| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
* Microbiology and Infectious Diseases,
Medical Sciences, and
Internal Medicine, University of Calgary, Alberta, Canada; and
Istituto di Fisiopatologia Respiratoria, CNR, Palermo, Italy
Correspondence: Christopher H. Mody, M.D., FRCPC, Rm. 273, Heritage Medical Research Building, 3330 Hospital Dr., N.W., University of Calgary, Calgary, Alberta, Canada, T2N-4N1. E-mail: cmody{at}ucalgary.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and IL-6. This study identifies a unique property of a
gram-negative-derived microbial product capable of activating multiple
cell types and suggests a mechanism by which exoenzyme S contributes to
the immunopathogenesis of cystic fibrosis and sepsis in patients
infected with P. aeruginosa.
Key Words: gram negative • inflammation • apoptosis • mitogen
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In an attempt to understand the factors that contribute to inflammation in response to gram-negative bacteria, we have been studying exoenzyme S (ExoS), a unique P. aeruginosa virulence factor that is produced by many Pseudomonas isolates from diverse tissue origins and virtually all pneumonia and cystic fibrosis (CF) pulmonary isolates [7 ]. Increased levels of ExoS correlate with increased pulmonary damage in animal models and CF patients [7 8 9 10 11 ]. Moreover, intratracheal injection of ExoS into rats produces a massive accumulation of neutrophils within the lungs, resembling the histopathology of the lungs during the early stages of P. aeruginosa infection in CF patients, suggesting that ExoS contributes to the inflammatory milieu [10 ]. In addition, P. aeruginosa is an important pathogen in sepsis, which is characterized by tremendous induction of proinflammatory cytokines [12 , 13 ]. To our knowledge, there are no studies linking other virulence factors besides LPS to the induction of proinflammatory cytokines in the pathogenesis of P. aeruginosa septic shock [14 15 16 ]. Therefore, the question remains: Does ExoS cause the production of cytokines that could contribute to inflammation and sepsis, and what cells are required for this effect?
ExoS is a complex, bifunctional toxin, possessing two distinct methods of action: extracellular and intracellular. ExoS can be translocated directly into target cells (intracellular), resulting in cellular cytotoxicity, and can also interact with target cells by acting as a soluble (extracellular) toxin [10 , 17 18 19 20 21 22 23 24 25 ].
Extracellular ExoS is mitogenic for T cells in human and murine systems [26 27 28 ]. Extracellular ExoS induces rapid up-regulation of CD69 on T cells, leading to moderate T-cell proliferation and subsequent T-cell apoptosis [28 29 30 ]. Induction of T-cell apoptosis may contribute to the defective T-cell responses that have been observed in CF patients [31 ]. When delivered extracellularly, ExoS has potent immunomodulatory effects, which are the focus of this study.
Despite inducing apoptosis, ExoS may also contribute to the
pathogenesis of P. aeruginosa-mediated inflammatory
responses through its ability to induce transcription of
proinflammatory mediators such as tumor necrosis factor- (TNF-),
interleukin (IL)-1, and IL-8 [32
]. Because ExoS is a
T-cell mitogen, we were surprised to find that the induction of
proinflammatory cytokine/chemokine genes was more pronounced than
T-helper cell type 1 (Th1) cytokine genes [32
]. Hence,
the cytokine/chemokine profile and kinetics of induction are more
consistent with a strong monocytic stimulus rather than the polarizing
effects of a T-cell mitogen [33
34
35
]. Because the
pulmonary compartment is rich in cells of the monocytic lineage, and
macrophages appear to be the primary source of proinflammatory
cytokines during pulmonary infection by P. aeruginosa, it
suggests that cells of the monocytic lineage may be the source of
ExoS-induced proinflammatory cytokines [36
].
