| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,§
,§
*
Department of Medicine, National Jewish Medical and Research Center, Howard Hughes Medical Institute, Denver, CO 80206; and Departments of
Pharmacology,
Biochemistry, Biophysics, and Genetics, and
§
Immunology and Medicine, University of Colorado Health Sciences Center, Denver, CO 80206
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This laboratory has frequently used staphylococcal enterotoxins, superantigens, as tools to study Ag-induced cell death of mature T cells (1, 7, 8, 9). Superantigens, when presented by the appropriate MHC class II molecule, activate T cells bearing particular Vßs as part of their TCRs. Superantigen treatment of mice thus allows for the activation of detectable but limited portions of the available repertoire of normal, primary T cells. Acute exposure of responsive T cells to a superantigen results in activation, expansion, and then deletion of these activated cells by apoptosis. This sequence of events is one postulated mechanism by which self-reactive T cells that have escaped detection during negative selection in the thymus can be removed from peripheral tissues.
Resistance to infections depends upon effective immune responses in which T cells stimulated by the invading organism do not all die rapidly. We and others have therefore reasoned that some property of infectious agents must act as an adjuvant to protect activated T cells against death. Along these lines, we have previously shown that bacterial LPS protects T cells stimulated by a superantigen from death (1). The protective effect of LPS appeared to act via LPS’s ability to induce components of the innate immune system, in particular proinflammatory cytokines, rather than costimulatory molecules, such as B7-1 or B7-2 (1, 9).
With this in mind, we wondered whether infectious agents such as viruses would affect activated T cells in a similar way. We chose vaccinia virus (VV)3 to test this idea. VV is well known for its efficacy as an immunogen, both for smallpox (10) and for protein Ags expressed from recombinant forms of the virus (11, 12), and for its ability to induce cytokine production by its hosts (13, 14). VV infects epithelial cells expressing receptors for epidermal growth factor (15, 16, 17), and thus infects many tissues in mice, including those of the ovaries, spleen, liver, and lung. Infection is acute: virus is completely cleared from immunocompetent mice in 7–10 days (15). Infection is also inflammatory by virtue of the inevitable cytopathicity VV induces in infected cells (16). Hence, we used infection with VV to determine whether or not infection with this model pathogen can act as adjuvant during superantigen-induced T cell responses, and if so by what means. We found that T cells responding to superantigen during active VV infection showed prolonged survival thereafter in mice and in culture. This enhanced survival of activated T cells was the only one of several adjuvant effects tested that was observed to operate in this model system, and appeared to be independent of cytokines such as IL-2, IL-4, IL-7, and IL-15 that induce Bcl-2 expression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Female B10.BR mice (The Jackson Laboratory, Bar Harbor, ME) were maintained in a specific pathogen-free environment at the Biological Resources Center of the National Jewish Medical and Research Center (Denver, CO). The mice in any given experiment were age-matched, but ranged in age from 8 to 14 wk in various experiments.
CD4 T cells were activated in vivo by injecting mice via the tail vein with 0.1 µg staphylococcal enterotoxin A (SEA; Toxin Technology, Sarasota, FL) in a volume of 0.25 ml balanced saline solution (BSS).
VV inoculation was performed by tail vein injection of 107 plaque-forming units (pfu) of the nonrecombinant WR strain of the virus (cultured from stocks provided by J. Bennink, National Institutes of Health, Bethesda, MD). Virus stocks were propagated in cultured cells by exposing 5 x 107 pfu of the virus to 0.125 µg/ml trypsin for 30 min at 37°C before adding the virus to 107 293 cells (American Type Culture Collection (ATCC), Manassas, VA) in medium containing 8% FBS for 72 h. At the end of this period, the cells were rinsed in BSS, frozen and thawed twice, and then lysed further in BSS by sonication. Stocks of virus were stored at -70°C before use. Cell lysates of uninfected 293 cells were prepared for use as mock inocula. Viral titers were measured by plaque assay on BS-C-1 cells (ATCC CCL 26) as described previously (17).
Lymphocyte preparation and analysis
Whole lymph nodes (LN) were removed for analysis from cervically dislocated mice at the specified times. The inguinal, brachial, axillary, and submandibular LN from similarly treated mice were pooled, dispersed by grinding through nylon mesh into culture medium, and counted with a Coulter cell counter (Coulter, Hileah, FL). The cells were resuspended at 1.5 x 106 cells per ml culture medium and plated in a volume of 0.2 ml in triplicate in wells of 96-well culture plates for 20 h, or were resuspended in flow cytometric staining buffer (BSS with 2% FBS and 0.01% sodium azide).
