| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Microbiology and Immunology and Cooperative Research Center for Vaccine Technology, University of Melbourne, Parkville, Victoria, Melbourne, Australia; and
Department of Cell Biology and Immunology, Vrije Universiteit, Amsterdam, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
aroAD strain.
Ag-specific CD8+ T lymphocyte responses were analyzed using
IFN-
ELISPOT, tetramer binding, and in vivo and in vitro CTL assays.
Our results showed that aroAD and
aroADinvA induced comparable levels
of Ag-specific CD8+ T lymphocyte responses as well as
protective, Ag-specific B and CD4+ T lymphocyte immunity.
Furthermore, experiments in macrophage-depleted mice showed that
CD8+ T lymphocyte responses were effectively induced in the
absence of macrophages. Together, our results imply that in this
infection model, SPI-1-mediated cell death does not affect the
immunological defense response and is not important for the induction
of CD8+ T lymphocyte responses. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bacterial entry into macrophages (or other professional phagocytes) may
occur either through Salmonella pathogenicity island 1
(SPI-1)3-mediated
invasion or through phagocytosis. However, infection of macrophages by
Salmonella is cytotoxic, and bacterial invasion can lead to
rapid death of the phagocyte (8, 9). Several in vitro
studies have focused on the mechanism inducing host cell death, which
was originally described as programmed cell death, or apoptosis.
Further studies, however, have shown that
Salmonella-induced macrophage death is distinct from
classical apoptosis, since it involves early disruption of the plasma
membrane in the absence of caspase-3 and caspase-7 activation, and that
it shares aspects of both apoptotic and necrotic cell death
(10). Despite this controversy, it has become evident that
activation of the rapid cell death pathway is mediated by
Salmonella invasion protein B (SipB), an effector protein
exported by the SPI-1-encoded type III secretion system
(11). SipB binds to IL-1-converting enzyme caspase-1,
resulting in secretion of IL-1
and IL-18 as well as death of the
host cell. In addition, it has been shown that Salmonella is
able to induce macrophage death in a caspase-1-independent manner in a
late phase killing process more consistent with apoptosis, involving
caspase-3 processing and cytochrome c release (12, 13). Furthermore, a role for caspase-2 in both early and late
phase macrophage death has been proposed (12).
Despite the many recent studies investigating the mechanism behind
Salmonella-induced macrophage death, its relevance in the
context of in vivo pathogenicity and immunobiology associated with
Salmonella infections has not been extensively investigated.
Most of the studies mentioned above were performed in vitro, many using
macrophage-like cell lines, and a few studies have shown a cytotoxic
effect of S. typhimurium infection on macrophages
in vivo (4, 14). Using confocal microscopy, the presence
of TUNEL-positive, S.
typhimurium-harboring macrophages in Peyer’s patches or
livers of infected mice was demonstrated. SipB-mediated macrophage
death coincides with the release of proinflammatory cytokines IL-1
and IL-18, which may not be beneficial to the invading bacterium, but
may help the host to combat the infection. In a recent in vivo study,
however, it was shown that mice lacking caspase-1 are more resistant to
S. typhimurium colonization after an oral
inoculation and have a 50% lethal dose that is 1000-fold higher than
that in normal mice (14), suggesting that
Salmonella bacteria induce macrophage death and the release
of proinflammatory cytokines to establish an infection.
Nevertheless, since macrophages are the first host defense cells
encountered by S. typhimurium invading through
the gastrointestinal tract, and since Salmonella-induced
cell death coincides with the release of proinflammatory cytokines, the
observed macrophage death might also affect the induction of
Salmonella-specific immunity. In addition, it has been
suggested that uptake of infected, apoptotic macrophages may result in
cross-priming and therefore in the induction of Ag-specific
CD8+ T lymphocyte responses (15, 16). Furthermore, the induction of macrophage death and
consequently an inflammatory response may explain why attenuated
Salmonella function so well as vectors for DNA vaccines
(17). In the present study we investigated the effect of
Salmonella-induced macrophage death on the induction of
Salmonella-specific immunity. We used an S.