To determine whether the previously described mRNA induction of proinflammatory cytokines and chemokines results in the production of proteins with biological activity, peripheral blood mononuclear cells (PBMC) were stimulated with ExoS; cytokine production and bioactivity were monitored by fluorescein-activated cell sorter (FACS) analysis, bioassay, and chemotaxis assays. To determine the cytokine cellular source, PBMC were separated into T cells and monocytes after stimulation, and RNase protection assay was performed. Using highly purified monocytes, highly purified T cells, as well as monocytic and T-cell lines, the independent effects on cellular activation, cytokine production, and apoptosis were determined, and the mechanism by which monocytes costimulate T cells was established.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
production induced by LPS (10 µg/ml), having no effect on
ExoS-induced TNF- production [32
]. P.
aeruginosa HABS-10 LPS was purchased from Sigma Chemical Co.
Preparation of recombinant ExoS (rExoS) and native P.
aeruginosa ExoS (ExoS/DG1)
rExoS was isolated from an Escherichia coli strain
BL21(DE3) pLysS bearing a plasmid-encoding, histidine-tagged ExoS
cloned from P. aeruginosa strain 388 (pETrHisExoS). rExoS
was purified by Ni2+-affinity chromatography
from cellular lysates and migrated as a 52-kDa band possessing
adenosine 5'-diphosphate (ADP)-ribosyl transferase activity
[26
, 37
]. ExoS/DG1 from P.
aeruginosa strain DG1 was isolated as described previously using
(NH4)2SO4 precipitation of culture
supernatants, ion exchange chromatography, and acetone precipitation,
followed by gel filtration, and migrated as a 50-kDa band without
ADP-ribosyl transferase activity [38
]. Monoclonal
antibodies (mAb) generated against ExoS/DG1 were shown to neutralize
T-cell activation induced by rExoS, indicating that both preparations
share the domain responsible for T-cell activation [26
].
Cell-purity analysis, intracellular cytokine detection, and CD69
up-regulation
To determine the purity of cell populations, cells were
collected and washed in ice-cold FACS wash buffer [phosphate-buffered
saline (PBS), 1% fetal bovine serum (FBS), 0.l% NaN3]
three times. PBMC were incubated for 25 min at 4°C with labeled
antibodies [anti-CD3-PerCP, anti-CD19-PE, or CD14-fluroescein
isothiocyanate (FITC; Becton-Dickinson, San Jose, CA)] and were washed
three more times in FACS buffer, fixed in 1% buffered formalin, and
fluorescence-measured by flow cytometric analysis (FACScan;
Becton-Dickinson) with 10,000–20,000 events analyzed. To determine
CD69 up-regulation, cells were collected and washed as above. Cells
were labeled with anti-CD3-PerCP and anti-CD69-PE or an isotype-matched
control antibody [immunoglobulin G (IgG; Becton-Dickinson)]. Cells
were washed, fixed, and analyzed as above. The percentage of
CD3+ T cells expressing CD69 was determined by the
following formula (%CD3+CD69+ cells) -
(%CD3+IgG+ cells).
To determine intracellular cytokine levels, cells were stimulated for
4 h with ExoS/DG1 or rExoS in the presence of monensin, collected,
and labeled for cell-surface molecules as above (if necessary). Cells
were then fixed in CytoFix-CytoPerm® for 20 min and washed in
PermWash® buffer twice, according to the manufacturer’s instruction
(PharMingen, Mississauga, ON). Cells were then incubated with
anti-TNF--PE, anti-IL-8-PE, or isotype-matched control antibodies
for 30 min and washed twice in PermWash® buffer. Cells were
resuspended in FACS buffer, and fluorescence was measured by flow
cytometric analysis. The values obtained from the control antibodies
were subtracted from the test group in each experiment.
