantovani. 2000. Toll-like receptors. Microbes Infect. 2:251.
Gupta, D., Y. P. Jin, and R. Dziarski. 1995. Peptidoglycan induces transcription
and secretion of TNF-␣ and activation of lyn, extracellular signal-regulated kinase, and rsk signal transduction proteins in mouse macrophages. J. Immunol.
155:2620.
Hermann, C., I. Spreitzer, N. W. Schroder, S. Morath, M. D. Lehner, W. Fischer,
C. Schutt, R. R. Schumann, and T. Hartung. 2002. Cytokine induction by purified
lipoteichoic acids from various bacterial species–role of LBP, sCD14, CD14 and
failure to induce IL-12 and subsequent IFN-␥ release. Eur. J. Immunol. 32:541.
Nishiguchi, M., M. Matsumoto, T. Takao, M. Hoshino, Y. Shimonishi, S. Tsuji,
N. A. Begum, O. Takeuchi, S. Akira, K. Toyoshima, and T. Seya. 2001. Mycoplasma fermentans lipoprotein M161Ag-induced cell activation is mediated by
Toll-like receptor 2: role of N-terminal hydrophobic portion in its multiple functions. J. Immunol. 166:2610.
Beachey, E. H., J. B. Dale, S. Grebe, A. Ahmed, W. A. Simpson, and I. Ofek.
1979. Lymphocytes binding and T cell mitogenic properties of group A streptococcal lipoteichoic acid. J. Immunol. 122:189.
Rasanen, L., and H. Arvilommi. 1982. Cell walls, peptidoglycans, and teichoic
acids of Gram-positive bacteria as polyclonal inducers and immunomodulators of
proliferative and lymphokine responses of human B and T lymphocytes. Infect.
Immun. 35:523.
Knigge, H., M. M. Simon, S. C. Meuer, M. D. Kramer, and R. Wallich. 1996. The
outer surface lipoprotein OspA of Borrelia burgdorferi provides co-stimulatory
signals to normal human peripheral CD4⫹ and CD8⫹ T lymphocytes. Eur. J. Immunol. 26:2299.
Ferlito, M., O. G. Romanenko, S. Ashton, F. Squadrito, P. V. Halushka, and
J. A. Cook. 2001. Effect of cross-tolerance between endotoxin and TNF-␣ or
IL-1␤ on cellular signaling and mediator production. J. Leukocyte Biol. 70:821.
Chuang, T. H., J. Lee, L. Kline, J. C. Mathison, and R. J. Ulevitch. 2002. Toll-like
receptor 9 mediates CpG-DNA signaling. J. Leukocyte Biol. 71:538.
da Silva, C. J., K. Soldau, U. Christen, P. S. Tobias, and R. J. Ulevitch. 2001.
Lipopolysaccharide is in close proximity to each of the proteins in its membrane
receptor complex: transfer from CD14 to TLR4 and MD-2. J. Biol. Chem.
276:21129.
Jiang, Q., S. Akashi, K. Miyake, and H. R. Petty. 2000. Lipopolysaccharide
induces physical proximity between CD14 and Toll-like receptor 4 (TLR4) prior
to nuclear translocation of NF-␬B. J. Immunol. 165:3541.
Basu, S., R. J. Binder, T. Ramalingam, and P. K. Srivastava. 2001. CD91 is a
common receptor for heat shock proteins gp96, hsp90, hsp70, and calreticulin.
Immunity 14:303.
Kang, D., G. Liu, A. Lundstrom, E. Gelius, and H. Steiner. 1998. A peptidoglycan recognition protein in innate immunity conserved from insects to humans.
Proc. Natl. Acad. Sci. USA 95:10078.
Dziarski, R., R. I. Tapping, and P. S. Tobias. 1998. Binding of bacterial peptidoglycan to CD14. J. Biol. Chem. 273:8680.
Ingalls, R. R., B. G. Monks, R. J. Savedra, W. J. Christ, R. L. Delude,
A. E. Medvedev, T. Espevik, and D. T. Golenbock. 1998. CD11/CD18 and CD14
share a common lipid A signaling pathway. J. Immunol. 161:5413.