typhimurium strain carrying a mutation in its SPI-1-encoded
invA gene, resulting in a nonfunctional type III secretion system and
therefore an inability to actively invade into host cells and secrete
any effector proteins (18). Since the invA
mutant causes a lethal infection in susceptible
(Slc11A1-/- (formerly NRAMP)) mice, the
invA mutation was introduced into a well-characterized,
attenuated S. typhimurium vaccine strain
(aroAD) to be able to perform in vivo experiments. Our
results showed that Salmonella SPI-1-mediated macrophage
death does not affect the in vivo induction of acquired, protective
immune responses, and that Salmonella-specific
CD8+ T lymphocyte responses can be elicited in
the absence of splenic and hepatic macrophages. Together, our results
imply that in this infection model, bacterial (SPI-1)-mediated host
cell death does not affect the immunological defense response, and that
systemic macrophages are not involved in the induction of specific
acquired immunity. The implications of this finding for our current
understanding of the immunobiology of murine Salmonella
infections will be discussed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study we used the aroA aroD mutant
of S. typhimurium SL1344 (BRD509 (streptomycin resistant); a
gift from Prof. G. Dougan, Imperial College, London, U.K.), the
invA mutant of SL1344 (
4370; provided by Prof. R. Curtiss
III, Department of Biology, Washington University, St. Louis,
MO), and virulent S. typhimurium SL1344. The
invA mutation (kanamycin resistant) from 4370 was
introduced into BRD509 using bacteriophage P22
(Int-)-mediated transformation to generate
BRD509invA. Recombinant S.
typhimurium strains expressed either C fragment of tetanus
toxin (TT) from the pTETtac4 expression plasmid (e.g., BRD509/TT4)
(19) or OVA from the pKK-OVA plasmid (e.g., BRD509/OVA)
(20), or harbored the pKK233 control plasmid (e.g.,
BRD509/pKK). Since all plasmids contain the ampicillin resistance gene,
recombinant bacteria were grown in 50 µg/ml ampicillin.
Invasion assay
S. typhimurium strains were grown overnight without shaking and then subcultured until the absorbance at 600 nm reached 0.5, representing the mid-log growth phase at which the bacteria are most invasive. Monolayers of RAW264 cells or Madine-Darby canine kidney (MDCK) cells were infected in triplicate at a multiplicity of infection (MOI) of 100:1 in RPMI 1640 (Life Technologies, Grand Island, NY) for 1 h at 37°C and 5% CO2. The monolayers were then washed with RPMI 1640 supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin (RP10) and incubated in RP10 containing 100 µg/ml gentamicin at 37°C in 5% CO2. After 1.5 h the cells were washed twice in PBS and lysed with 0.1% Triton X-100 in PBS. The number of viable bacteria in the cell lysates was determined by plating serial dilutions on Luria-Bertoni agar plates.
Host cell death assays
S. typhimurium were grown statically overnight, and subcultured to grow to mid-log phase (A600 = 0.5). Monolayers of macrophages obtained by lavage of the peritoneal cavity of C57BL/6 mice were infected for 1 h with S. typhimurium at an MOI 100:1 at 37°C in 5% CO2. Bacteria were washed away from the macrophages with RP10, and the macrophages were incubated for 2 h with RP10 containing 100 µg/ml gentamicin. Subsequently, the macrophages were detached from the tissue culture trays using trypsin, washed, and analyzed for cell death changes using either propidium iodide labeling of fragmented DNA as described previously (21) or labeling of cells with fluorescently tagged annexin V (BD PharMingen, San Diego, CA) according to the manufacturer’s instructions. Cells were kept on ice and were analyzed by flow cytometry using a FACScan flow cytometer and CellQuest software (BD Biosciences, San Jose, CA).
Immunization of mice
Male C57BL/6 mice, 6–8 wk of age, were obtained from the animal care facility at Department of Microbiology and Immunology, University of Melbourne. Mice were orally inoculated with 1010 recombinant S. typhimurium as described previously (22). When indicated, mice were challenged orally with 107 wild-type S. typhimurium SL1344. As a positive control for the induction of OVA-specific CTL, mice were immunized i.v. with OVA-loaded splenocytes 7 days before analysis of CTL responses (23).
In vivo macrophage depletion
Macrophages can be depleted in vivo using dichloromethylene diphosphonate (Cl2MDP)-loaded liposomes (24). We have previously established that Cl2MDP does not affect the in vitro growth of S. typhimurium (22). Liposomes were prepared by Dr. N. van Rooijen as described previously (24); Cl2MDP was a gift from Roche (Mannheim, Germany). Mice were treated i.v. with 200 µl Cl2MDP liposomes 2 days before immunization, resulting in the elimination of macrophages from the spleen (red pulp macrophages, marginal zone macrophages, and marginal metallophilic macrophages) and the liver (Kupffer cells). Subsequently, mice received 50 µl Cl2MDP liposomes i.v. every 5 days for the duration of the experiment to remove newly immigrated macrophages. Successful (continuous) elimination of macrophages from spleen and liver was confirmed by immunohistology (24).