TNF- bioassay
The WEHI 164.13 murine fibrosarcoma cell line (kindly donated by
Dr. Pere Santamaria) was used to determine secretion of biologically
active TNF- as described [39
]. Briefly, the WEHI
164.13 cells were maintained in media containing RPMI-1640, 10%
heat-inactivated FBS, 2 mM L-glutamine, 100 units/ml penicillin, 100
µg/ml streptomycin, 0.2 µg/ml amphotericin B, 1 mM sodium pyruvate,
and 1 mM nonessential amino acids. WEHI 164.13 cells were resuspended
in media containing 5 ng/ml actinomycin D (Calbiochem, San Diego, CA),
and in quadruplicate, WEHI 164.13 cells were added to wells containing
test supernatant or TNF- standard (R&D Systems, Minneapolis, MN) and
were incubated for 20 h at 37°C and 5% CO2. The
assay was developed by addition of
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT;
Sigma Chemical Co.) and incubated for 4 h. Isopropanol/0.04 M HCl
was added to dissolve crystals of oxidized MTT. After overnight
incubation, the assay was read spectrophotometrically at 550 nm.
Chemotaxis assays
Chemotaxis assays were performed using a 48-well microchemotaxis
chamber with 3 µm polycarbonate filters for neutrophils and 8 µm
filters for T cells (Neuro Probe, Bethesda, MD). Neutrophils were
isolated using Dextran sedimentation (6% Dextran in normal saline) for
45 min at room temperature [in ethylenediaminetetraacetate (EDTA)],
followed by Ficoll-Hypaque separation. Red blood cells were removed
from the neutrophil fraction by hypotonic lysis. In duplicate, the top
wells were filled with 50 µl 2 x 105 neutrophils or
PBMC, and the bottom well was filled with media alone or supernatants
from ExoS-stimulated PBMC. Cells were allowed to migrate for 1 h
at 37°C. The transmigrated cells on the filter were stained by
Dif-Quick® (Becton-Dickinson), and two random, high-power fields were
enumerated.
Positive selection
To determine the cellular source of cytokine production, PBMC
were stimulated, and the cells were purified by positive selection.
Stimulated PBMC were collected and washed twice in buffer (PBS, 1%
FBS, and 20 mM EDTA). PBMC were incubated with antibodies conjugated to
magnetic beads (anti-CD14 or anti-CD3) for 15 min, and the PBMC were
washed twice more in buffer. PBMC were then passed through a
positive-selection column (Miltenyi Biotech, Auburn, CA). Typically,
positive selection resulted in >95% purity as assessed by flow
cytometric analysis.
Negative selection
To determine the requirement for monocytes in T-cell responses,
CD14+ monocytes were purified by positive selection
(above). CD3+ T cells were purified by using
CD14+-depleted PBMC, and these cells were depleted further
by using the Pan T-cells isolation cocktail, using antibody-conjugated
beads and a depletion column (Miltenyi Biotech). Typically, depletion
by this method resulted in >99% purity of CD3+ T cells,
as assessed by flow cytometric analysis.
Monocyte—T-cell coculture
To assess whether monocytes were required for T-cell activation
and whether T cells were required for monocyte activation in response
to ExoS, purified CD3+ T cells and CD14+
monocytes were mixed together at various ratios and stimulated with 1
µg/ml rExoS or ExoS/DG1 in 10 µg/ml polymyxin B. In some
experiments, monocytes were first pretreated with 2 µg/ml rExoS in 10
µg/ml polymyxin B. After 30 min or 3 h, rExoS-conditioned
supernatants, rExoS-activated monocytes, or fixed rExoS-activated
monocytes (1% buffered formalin for 20 min on ice) were transferred
(1:1 ratio) to unactivated T cells, and CD69 up-regulation was assessed
3 h later. Antibodies to TNF-, IL-1ß, IL-6, IL-12, and IL-15
were used to neutralize soluble cytokines and were compared with an
isotype-matched control antibody (all from R&D Systems). Antibodies
were used at a final concentration of 10 µg/ml.
RNA extraction and RNase protection assay
RNA was extracted using Qiagen RNA isolation spin columns
(Qiagen Inc., Mississauga, ON). RNase protection assay was performed
according to the manufacturer’s instructions (PharMingen) and as
described previously [32
].