Culture with Fas cross-linking Abs was performed by adsorbing Jo2 anti-Fas MAb (PharMingen, San Diego, CA) or polyclonal hamster Ig (Sigma, St. Louis, MO) at 10 µg/well of a 96-well, polystyrene plate for 3 h at 37°C. Before cell culture, the wells were rinsed three times with culture medium. The effectiveness of Fas cross-linking was confirmed in each experiment by culturing L1210.Fas cells in wells coated with Ab (18). Culture of control cells with anti-Fas produced 40% and 60% apoptotic cells after 20 and 40 h, respectively, while culture with control Ig produced 3% apoptotic cells after any period of culture.
Lymphocyte survival was assessed by measuring the proportion of "live-gated" cells that bore a particular Vß before and after culture. The validity of the live-gate measurement was confirmed by comparison to other assays of apoptosis, including propidium iodide detection of DNA fragmentation (1), propidium uptake (cells stained for cytometry were incubated in 0.5 mg/ml propidium iodide in BSS for 15 min on ice before analysis), and annexin V staining (Sigma; cells stained for cytometry were incubated for 15 min in BSS containing 2.5 mM CaCl2 and the manufacturer’s recommended dilution of annexin V-FITC). Each measure of apoptosis gave similar values (19, and data not shown).
Recombinant mouse cytokines IL-2, IL-4, and IL-7 and human IL-15 were
purchased from R&D Systems (Minneapolis, MN). The cytokines were
suspended in culture medium and frozen at -70°C before use at the
concentrations indicated in Fig. 9
.
|
Flow cytometric analysis of cell-surface markers was performed with 106 uncultured cells or 3 x 105 cultured cells per stain, using mAbs specific for CD4, TCR Vß3, TCR Vß6, or CD69 (clones GK1.5, KJ25, RR4–7, and H1.2F3, respectively; PharMingen). The fluorochrome conjugates used for 3-color FACS analysis were CyChrome, phycoerythrin, and FITC. Intracellular Bcl-2 was detected by staining in buffer containing 0.03% saponin (Sigma), first with mAb clone 3F11, or human Bcl-2-specific clone 6C8 as control, and then with a mixture of anti-hamster IgG clones G70-204 and G94-56 coupled to FITC (PharMingen) essentially as described (20). Bcl-x expression was measured in cells after cell-surface stains were performed after fixation in 1% paraformaldehyde. Polyclonal rabbit anti-Bcl-x (Transduction Laboratories, Lexington, KY) was incubated at ambient temperature with fixed cells in staining buffer containing 0.3% saponin. The cells were washed three times, stained with goat anti-rabbit Ig that was fluoresceinated (Fisher Scientific, Pittsburgh, PA), washed again, and then analyzed flow cytometrically. Flow cytometry was performed on FACScan or FACScalibur instruments, and analysis was with PC-Lysis or CellQuest software (Becton Dickinson, San Jose, CA).
Cells used for mixing experiments were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes, Eugene, OR) immediately before culture. Cells were labeled by incubating 107 cells in 0.2 ml BSS containing 0.5 µM CFSE for 10 min at 37°C. After incubation, the cells were washed three times in culture medium supplemented with FBS. Cells used for transfer experiments were labeled by incubating 2 x 107 cells per ml BSS in 2 µm CFSE for 10 min at 37°C. The cells were washed once in BSS and then injected into recipient mice i.v.
RNA analyis
Total cellular RNA was isolated from whole LN using the Rapid
Total RNA Isolation Kit (large scale) as directed by the manufacturer
(5 Prime 
3 Prime, Boulder, CO). Yields were typically 0.5–1.5 µg
RNA per 106 LN cells.
RNA transcripts of cytokine genes were detected and quantified using
the RiboQuant Multiprobe RNase Protection Assay system, as directed by
the manufacturer (PharMingen). Briefly, radiolabeled RNA probes were
synthesized in vitro using the RiboQuant In Vitro Transcription kit
with incorporation of uridine 5'-triphosphate [
-32P],
6000Ci/mmol (New England Nuclear Life Science Products, Boston, MA).
RNA probes were prepared from PharMingen template sets mCK-1, mCK-2b,
mCK-3, and mCK-4. For each probe set, 106 Cherenkov units
were added to 75 µg total RNA obtained from LN of mice after selected
treatments and hybridized in solution by heating to 94°C for 2 min
and then incubating at 56°C for 12–15 h. Following hybridization,
any radiolabeled probe that was not annealed to its complementary
cellular transcript was digested with RNases T1 and A. The remaining
RNA was resolved by electrophoresis through acrylamide containing urea,
transferred to Whatmann filter paper, and dried.