Viable counts of S. typhimurium in organs
At the indicated time points, organs were collected from mice, and tissue homogenates were prepared as described previously (22). Peyer’s patches and small intestines were cleaned and washed in HBSS containing 100 µg/ml gentamicin, then incubated for 1.5 h in HBSS/100 µg/ml gentamicin at 37°C to kill any extracellular bacteria before homogenizing the tissues. The number of bacteria in each organ was determined by plating serial dilutions of tissue homogenates on Luria-Bertoni agar plates containing the appropriate antibiotics.
Measurement of Ab responses by ELISA
Mice were bled from the orbital sinus at weekly intervals. Serum was analyzed for the presence of S. typhimurium LPS- or TT-specific Abs by ELISA as described by Wijburg et al. (22). Serum titers are presented as the highest dilution with an OD of 0.1 above normal mouse serum.
Single-cell ELISPOT assay for IFN--secreting cells
Microtiter 96-well plates (Maxisorp white; Nunc, Copenhagen,
Denmark) were coated overnight at 4°C with 10 µg/ml
IFN--specific mAb HB170 diluted with carbonate buffer (0.1 M
NaHCO3, pH 9.6). Plates were emptied, washed with
RP10, and blocked with RP10 for 2 h at 37°C in 5%
CO2. After washing the plates with RP10, serial
dilutions (starting at 106 cells/well) of
splenocytes from immunized mice were added in duplicate wells with
105 irradiated naive spleen cells/well and 10
U/well recombinant human IL-2 with or without Ag. To measure the number
of OVA-specific CD8+ T lymphocytes, 0.1 µg/well
OVA257–264 peptide was used as the Ag, whereas
in other assays 106 heat-killed S.
typhimurium/well were used to stimulate
Salmonella-specific CD4+ T
lymphocytes. The cells were cultured for 48 h at 37°C in 5%
CO2 and subsequently lysed by incubation for 10
min at room temperature with PBS/0.5% Tween 20, followed by three
additional washing steps with PBS/0.5% Tween 20. Plates were then
incubated for 2 h at 37°C with biotinylated anti-IFN- mAb
XMG1.2 (BD PharMingen) diluted with PBS/0.1% BSA, thoroughly washed
with PBS/0.5% Tween 20, followed by a 2-h incubation with
HRP-conjugated streptavidin (Silenus, Hawthorn, Australia). Bound Ab
was visualized using 3-amino-9-ethyl-carbazole (Sigma-Aldrich, St.
Louis, MO) in 0.1 M sodium acetate buffer (pH 5.2) and freshly added
H2O2. The reaction was
stopped by rinsing the plates in tap water, and spots were enumerated
using a magnifying glass. Presented are the number of IFN--secreting
cells per 106 cells detected when cultured in the
presence of Ag, corrected for background levels measured in the absence
of Ag, which was always between 5–10 cells/106
cells.
Binding of tetrameric MHC class I complexes
H-2KbOVA257–264 tetramers were prepared and tested by Dr. A. Brooks (University of Melbourne). Single-cell suspensions (5 x 106 cells/sample) prepared from spleens obtained from immunized mice were surface-stained for 30 min on ice with allophycocyanin-conjugated anti-CD8 mAb (clone 53.5-6.7; BD PharMingen) diluted with ice-cold PBS/1% FCS. The cells were washed in PBS/1% FCS and surface-stained for 30 min at 37°C with PE-conjugated H-2KbOVA257–264 tetramer in PBS/1% FCS. After washing in PBS/1% FCS, cells were kept on ice and acquired on a FACScan flow cytometer (BD Biosciences). The viability of the cell suspensions was analyzed using propidium iodide, which was added to the cells immediately before acquisition. The data were analyzed using CellQuest software (BD Biosciences).
Stimulation of OVA-specific CTL in vitro
Splenocytes (3 x 107) obtained from immunized mice were cultured for 6 days at 37°C in 5% CO2 with 3 x 106 irradiated (2000 rad) OVA257–264-pulsed naive spleen cells in 10 ml RP10.
51Cr release cytotoxicity assay
Viable cells were harvested from the 6-day spleen cell cultures and used as effector cells in a cytotoxicity assay in which OVA257–264-pulsed and unpulsed Na251CrO4-labeled EL4 cells were used as target cells. After 5 h the amount of 51Cr released from target cells was measured in 25 µl supernatant, using Lumaplates and a TopCount NXT Microplate Scintillation and Luminescence Counter (Packard, Canberra, Australia). Results are presented as the percentage of specific lysis defined as [(experimental lysis) - (spontaneous lysis)]/[(total detergent lysis) - (spontaneous lysis)]. Maximum spontaneous release values were always <10% of the total detergent lysis.