Analysis of hypodiploid cells
Apoptosis in response to ExoS was analyzed by DNA content as
described previously [29
]. Briefly, cells were collected
and washed two times in FACS buffer. To the cell pellets, ice-cold 70%
ethanol was added slowly, and the cells were incubated at -20°C for
a minimum of 1 h. Cells were then washed two additional times with
FACS buffer, and the cell pellets were then incubated in a DNA-labeling
solution [PBS, 50 µg/ml propidium iodide (PI), 500 µg/ml RNAse A;
Sigma Chemical Co.] for 15 min at room temperature. Cells were
analyzed within 1 h by flow cytometric analysis.
Statistical analysis
To analyze the data in Figures 2
and 3
, we performed one-way
analysis of variance (ANOVA) and the Fisher’s protected least
significant difference (PLSD). For experiments in Figure 8a
one-tailed Student’s t-test was used to determine if the
means of experimental groups were significantly different from one
another, and the P values are given. Data in these figures
are shown as the mean ± SE. P < 0.05
was considered significant.
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and IL-8 protein, and IL-8 [15
,
16
, 40
41
42
]. The ability of rExoS and
ExoS/DG1 to induce production of TNF- and IL-8 proteins was examined
using flow cytometric analysis. The two different preparations of ExoS
induced dose-dependent production of TNF- and IL-8 protein. Cytokine
induction could be observed at concentrations as little as 10 ng/ml
ExoS (Fig. 1A
and B). Although all experimental groups in this
figure and in the study contained polymyxin B (10 µg/ml), additional
studies were performed to ensure that contaminating LPS was not
affecting cytokine production. As expected, polymyxin B abrogated
LPS-induced TNF- production but did not affect ExoS/DG1-induced
(Fig. 1C)
or rExoS-induced (unpublished results) TNF- production,
demonstrating that the effect was not because of contaminating LPS.
|
and IL-8 to a
similar extent.
ExoS-induced cytokines and chemokines are biologically active
Although protein production is strongly suggestive of bioactivity,
there are circumstances where protein production does not result in
bioactivity [45
46
47
]. Additionally, ExoS is known to
modify several serum proteins, leaving open the possibility that
secreted cytokines could be inactivated by modification
[19
]. PBMC were stimulated with ExoS/DG1, and
supernatants were collected at various times after stimulation. To
verify TNF- bioactivity, we performed experiments using the
TNF--sensitive fibrosarcoma cell line WEHI 164.13. Although TNF-
production peaked at 24 h, induction was rapid, with significant
production at 3 h (Fig. 2 A
). ExoS/DG1 induced the production of biologically active TNF-
in a dose-dependent manner (Fig. 2B)
. Supernatants from ExoS-stimulated
PBMC were also tested for their ability to induce neutrophil and
lymphocyte chemotaxis, two immunological processes important in the
pathogenesis of Pseudomonas infections and CF
[40
, 48
, 49
]. ExoS/DG1 and
rExoS induced neutrophil and lymphocyte chemotaxis. ExoS/DG1 induced
chemotactic factors rapidly (3 h), and this activity was still present
after three days (Fig. 3
). Neither ExoS/DG1 nor rExoS alone was chemotactic for neutrophils
or T cells when they were added directly to the chemotaxis chamber
[the number of transmigrated neutrophils/high-power field (media
alone=24.0±4.58) vs. (media+1 µg/ml ExoS/DG1=10.6±1.85) and
(media+1 µg/ml rExoS=22.8±1.78)]. Cellular chemotaxis and
TNF- production were characterized by rapid induction that slowly
decreases over time, which paralleled mRNA induction of TNF- and
chemokine genes [32
]. In addition, neutralizing
antibodies to IL-8 abrogated the neutrophil chemotactic response only
partially (48.9±5.0%), which was expected because of the large number
of chemotactic factors induced by ExoS [32
].