Visualization and quantitation of resolved RNA was performed with a Storm 800 PhosphorImager (Molecular Dynamics, Sunnyvale, CA) after exposing phosphor screens to dried acrylamide gels for 24–48 h. Counts corresponding to each cytokine transcript were normalized from one lane to another by comparison of the sum of the signals from L32 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcripts in each lane. Each lane of the analysis consisted of one multiprobe set annealed to RNA obtained from one treatment group of mice.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We tested T cells in VV-infected mice for their susceptibility to superantigen-induced deletion to assess the adjuvant effects of a cytopathic virus infection on activated T cells. B10.BR mice (H-2k) were either inoculated with VV or not and treated 3 days later with the superantigen SEA, which, when bound to MHC class II proteins, activates CD4 and CD8 T cells bearing the Vß1, -3, or -11 regions of TCR. Flow cytometric detection of Vß1+ T cells is currently not possible, and B10.BR mice have very few Vß11+ T cells (7), so Vß3+ T cells were examined in all experiments.
Fig. 1 shows that, 2 days after SEA
injection, mice contained greatly expanded numbers of
Vß3+ T cells whether they were also infected with VV or
not. The enhanced numbers of Vß3+ cells were found
for a much longer time in VV-infected mice, indicating that virus
infection sustained the survival of the activated T cells. This pattern
was seen with both CD4 and CD8 T cells in spleen (data not shown) and
LN (Fig. 1) when tested either for the percent of Vß3+ T
cells, or for the absolute numbers of T cells bearing this TCR. T cells
bearing Vß6, which does not bind SEA, were unaffected (data not
shown). Vß3+ T cells in mice infected with VV alone were
also unaffected, demonstrating that VV itself does not stimulate
specifically a large percent of Vß3+ T cells (Fig. 1).
These observations showed that T cells activated by a model Ag, SEA,
persisted longer in mice when activation occurred during
"bystander" virus infection.
|
CFSE labeling experiments were used to track the proliferative
responses of SEA-stimulated cells in VV-infected mice. This was done to
determine whether or not CD4+Vß3+ T cells
proliferated preferentially upon SEA challenge because they had been
cross-primed by VV Ags. LN cells from B10.BR mice were loaded with the
fluorescent dye CFSE and transferred to syngeneic mice. Groups of
recipient mice were then given 107 pfu VV on day -3
(VV), 0.1 µg SEA on day 0 (SEA), or 107 pfu VV on day -3
and 0.1 µg SEA on day 0 (VV + SEA), as described in Fig. 1
. Two days
after SEA treatment, CD4+Vß3+ cells were
gated, and CFSE fluorescence was measured. Fig. 2A shows that the pattern of
CFSE loss, indicating successive cellular divisions, was similar in
cells from uninfected as in cells from VV-infected mice. Measurement of
the percent of CD4+ cells bearing Vß3+ on
subsequent days showed that cells with any amount of CFSE label, those
that had been transferred, showed a similar pattern of prolonged
survival in vivo as those that had been resident in the recipient mice.
Therefore, transferred cells had not lost all CFSE dye due to extensive
proliferation, and had proliferated no more in VV-infected mice than
they had in uninfected mice.
|
Prolonged survival of responding T cells has previously been
reported to occur as a result of increased Bcl-x expression following
engagement of the CD28 T cell coreceptor by B7 proteins on the surfaces
of APC (21, 22, 23, 24). We measured Bcl-x in cells activated 2 days previously
by SEA in vivo and found that expression of this antiapoptotic protein
was increased in activated cells, relative to resting cells, to the
same extent whether or not activation had occurred in mice that had
been infected with VV (Table I). Hence,
SEA-activated cells were destined to die following their stimulation in
vivo in spite of having expressed high levels of Bcl-x (Fig. 1), and
the increased persistence of activated T cells in infected mice was not
correlated with further increases in Bcl-x expression.
|
Activated T cells from infected mice survive longer in culture than activated T cells from uninfected mice
Analysis of the survival of activated T cells in mice is complicated by cellular proliferation and trafficking and by the rapid clearance of dead cells that occurs in vivo. Previous work from this laboratory and elsewhere has shown that T cells that have been activated and expanded in vivo die more quickly than resting T cells when cultured for short periods of time (1, 26). Measurement of survival in vitro therefore permits an estimate of a cell’s likelihood to die or to live in vivo, in a manner that is not influenced by that cell’s ability to traffic. Moreover, cells whose survival is measured in culture can be analyzed to determine the extent of their proliferation, if any, from the beginning of the culture period to the end.