Measurement of OVA-specific in vivo lysis of target cells
In vivo CTL activity was determined using a method described by Barchet et al. (25), based on Ag-specific in vivo elimination of target cells labeled with fluorescent dye CFSE (Molecular Probes, Eugene, OR). Briefly, single-cell suspensions were prepared from spleens obtained from naive C57BL/6 mice. Half of the cell suspension was pulsed with OVA257–264 peptide (1 µg/ml, 107 cells/ml) for 1 h at 37°C in RP10; the other half of the cells were mock-treated. Cells were washed three times in PBS/0.1% BSA and labeled with either 10 µM CFSE (peptide-pulsed cells, CFSEhigh) or 1 µM CFSE (mock-treated cells, CFSElow) for 10 min at 37°C. Cells were washed twice in PBS/1% FCS, the two populations were mixed, and 2 x 107 target cells were injected i.v. into immunized or naive recipient mice. Mice were bled 24 h after injection, and PBMC were used for flow cytometric analysis using FACScan and CellQuest software (BD Biosciences). Analysis of in vivo elimination of CFSE-labeled target cells was performed on 10,000 acquired CFSE-labeled cells. The percentage of specific lysis was calculated as follows: [1 - ((r(primed mice)/(r(unprimed mice))] x 100, where r = (number of CFSEhigh cells/number of CFSElow cells).
Statistical analysis
Student’s t test was used to determine statistically significant differences between groups of mice.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The S. typhimurium invA mutant 4370
carries a mutation in its SPI-1-encoded invA gene and therefore has a
nonfunctional type III secretion system, resulting in an inability to
secrete effector proteins (such as SipB) into host cells. 4370 has
been well characterized for its inability to invade epithelial cells,
but animal studies have shown that inoculation of mice with this mutant
strain is lethal (18, 26). The invA mutation
from 4370 was therefore introduced into the S.
typhimurium aroA aroD mutant strain BRD509,
for which the murine immunobiology has been well established (22, 27). The results of in vitro invasion assays using epithelial
MDCK cells and the macrophage-like cell line RAW264 shows that
BRD509invA did not invade MDCK cells, whereas comparable
numbers of all Salmonella strains were taken up by RAW264
cells (Fig. 1
A). In addition,
the ability of the mutant strains to invade the host in vivo was
studied. Mice were orally inoculated with 1010
bacteria, and 20 h later bacterial numbers present in
Peyer’s patches and intestinal epithelium were established. In Fig. 1B it is shown that there was no significant difference in
the number of bacteria present in Peyer’s patches after inoculation
with either BRD509 or BRD509invA, whereas significantly
fewer (p < 0.05) bacteria were detected in the
intestinal epithelium of mice inoculated with BRD509invA
compared with BRD509. In subsequent studies it was shown that the
systemic bacterial load after oral inoculation with BRD509 and
BRD509invA was comparable, and that both
Salmonella strains were cleared from mice at the same rate
(results not shown). Together, the results of these invasion assays
showed that BRD509invA was unable to invade host cells,
but was taken up by phagocytic cells to levels comparable with those of
BRD509.
|
To establish the ability of BRD509 and BRD509invA to
induce the death of host macrophages, peritoneal macrophages were
infected with bacteria for 1 h. The cells were subsequently
cultured for 2 h, detached from the tissue culture trays, washed,
and then analyzed for cell death changes by flow cytometry, using
either staining of fragmented DNA with propidium iodide or binding of
fluorescently labeled annexin V, which binds to phosphatidylserine
present at an early stage of apoptosis in the outer membrane of cells.
In concordance with RAW264 cells (Fig. 1
A), all bacterial
strains were taken up by the peritoneal macrophages in similar numbers
(results not shown). Fig. 2 shows that
infection with BRD509 was cytotoxic to the macrophages, whereas uptake
of BRD509invA and invA mutant 4370 did
not affect the host cells.
|
To investigate the effect of the invA mutation on
the induction of Salmonella-specific immunity elicited with
vaccine strain BRD509, groups of five mice were immunized orally with
either BRD509 or BRD509invA, expressing the C fragment of
the pTETtac4 plasmid. After 4 wk the presence of
Salmonella-LPS-specific and TT-specific Abs in serum was
analyzed by ELISA. Fig. 3A
shows that both BRD509 and BRD509invA induced high levels
of LPS-specific Abs, and no significant differences were observed
between groups of mice. In addition, induction of LPS-specific Abs
occurred at the same rate in BRD509- and BRD509invA-immunized mice
(results not shown). Furthermore, TT-specific Ab titers (Fig. 3B) were similar in mice immunized with BRD509/TT4
(log103.9) compared with those in mice that
received BRD509invA/TT4 (log10
4.1).