CD14+ monocytes are the predominant source of
proinflammatory cytokines and chemokines
Previously, we had demonstrated that ExoS is mitogenic for T cells
[28
]. However, strong proinflammatory cytokine
production and modest T-cell (Th1) cytokine production suggested that
cells other than T cells might be responsible for cytokine production
[32
]. To identify the cellular source of the
ExoS/DG1-induced cytokines, PBMC were stimulated for 5 and 24 h
with ExoS/DG1. After stimulation, CD14+ monocytes and
CD3+ T cells were purified, and multiprobe RNAse protection
assay (RPA) was performed using three probe template sets. ExoS induced
mRNA expression for the cytokines (TNF-, IL-1, IL-1ß, IL-6,
IL-12p40, and IL-10) and chemokines [IL-8, macrophage inflammatory
protein (MIP)-1, MIP-1ß, I-309, and macrophage chemotactic protein
(MCP-1)] from CD14+ monocytes after 5 h (Fig. 4A
4B
4C
) and 24 h (unpublished results) of incubation. To
verify whether TNF- mRNA expression by monocytes resulted in protein
production, we performed intracellular cytokine analysis on these cells
after 3 h stimulation with ExoS/DG1 (Fig. 4D)
and rExoS
(unpublished results). Dose-dependent TNF- protein production was
observed only in CD14+ monocytes and not in
CD3+ T cells, although both cell populations are capable of
producing TNF- [50
]. In addition, ExoS induced
TNF- production in freshly isolated monocytes and monocyte-derived
macrophages, indicating that both monocyte cell populations were
responsive to ExoS (unpublished results).
|
in response to rExoS
or ExoS/DG1, and addition of purified CD3+ T cells did not
affect the TNF- production from monocytes (Fig. 5 A
). rExoS and ExoS/DG1 induced dose-dependent induction of
TNF- from highly purified CD14+ monocytes (Fig. 5B)
.
|
from a
monocytic cell line that was devoid of T cells, we stimulated THP-1
cells with ExoS. ExoS/DG1 induced TNF- from THP-1 cells (Fig. 5C)
.
Similar results were observed with rExoS (unpublished results). Thus,
ExoS stimulates monocytes directly and does not require the
contribution of T cells.
CD3+ T cells are activated independently by ExoS, but
maximal activation requires the presence of CD14+ monocytes
Because we have shown that ExoS is a powerful stimulus for
monocytes, we considered the possibility that the ExoS-induced T-cell
activation observed in earlier studies was solely a result of the
effects of ExoS-activated monocytes on T cells [28
].
Other monocytic stimuli such as LPS, which are not thought to activate
T cells directly, can induce T-cell activation by activating monocytes
first, which then activate T cells via a B7/CD28 contact-dependent
costimulatory mechanism [34
]. To determine if monocytes
were involved in early T-cell activation, purified CD14+
monocytes and highly purified CD3+ T cells (>99%
CD3+; <<1% CD14+, CD19+) were
mixed at various ratios and stimulated with ExoS/DG1. ExoS/DG1 induced
CD69 on highly purified CD3+ T cells, however when
increasing numbers of CD14+ monocytes were added, an
increased percentage of T cells expressed CD69 (Fig. 6 A
). ExoS (rExoS and ExoS/DG1) was capable of activating highly
purified CD3+ T cells directly in a dose-dependent manner
(Fig. 6B)
. The preparation of CD3+ T cells was free of
accessory cell contamination, because these purified CD3+ T
cells did not proliferate to mitogenic stimulation using concanavalin,
but proliferation was restored when CD14+ monocytes were
added to the cultures (unpublished results).