Sato, S., F. Nomura, T. Kawai, O. Takeuchi, P. F. Muhlradt, K. Takeda, and
S. Akira. 2000. Synergy and cross-tolerance between Toll-like receptor (TLR 2and TLR4-mediated signaling pathways. J. Immunol. 165:7096.
Olson, J. C., J. E. Fraylick, E. M. McGuffie, K. M. Dolan, T. L. Yahr,
D. W. Frank, and T. S. Vincent. 1999. Interruption of multiple cellular processes
in HT-29 epithelial cells by Pseudomonas aeruginosa exoenzyme S. Infect. Immun. 67:2847.
Daeron, M., O. Malbec, S. Latour, M. Arock, and W. H. Fridman. 1995. Regulation of high-affinity IgE receptor-mediated mast cell activation by murine lowaffinity IgG receptors. J. Clin. Invest. 95:577.
Downloaded from http://www.jimmunol.org/ by guest on June 18, 2017
35.
down-regulation of surface Toll-like receptor 4 expression. J. Immunol.
164:3476.
Li, L., S. Cousart, J. Hu, and C. E. McCall. 2000. Characterization of interleukin-1 receptor-associated kinase in normal and endotoxin-tolerant cells. J. Biol.
Chem. 275:23340.
Mizel, S. B., and J. A. Snipes. 2002. Gram-negative flagellin-induced self-tolerance is associated with a block in IRAK release from TLR5. J. Biol. Chem.
277:22414.
Sato, S., O. Takeuchi, T. Fujita, H. Tomizawa, K. Takeda, and S. Akira. 2002. A
variety of microbial components induce tolerance to lipopolysaccharide by differentially affecting MyD88-dependent and -independent pathways. Int. Immunol.
14:783.
Wenzel, R. P. 1992. Anti-endotoxin monoclonal antibodies: a second look.
N. Engl. J. Med. 326:1151.
Woods, D. E., M. S. Schaffer, H. R. Rabin, G. D. Campbell, and P. A. Sokol.
1986. Phenotypic comparison of Pseudomonas aeruginosa strains isolated from
a variety of clinical isolates. J. Clin. Microbiol. 24:260.
Nicas, T. I., D. W. Frank, J. D. Lile, and B. H. Iglewski. 1985. Role of exoenzyme
S in chronic Pseudomonas aeruginosa lung infections. Eur. J. Clin. Microbiol.
4:175.
Grimwood, K., M. To, H. R. Rabin, and D. E. Woods. 1989. Subinhibitory antibiotics reduce Pseudomonas aeruginosa tissue injury in the rat lung model.
J. Antimicrob. Chemother. 24:937.
Woods, D. E., W. S. Hwang, M. S. Shahrabadi, and J. U. Que. 1988. Alteration
of pulmonary structure by Pseudomonas aeruginosa exoenzyme S. J. Med. Microbiol. 26:131.
Frithz-Lindsten, E., Y. Du, R. Rosqvist, and A. Forsberg. 1997. Intracellular
targeting of exoenzyme S of Pseudomonas aeruginosa via type III-dependent
translocation induces phagocytosis resistance, cytotoxicity and disruption of actin
microfilaments. Mol. Microbiol. 25:1125.
Wurtele, M., E. Wolf, K. J. Pederson, G. Buchwald, M. R. Ahmadian,
J. T. Barbieri, and A. Wittinghofer. 2001. How the Pseudomonas aeruginosa
ExoS toxin downregulates Rac. Nat. Struct. Biol. 8:23.
Knight, D. A., and J. T. Barbieri. 1997. Ecto-ADP-ribosyltransferase activity of
Pseudomonas aeruginosa exoenzyme S. Infect. Immun. 65:3304.
Yahr, T. L., L. M. Mende-Mueller, M. B. Friese, and D. W. Frank. 1997. Identification of type III secreted products of the Pseudomonas aeruginosa exoenzyme S regulon. J. Bacteriol. 179:7165.