|
ELISPOT
assay, in which spleen cells from immunized mice were restimulated with
heat-killed BRD509. Fig. 4 shows that
there was no significant difference in the number of IFN--secreting
cells in the spleens of BRD509-immunized mice compared with
BRD509invA-immunized mice. Within all groups of immunized
mice, an average of log10 2.5 to
log10 3 spleen cells per
106 spleen cells secreted IFN- in response to
in vitro stimulation with heat-killed BRD509.
|
In the next set of experiments we studied the induction of
Ag-specific CD8+ T lymphocytes. Since no
Salmonella-derived MHC class I-binding epitopes have been
mapped, we used OVA as a surrogate Salmonella Ag, expressed
by the mutant Salmonella strains from the pKK-OVA plasmid.
Four weeks after a single oral immunization, the presence of
OVA-specific CD8+ T lymphocytes in the spleen was
analyzed using an IFN- ELISPOT assay,
H-2KbOVA257–264 tetramer
binding, and in vivo CTL assays (Fig. 5).
In both BRD509/OVA and BRD509invA/OVA-immunized mice we
found comparable numbers of OVA-specific IFN--secreting
CD8+ T lymphocytes, ranging from
log10 1.3 to log10 1.8,
with one low or nonresponding mouse per group (Fig. 5A). The
number of IFN--secreting cells in spleens from mice immunized with
the control Salmonella strains (pKK) were approximately the
same background levels (log10 0.3) as those found
in naive mice, whereas IFN- responses in the positive control group
ranged from log10 1.8 to
log102.3.
|
B). Tetramer binding to spleen
cells obtained from mice immunized with BRD509/OVA or
BRD509invA/OVA was equivalent, averaging 1.4 and 1.2% of
CD8+ T lymphocytes, respectively. In the positive
control group 2.3% of the CD8+ T lymphocytes
were binding
H-2KbOVA257–264.
Finally, we investigated whether the OVA-specific
CD8+ T lymphocytes were able to lyse target cells
using an in vivo CTL assay. Fluorescently labeled, peptide-loaded
and mock-treated target cells were injected into mice 4 wk after
immunization with recombinant BRD509 or BRD509invA
strains. After 24 h lysis of target cells was determined using
flow cytometric analysis of PBL. It is evident from Fig. 5C
that although responses varied between mice in the same group,
OVA-specific CD8+ T lymphocytes induced by
immunization with either recombinant BRD509- or
BRD509invA-expressing OVA were able to lyse target cells
in an Ag-specific manner. In addition, no significant difference was
observed when comparing BRD509-immunized mice to
BRD509invA-immunized mice. Taken together, these results
suggest that Salmonella SPI-1-mediated macrophage death does
not affect the in vivo induction of either CD4+ T
lymphocyte or CD8+ T lymphocyte Ag-specific
immune responses.
Spleen and liver macrophages are not involved in the induction of Ag-specific CTL responses elicited by recombinant Salmonella mutants
Since Salmonella are generally considered to mainly
reside inside macrophages in vivo, and since macrophages are the major
target cell for Salmonella-mediated host cell death,
investigation of whether macrophages play a role in the induction of
Ag-specific CTL responses at all was warranted. In this study we used a
liposome-based macrophage depletion technique to eliminate spleen and
liver macrophages in vivo. Mice were i.v. injected with 200 µl
Cl2MDP liposomes 2 days before immunization with
recombinant Salmonella strains. This treatment resulted in
the complete elimination of macrophages residing in the red pulp and
marginal zone area of the spleen (Fig. 6)
and in the liver (results not shown). Continuous depletion of
macrophages from the spleen and liver was achieved by injection of
Cl2MDP liposomes every 5 days.
|
invA-expressing OVA. The presence of OVA-specific
CD8+ T lymphocytes in the spleens of normal and
Cl2MDP liposome-treated mice was assayed 4 wk
after immunization. Analogous to previous experiments, equivalent
numbers of IFN--secreting cells were detected in spleens of mice
immunized with either BRD509/OVA or BRD509invA/OVA (Fig. 7A). Interestingly, treatment
of mice with Cl2MDP liposomes did not affect the
number of OVA-specific, CD8+ IFN- secreting
cells in either BRD509-immunized mice or
BRD509invA-immunized mice (Fig. 7A),
suggesting that spleen and liver macrophages are not involved in the
induction of OVA-specific CD8+ T lymphocytes.