|
Soluble factors are responsible for monocyte-mediated enhancement
of T-cell activation by ExoS
Having demonstrated that monocytes enhance T-cell activation,
experiments were undertaken to determine the mechanism by which
monocytes enhance T-cell activation in response to ExoS. We anticipated
that the effect would be contact-dependent, as it is for LPS
[34
]. Purified CD14+ monocytes were
stimulated with rExoS for 30 min or 3 h. The monocytes or the
supernatants were then added to purified CD3+ T cell, and
CD69 up-regulation on the CD3+ T cells was evaluated after
an additional 3 h. After 30 min of stimulation, only
rExoS-activated monocytes and not their supernatants could induce CD69
up-regulation on resting T cells (Fig. 7
). This suggested that the factor was contact-dependent or that the
factor had not been released into the supernatants at 30 min and that
monocytes, which were placed in culture with T cells, subsequently
released the factor. However, after 3 h, rExoS-activated monocytes
and their rExoS-conditioned monocyte supernatants possessed the ability
to activate resting T cells (Fig. 7)
. Additionally, the ability of
rExoS-conditioned monocytes to activate resting T cells was ablated by
fixation in 1% buffered formalin. This is most consistent with a
mechanism in which a soluble factor was released by monocytes. The
factor was not provided by cell contact, because fixed cells were
incapable of mediating the effect, and supernatants were fully capable
of providing the effect.
|
, IL-1ß, IL-6, and IL-15)
[50
]. TNF- and IL-6 were the critical cytokines
involved in rExoS-induced up-regulation of CD69 on T cells, because
neutralizing antibodies directed against them blocked rExoS-induced
CD69 up-regulation by 83.6% ± 12.0% and 43.6 ± 9.0,
respectively (Fig. 8
). Although it is likely that these cytokines were produced by
monocytes and acted directly on T cells, additional experiments were
undertaken to show a direct effect of monokines on T cells. Purified
CD14+ monocytes were stimulated with rExoS for 3 h,
and the supernatants were treated with an isotype-matched control
antibody (10 µg/ml) or anti-TNF-, anti-IL-1ß, and IL-6 (3.33
µg/ml each). These supernatants were added to purified
CD3+ T cells for an additional 3 h. This combination
of anticytokine antibodies blocked CD69 induction (CD69 up-regulation
was neutralized by 98.2%±28.8%; n=3 vs. control
antibody). Additionally, neutralization of IL-12 did not block CD69
up-regulation in response to rExoS (unpublished results).
ExoS induces T-cell death directly
We have demonstrated previously that ExoS activates T cells,
resulting in DNA hypodiploidy, DNA fragmentation, and
phosphatidylserine translocation to the outer leaflet of the plasma
membrane, all of which are characteristic of apoptosis
[29
]. Thus, the studies above raised questions about the
ability of ExoS to induce T-cell death directly or whether monocytes
were required. When PBMC were used, rExoS induced apoptosis (a 27.7%
increase in the hypodiploid PBMC; Fig. 9 A
). When a highly purified population of T cells was used, rExoS
caused an increase in the number of apoptotic T cells (a 24.2%
increase in hypodiploid T cells; Fig. 9B
). There was no increase in the
death of cells whether monocytes were present or not. Thus, ExoS can
kill T cells directly.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and IL-6
secretion.
Although bacterial products from gram-positive bacteria have been shown
to activate monocyte-cytokine production and T-cell activation
[52
53
54
55
56
57
], much less is known about how products from
gram-negative bacteria, other than LPS, induce inflammatory responses
[2
]. Furthermore, strategies designed to neutralize LPS
have had only limited therapeutic value [1
], suggesting
that other products of gram-negative bacteria play an important role in
the pathogenesis of these infections and perhaps themselves can be
therapeutic targets. There are a few studies of other P.
aeruginosa products that stimulate proinflammatory cytokine
production. For example, nitrate reductase, flagellin, and porins have
been shown to induce TNF- from macrophages [6
,
58
, 59
], and pyocyanin induces IL-8
production from epithelial cell lines [4
]. The role of
exotoxin A is controversial, because it has been demonstrated to act as
a T-cell superantigen and induce TNF- and IL-1 from monocytes but
also because it inhibits T-cell responses and cytokine production
[60
, 61
]. Together, there is only a small
list of factors from P. aeruginosa capable of inducing
inflammatory responses directly, and to date, their characterization is
incomplete.