Vallis, A. J., T. L. Yahr, J. T. Barbieri, and D. W. Frank. 1999. Regulation of
ExoS production and secretion by Pseudomonas aeruginosa in response to tissue
culture conditions. Infect. Immun. 67:914.
Goehring, U. M., G. Schmidt, K. J. Pederson, K. Aktories, and J. T. Barbieri.
1999. The N-terminal domain of Pseudomonas aeruginosa exoenzyme S is a
GTPase-activating protein for Rho GTPases. J. Biol. Chem. 274:36369.
Bruno, T. F., D. E. Buser, R. M. Syme, D. E. Woods, and C. H. Mody. 1998.
Pseudomonas aeruginosa exoenzyme S is a mitogen but not a superantigen for
human T lymphocytes. Infect. Immun. 66:3072.
Epelman, S., G. G. Neely, L. L. Ma, M. Gjomarkaj, E. Pace, M. Melis,
D. E. Woods, and C. H. Mody. 2002. Distinct fates of monocytes and T cells
directly activated by Pseudomonas aeruginosa exoenzyme S. J. Leukocyte Biol.
71:458.
Bruno, T. F., D. E. Woods, D. G. Storey, and C. H. Mody. 1999. Recombinant
Pseudomonas exoenzyme S and exoenzyme S from Pseudomonas aeruginosa
strain DG1 share the ability to stimulate T lymphocyte proliferation. Can.
J. Microbiol. 45:607.
Epelman, S., T. F. Bruno, G. G. Neely, D. E. Woods, and C. H. Mody. 2000.
Pseudomonas aeruginosa exoenzyme S induces transcriptional expression of proinflammatory cytokines and chemokines. Infect. Immun. 68:4811.
Bruno, T. F., D. E. Woods, and C. H. Mody. 2000. Exoenzyme S induces apoptosis of T lymphocytes. J. Leukocyte Biol. 67:808.
Knight, D. A., V. Finck-Barbancon, S. M. Kulich, and J. T. Barbieri. 1995.
Functional domains of Pseudomonas aeruginosa exoenzyme S. Infect. Immun.
63:3182.
Cormack, B. P., R. H. Valdivia, and S. Falkow. 1996. FACS-optimized mutants
of the green fluorescent protein (GFP). Gene 173:33.
Hirschfeld, M., Y. Ma, J. H. Weis, S. N. Vogel, and J. J. Weis. 2000. Cutting
edge: repurification of lipopolysaccharide eliminates signaling through both human and murine Toll-like receptor 2. J. Immunol. 165:618.
Vincent, T. S., J. E. Fraylick, E. M. McGuffie, and J. C. Olson. 1999. ADPribosylation of oncogenic Ras proteins by Pseudomonas aeruginosa exoenzyme
S in vivo. Mol. Microbiol. 32:1054.
Hacker, H., H. Mischak, T. Miethke, S. Liptay, R. Schmid, T. Sparwasser,
K. Heeg, G. B. Lipford, and H. Wagner. 1998. CpG-DNA-specific activation of
antigen-presenting cells requires stress kinase activity and is preceded by nonspecific endocytosis and endosomal maturation. EMBO J. 17:6230.
Pol, A., A. Lu, M. Pons, S. Peiro, and C. Enrich. 2000. Epidermal growth factormediated caveolin recruitment to early endosomes and MAPK activation: role of
cholesterol and actin cytoskeleton. J. Biol. Chem. 275:30566.
Lehner, M. D., S. Morath, K. S. Michelsen, R. R. Schumann, and T. Hartung.
2001. Induction of cross-tolerance by lipopolysaccharide and highly purified lipoteichoic acid via different Toll-like receptors independent of paracrine mediators. J. Immunol. 166:5161.
Tominaga, K., S. Saito, M. Matsuura, and M. Nakano. 1999. Lipopolysaccharide
tolerance in murine peritoneal macrophages induces downregulation of the lipo-
ExoS ACTIVATES TLRs