These results were confirmed by analysis of tetramer binding to spleen
cells obtained from immunized mice. Fig. 7B shows that
H-2KbOVA257–264 bound to
2.5–5% of all splenic CD8+ T lymphocytes
from mice immunized with recombinant Salmonella-expressing
OVA. Although there was some variability in binding between mice in
each group, there was no significant difference in tetramer binding
when comparing spleen cells obtained from mice immunized with
BRD509/OVA to spleen cells derived from
BRD509invA/OVA-immunized mice. No tetramer binding was
observed in spleen cell suspensions obtained from mice immunized with
control (pKK) Salmonella.
|
C). Furthermore, the results of this in vitro CTL assay
confirmed our findings using the in vivo assay, which showed that
immunization with OVA-expressing BRD509 or BRD509invA
results in equivalent levels of OVA-specific CTL responses.
In summary, the results of these experiments showed that treatment of
mice with Cl2MDP liposomes before and during
immunization with BRD509/OVA or BRD509invA/OVA did not
affect the induction of OVA-specific CD8+ T
lymphocytes and therefore suggest that splenic and hepatic macrophages
are not involved in the induction of Ag-specific CTL responses elicited
with recombinant, attenuated Salmonella strains.
Lack of Salmonella SPI-1-induced macrophage death does not affect the induction of protective immunity
Finally, we compared the ability of BRD509 and
BRD509invA to induce protective immunity. To do so, mice
were orally immunized with either 1010 BRD509 or
BRD509invA and 8 wk later were challenged with a lethal
dose of wild-type S. typhimurium SL1344. As a
control, naive mice were infected with SL1344 as well. The bacterial
load in the spleen was determined 5 days after challenge as a measure
of protection. Fig. 8 shows that naive
mice were unable to restrict the infection and had 2 x
105 SL1344 bacteria in their spleens. In
contrast, the bacterial load in BRD509-immunized mice was very low
(102/spleen), indicating that these mice were
able to control the challenge inoculum. No significant difference in
the splenic bacterial load was observed between BRD509- and
BRD509invA-immunized mice, suggesting that both
attenuated S. typhimurium strains are equally capable of
inducing protective immunity in mice.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-18, and it has
therefore been argued that induction of host cell death by bacteria, or
rather, the consequent release of cytokines, might be important in the
host defense. In this study we investigated the importance of this
Salmonella SPI-1-induced macrophage death on the induction
of acquired, protective immune responses in vivo. To be able to perform
in vivo experiments, we generated an attenuated S.
typhimurium (aroAD) vaccine strain defective
in its ability to kill host macrophages due to a mutation in its
SPI-1-encoded invA gene (aroADinvA). By
comparing the specific immune responses raised in mice against protein
Ags expressed by these attenuated S. typhimurium
strains, we were able to show that Salmonella SPI-1-induced
macrophage death does not affect the induction of serological responses
or the activation of Ag-specific CD4+ or
CD8+ T lymphocytes. In addition, we demonstrated
that the induction of protective immunity was not altered following
immunization with the aroADinvA S.
typhimurium vaccine strain. Furthermore, we showed that the
induction of Ag-specific CD8+ T lymphocyte
responses was not affected by the depletion of macrophages from the
spleen and liver before and during systemic immunization, confirming
that these cells do not play a significant role in vivo in the
induction of CD8+ T lymphocyte responses.
Acquired immunity against S. typhimurium is comprised of
both Salmonella-specific serological and T lymphocyte
responses, and in the absence of a T lymphocyte response the infection
is fatal to mice (7, 28). Immunity to reinfection is
considered to be predominantly Th1-mediated (7, 28). The
production of IFN- is necessary for control of in vivo growth of
attenuated S. typhimurium mutants (e.g., aromatic mutants),
and to effect immunity following successful immunization (7, 28). Infection also elicits MHC class I-restricted CTLs which
may help clear the infection (20, 28). In this paper we
showed that immunization with attenuated S. typhimurium
elicited both CD4+ and CD8+
IFN--producing cells. The number of IFN--secreting cells
activated in vitro with heat-killed BRD509 was 10-fold higher
compared with stimulation with OVA257–264. This
difference is most likely explained by the fact that heat-killed BRD509
activates a pool of T lymphocytes specific for a variety of
Salmonella Ags, whereas OVA257–264
only stimulates OVA257–264-specific
CD8+ T lymphocytes.