The proinflammatory responses of gram-positive products have been elucidated more clearly. For example, Staphylococcus and Streptococcus secrete a host of superantigens that induce cytokine production from T cells and monocytes, as well as induce T-cell apoptosis [33 , 53 , 62 ]. Although we have shown that ExoS is not a superantigen [28 ], it does possess similar characteristics to superantigens. ExoS and superantigens induce T-cell activation, T-cell-independent monocyte-cytokine production, and accessory cell-dependent T-cell proliferation (although T-cell activation is only partially accessory cell-dependent) [28 , 30 , 53 ]. However, superantigens are potent inducers of T-cell cytokines, and ExoS induces only very weak induction of T-cell cytokines. In addition to the superantigens from gram-positive bacteria, lipotechoic acid (LTA) and peptidoglycan (PepG) have been shown to be mitogenic for T cells and have well-characterized abilities to induce cytokine production from monocytes [52 , 54 , 55 ]. These molecules are recognized by a diverse group of pattern-recognition receptors present on monocytes and T cells [54 , 63 , 64 ]. It is not known whether ExoS binds to some form of a pattern-recognition receptor.
We have shown previously that ExoS induces mRNA for multiple
proinflammatory cytokines and chemokines [32
], however
cytokine expression can be controlled at multiple levels
[65
66
67
], and in some cases, neither mRNA induction nor
translation guarantees a biologically active protein
[68
]. Additionally, ExoS is capable of ADP-ribosylating
serum antibody molecules giving rise to the possibility that ExoS could
inactivate cytokines, which would complement the ability of P.
aeruginosa alkaline protease and elastase, which also inactivate
cytokines [19
, 46
, 69
,
70
]. The current studies indicate that ExoS induced
protein production of TNF- and IL-8 and that both of these cytokines
were secreted in a biologically active form.
We have observed previously that ExoS is mitogenic for human and murine
T cells [27
, 28
, 30
]. The
cytokine profiles induced by T-cell mitogens and superantigens have
been well-characterized, and cytokines such as TNF- and IL-8 can be
induced from T cells and monocytes [33
,
53
]. Based on these previous observations and the potent
ability of ExoS to activate T cells [28
,
30
], we were surprised to find that the proinflammatory
cytokines and chemokines induced by ExoS were from monocytes, and
virtually none came from T cells. In addition, ExoS activated monocytes
directly, in a T-cell-independent manner. Therefore, tremendous early
T-cell activation had no effect on early monocyte-cytokine production.
These results left open the possibility that the T-cell activation
induced by ExoS may simply be a byproduct of a strong and solely
monocytic stimulus. LPS is known to induce T-cell activation by a
similar mechanism; LPS directly activates monocytes to up-regulate B7,
and then through a contact-dependent CD28/B7 mechanism, LPS-activated
monocytes induce T-cell activation [34
]. The T-cell
response induced by ExoS is characterized by activation of a very large
percentage of T cells that results in a modest, proliferative response
and marked apoptosis [27
28
29
30
]. Stimulated monocytes are
capable of inducing T-cell activation [34
], however T
cells are stimulated via two mechanisms by ExoS. The first is a direct
mechanism, where highly purified peripheral T cells and the Jurkat
T-cell line were directly activated by ExoS. The second is a
monocyte-dependent mechanism that complements the ability of ExoS to
activate T cells. It is interesting that this mechanism is distinct
from the contact-dependent B7/CD28 mechanism used by LPS, as it relies
on the secretion of soluble factors that were identified as TNF- and
IL-6.
The death of ExoS-activated T cells could also be induced directly and
did not require the presence of monocytes. This was also a surprising
result, because the early stages of T-cell activation (CD69
up-regulation) were in part monocyte-dependent, and activated monocytes
can induce T-cell death through a TNF--dependent mechanism
[71
]. Because monocytes are not required for T-cell
activation or T-cell death but are required for T-cell proliferation
[28
], this suggests that monocytes may provide a
survival/growth signal to activated T cells. Activation of T cells and
monocytes by ExoS produced very different results (proinflammatory
cytokine production vs. apoptosis). This is somewhat analogous to the
paradoxical effects of TNF-, which can enhance activation, induce
inflammatory responses, or induce apoptosis, depending on the cell
type, activation state of that cell, and the context in which TNF-
is seen [72
, 73
].