To demonstrate the presence of Salmonella-specific
CD8+ T lymphocytes, we used OVA expressed by the
S. typhimurium strains as a surrogate Salmonella
Ag. We thoroughly analyzed the presence of OVA-specific
CD8+ T lymphocytes using IFN- ELISPOT assays,
flow cytometric analysis of binding of
H-2KbOVA257–264 tetramers
and in vivo CTL assays. All three assays confirmed that immunization of
mice with S. typhimurium results in the induction
of Ag-specific CD8+ T lymphocytes that are able
to secrete IFN- and lyse target cells in vivo in an Ag-specific
manner. However, the means by which the endosome-bound bacterium gains
access to the MHC class I processing pathway are not clear. In a recent
in vitro study it was shown that following infection by S.
typhimurium, fragments of dead (apoptotic) macrophages were taken
up by bystander dendritic cells, resulting in presentation of
bacteria-encoded Ags to Ag-specific CD4+ and
CD8+ T hybridoma cells (15). Such a
mechanism to elicit Ag-specific CTL responses has also been described
in other in vitro infection models, e.g., influenza virus
(16). In this study we studied the consequences of
Salmonella SPI-1-induced macrophage death on the in vivo
induction of Salmonella-specific acquired immune responses.
By comparing the Ag-specific immune responses elicited with
aroAD and aroADinvA S.
typhimurium, we demonstrated that the lack of
Salmonella SPI-1-induced macrophage death did not affect the
induction of either serological or cellular immune responses.
Furthermore, we showed that elimination of splenic and hepatic
macrophages before and during systemic immunization with either
aroAD or aroADinvA S.
typhimurium did not alter the Ag-specific
CD8+ T lymphocyte response, suggesting that the
presence of spleen and liver macrophages in vivo is not required at all
for the induction of Salmonella-specific
CD8+ T lymphocyte responses. Together, these
results suggest that in vivo, uptake of dead (apoptotic)
Salmonella-infected macrophages is not the main mechanism by
which Salmonella Ags are delivered to a MHC class I
processing pathway. However, the possibility that dead host cells other
than macrophages, e.g., hepatocytes, are involved in delivery of
Salmonella-derived Ags to an MHC class I processing pathway
cannot be excluded.
Alternatively, since in vitro studies have demonstrated that S.
typhimurium can kill macrophages by two mechanisms, i.e., rapid
cell death mediated by SPI-1 (SipB) (11), or delayed cell
death mediated by SPI-2 (12, 13), and since this study was
performed using a SPI-1 (invA) mutant S.
typhimurium, a role for SPI-2-mediated delayed cell death in the
induction of Salmonella-specific CD8+
T lymphocyte responses cannot be excluded. However, results from in
vitro studies in our laboratory have demonstrated that non-phagocytic
cells such as epithelial MDCK cells, but also phagocytic dendritic
cells, are resistant to S. typhimurium-mediated cell death
(both aroAD and aroADinvA)
for at least 48 h after infection (results not shown). Therefore,
a role for SPI-2-mediated host cell death in the induction of cellular
immune responses seems unlikely.
Our finding that spleen and liver macrophages, which are regarded as the main host cell for S. typhimurium and have the capacity to present Ags to T lymphocytes, are not involved in the induction of Ag-specific immunity, raises the question of which cells function as the principal APC during immunization with attenuated Salmonella strains. It has been widely accepted that dendritic cells are superior to any other APC in presenting protein Ags and activating naive T lymphocytes. The interaction of dendritic cells with Salmonella has been investigated recently, and several in vitro studies have shown that murine dendritic cells can take up bacteria, resulting in maturation and cytokine expression and presentation of Salmonella-derived Ags to T lymphocytes (29, 30, 31). More recent in vivo studies have demonstrated that attenuated S. typhimurium strains expressing green fluorescent protein can be visualized within dendritic cells in Peyer’s patches at early time points after oral inoculation (32), and that green fluorescent protein-expressing, CD11c+ MHC-II+ cells can be detected in the spleen following systemic administration (33). Furthermore, uptake of Salmonella directly from the intestinal lumen by dendritic cells has recently been demonstrated (34). Together, these results suggest that after immunization with attenuated S. typhimurium, dendritic cells are the main APC involved in the induction of acquired immune responses. Recently, Bumann (35) demonstrated the local activation of Ag-specific CD4+ T lymphocytes in Peyer’s patches following inoculation of an attenuated S. typhimurium strain. Possibly, dendritic cells are taking up invaded Salmonella bacteria in the subepithelial dome and process and present Salmonella-derived Ags to T lymphocytes in Peyer’s patches and/or draining mesenteric lymph nodes. On the other hand, we suggest that S. typhimurium exploits the macrophage as a site for rapid replication and uses the macrophage to disseminate systemically to the liver and spleen. During secondary infections, macrophages function as important effectors of acquired immune responses.