ExoS possesses a C' ADP-ribosyl-transferase domain and an N' GTPase-activating protein (GAP) for the Rho family of GTPases. When translocated intracellularly by a bacterial type III secretion system, intracellular ExoS inhibits DNA synthesis and actin polymerization [17 , 23 24 25 ]. Although we do not know whether ExoS GAP activity affects monocyte or T-cell activation, our studies clearly indicate that the ADP-ribosyl-transferase activity possessed by ExoS is not required for monocyte-cytokine production, T-cell activation, and death, as all were induced with an enzymatically inactive preparation.
The mechanism and biological outcomes of monocyte and T-cell activation
by ExoS correlate with clinical characteristics present in patients
infected with P. aeruginosa. We speculate that in the
context of pulmonary infection by P. aeruginosa, ExoS
contributes to the inflammatory response by activating alveolar
macrophages as well as newly recruited monocytes or macrophages to
produce proinflammatory cytokines and chemokines, which together create
the inflammatory milieu. The chemokine profile induced by ExoS suggests
that neutrophils, T cells, and monocytes would be recruited to the lung
[32
], and we have demonstrated that ExoS-induced
chemotactic factors stimulate neutrophil and T-cell chemotaxis.
Alveolar macrophages activated by ExoS would be important mediators in
pulmonary neutrophil infiltration, and in CF patients, neutrophil
infiltration correlates with increased pulmonary damage, suggesting
that ExoS-recruited neutrophils would contribute to pulmonary
inflammation [40
, 42
, 74
]. In
fact, immunocytochemistry of the bronchoalveolar lavage from CF
patients indicates that alveolar macrophages are a major source of
TNF-, IL-1, IL-6, and IL-8 during pulmonary infection
[42
], and depletion of alveolar macrophages results in
decreased proinflammatory cytokine production and decreased neutrophil
chemotaxis during infection [36
]. Activated neutrophils
and macrophages can secrete lytic enzymes that destroy pulmonary
tissue, reduce pulmonary function, and decrease pathogen clearance
[40
]. Infiltrating T cells would die by apoptosis, which
may contribute to the deficient cell-mediated immune response to
P. aeruginosa in CF patients [31
].
Infection is the leading cause of morbidity and mortality in severe
burn-wound patients, and P. aeruginosa is the most frequent
invader, making up 45% of all burn-wound infections
[12
]. Once colonized, a burn is the site of bacterial
replication that can be associated with invasion, bacteremia, and
sepsis. Lethal sepsis has traditionally been associated with production
of inflammatory mediators such as TNF-, IL-1, and IL-6 that results
in symptoms such as shock and multiorgan system failure. In animal
models, P. aeruginosa-induced sepsis can only be reduced
partially with anti-LPS antibodies, even if the antibodies are
administered prior to challenge with live bacteria, suggesting that
molecules other than LPS are contributing [16
]. In
addition, burn-wound patients and septic patients have impaired T-cell
function as well as increased thymocyte apoptosis, which in the case of
sepsis, has been shown to be independent of LPS
[75
76
77
]. It is possible that ExoS can also contribute
to the pathogenesis of burn-wound sepsis, because ExoS is released
during infection and can be detected in the serum [78
].
Systemic ExoS can induce proinflammatory cytokine production and
apoptosis from circulating monocytes and T cells, respectively
[7
].
In summary, ExoS has distinct functions depending on its presentation to target cells (extracellular vs. intracellular). These different functions are likely to work in concert to prevent clearance and favor pathogen persistence. This study has demonstrated that ExoS directly induced proinflammatory cytokine production from monocytes and T-cell apoptosis. This study also highlights the importance of thoroughly examining the different mechanisms used by gram-negative bacteria to induce and modulate immune responses.
|
ACKNOWLEDGEMENTS |
|---|
Received April 1, 2001; revised November 12, 2001; accepted November 17, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