Recent studies in caspase-1-/- mice showed that in the absence of caspase-1, S. typhimurium is unable to effectively colonize murine Peyer’s patches, resulting in a 1000-fold increased 50% lethal dose (14). These results suggested that the induction of macrophage death and the concomitant release of proinflammatory cytokines by S. typhimurium are a virulence mechanism exploited by this bacterium. In a previous study we showed that naive mice that were depleted of macrophages in the spleen and liver before infection were able to control an infection with wild-type S. typhimurium much better than normal mice (22). In addition, mice that were continuously depleted of macrophages did not succumb to the infection for >10 days, and the systemic bacterial load remained low. Together, these results support our hypothesis that macrophages are involved in the pathogenesis of virulent S. typhimurium infections, presumably through the bacterial-induced release of proinflammatory cytokines. In the absence of macrophages the bacterium cannot find the right milieu to replicate and disseminate throughout the host.
In conclusion, our data show that the in vivo induction of protective, acquired immune responses is not affected by Salmonella SPI-1-induced macrophage death during vaccination with an attenuated vaccine strain, and further, that the in vivo elimination of spleen and liver macrophages during immunization does not influence the induction of immunity. These results suggest that macrophage death as a result of invasion by S. typhimurium is not a host defense response, but more likely is a virulence mechanism exploited by the bacterium to establish an infection. Indeed, these results are in line with our previous findings that showed that macrophages are involved in the pathogenesis of wild-type S. typhimurium infections and are important for clearance of a recurrent infection in immunized animals, but are not involved in the induction of protective immunity (22).
|
Acknowledgments |
|---|
4370. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Odilia L. C. Wijburg, Department of Microbiology and Immunology, University of Melbourne, Royal Parade, Parkville, Victoria 3052, Australia. E-mail address: odilia{at}unimelb.edu.au
3 Abbreviations used in this paper: SPI-1, Salmonella pathogenicity island 1; Cl2MDP, dichloromethylene diphosphonate; MDCK, Madine-Darby canine kidney; MOI, multiplicity of infection; SipB, Salmonella invasion protein B; TT, tetanus toxin.
Received for publication January 17, 2002. Accepted for publication July 12, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
antibodies on histopathology of primary Salmonella infections. Infect. Immun. 63:3674.[Abstract]
and IFN in recall of immunity to oral challenge with virulent salmonellae in mice vaccinated with live attenuated aro- Salmonella vaccines. Microb. Pathog. 13:477.[Medline]
aroA aroD mutants expressing a foreign recombinant protein induce specific major histocompatibility complex class I-restricted cytotoxic T lymphocytes in mice. Infect. Immun. 61:5374. cells and IFN- in bacterial clearance independent of intracellular location. J. Immunol. 156:3321.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  J. D. Price, K. R. Simpfendorfer, R. R. Mantena, J. Holden, W. R. Heath, N. van Rooijen, R. A. Strugnell, and O. L. C. Wijburg Gamma Interferon-Independent Effects of Interleukin-12 on Immunity to Salmonella enterica Serovar Typhimurium Infect. Immun., December 1, 2007; 75(12): 5753 - 5762. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Jones-Carson, B. D. McCollister, E. T. Clambey, and A. Vazquez-Torres Systemic CD8 T-Cell Memory Response to a Salmonella Pathogenicity Island 2 Effector Is Restricted to Salmonella enterica Encountered in the Gastrointestinal Mucosa Infect. Immun., June 1, 2007; 75(6): 2708 - 2716. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Fu, O. L. C. Wijburg, P. U. Cameron, J. D. Price, and R. A Strugnell Salmonella enterica Serovar Typhimurium Infection of Dendritic Cells Leads to Functionally Increased Expression of the Macrophage-Derived Chemokine Infect. Immun., March 1, 2005; 73(3): 1714 - 1722. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Tobar, P. A. Gonzalez, and A. M. Kalergis Salmonella Escape from Antigen Presentation Can Be Overcome by Targeting Bacteria to Fc{gamma} Receptors on Dendritic Cells J. Immunol., September 15, 2004; 173(6): 4058 - 4065. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Pearce, D. J. Shedlock, and H. Shen Functional Characterization of MHC Class II-Restricted CD8+CD4- and CD8-CD4- T Cell Responses to Infection in CD4-/- Mice J. Immunol., August 15, 2004; 173(4): 2494 - 2499. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Johansson and M. J. Wick Liver Dendritic Cells Present Bacterial Antigens and Produce Cytokines upon Salmonella Encounter J. Immunol., February 15, 2004; 172(4): 2496 - 2503. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. R. Ganta, C. Cheng, M. J. Wilkerson, and S. K. Chapes Delayed Clearance of Ehrlichia chaffeensis Infection in CD4+ T-Cell Knockout Mice{dagger} Infect. Immun., January 1, 2004; 72(1): 159 - 167. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