To determine whether or not T cells that were activated during viral infection survived longer upon culture than other activated cells, B10.BR mice were left uninfected or were inoculated with VV, given SEA, and sacrificed 2 days later. LN cells from treated mice were analyzed by cytofluorometry either immediately after harvest or after culture for 20 h. Survival of SEA-activated cells was assessed by measuring the percent of live CD4 or CD8 T cells that bore Vß3. CD8 T cells behaved like CD4 T cells whenever tested, however, for simplicity in the rest of this paper, only the results for CD4 cells will be shown.
Live cells were defined by their forward vs side-light scattering
properties, which we confirmed gave values that were equivalent to
those determined with annexin V and propidium iodide staining (as
described in Materials and Methods). A proportion of both
the resting and activated T cell populations died during the 20-h
culture (data not shown), but a greater proportion of activated than
resting cells died, as demonstrated by the fact that after activation
with SEA the percentage of live, CD4 T cells bearing Vß3 fell during
culture (Fig. 3A). This
increased death rate of the activated T cells was prevented if the T
cells came from VV-infected mice. A summary of several experiments in
which SEA was given to uninfected or VV-infected mice showed no
statistical difference between the starting and ending proportion of
live, CD4 T cells bearing Vß3 (p = 0.69,
paired t test) if the cells had been activated in vivo
during bystander VV infection.
|
Infection with VV 3 days before giving SEA was sufficient to
confer longer life on the SEA-activated cells, but it was not clear
when during the course of infection the effect started and when it
ended. A time course of the virally induced protective effect was
therefore performed to determine when the effect peaked in vivo. Groups
of B10.BR mice were inoculated with VV on successive days and then
given SEA, such that Vß3+ T cells were activated on days
1–7 and on day 14 of infection. LN cells were removed from all mice 2
days after activation and tested for their survival in culture. Fig. 4 shows that injection of SEA on day 3,
4, or 5 of infection produced activated T cells with enhanced survival
in culture, whereas injection before or after this day 3–5 period had
little effect.
|
One explanation for the finding that VV infection increased the
life expectancy of SEA-activated T cells was that VV inhibited T cell
activation by SEA. The fact that VV did not affect the expansion of
Vß3+ T cells induced by SEA (Fig. 1), or their increased
Bcl-x levels (Table I), suggested that this was not true. However, as
an additional test, we measured induction of CD69, an early marker of T
cell activation, on Vß3+ T cells in VV-infected or normal
mice. In normal mice only, 11% of Vß3+CD4+ T
cells were CD69+ (data not shown). This percentage
increased slightly, to 16%, in VV-infected animals on day 5 of
infection, probably due to activation of the few
CD4+Vß3+ T cells that were VV-specific. SEA
injection caused 
50% of CD4+Vß3+ T cells
to be CD69+ 1 day later, whether or not the mice were also
infected with VV. These results showed that VV infection did not reduce
the ability of SEA to activate Vß3+ T cells and confirmed
that the effects of VV on SEA-activated cells were not due to the fact
that the rescued Vß3+ T cells were specific for VV as
well as SEA.
Also, VV infection did not affect the ability of SEA to reactivate
cells after 20 h of culture. In cells that had been cultured
overnight following day 2 SEA treatment, all but 2% of
CD4+Vß3+ T cells were CD69-.
Culture of these cells with SEA for an additional 20 h, however,
induced CD69 expression on 75% of the live Vß3+ T
cells, whether they came from VV infected mice or not.
Activated T cells that are protected from in vitro death by VV infection remain susceptible to killing via Fas
Previous results from this laboratory suggested that bacterial LPS
functioned as an adjuvant, at least in part, by rendering the activated
cells insensitive to cytopathic signaling via the death receptor Fas
(1). To test whether or not this was true for VV as well, T cells that
had been activated in mice with or without concurrent VV infection were
tested for their susceptibility to death induced by cross-linking Fas.
Whole LN populations were removed from mice after having received the
treatments indicated in Fig. 5. The
lymphocytes were plated in tissue culture wells that had been coated
previously with anti-Fas Ab, or with hamster Ig as control. At the
outset of culture, and after 20 or 40 h of culture, the
proportions of live CD4 cells bearing Vß3 TCR were determined.
|
).
Vß3+ cells from mice given only SEA died rapidly in vitro
with or without exposure to Fas Ab. Activated Vß3+ cells
from SEA treated, VV-infected mice, however, died quickly when the
culture wells contained Fas Ab. This behavior differed from that of the
Vß3+ cells from mice given SEA and LPS, which were
insensitive to killing via Fas. Hence, VV infection did not act as
adjuvant by making activated T cells resistant to Fas, but rather by
some other mechanism that was manifested as prolonged survival in
culture and in animals. The response to infection that protects activated T cells from rapid death is not ongoing during the culture period
Mixing experiments were performed to determine whether or not the adjuvant properties of viral infection operated in trans during culture. To do this, SEA-activated cells from VV-infected mice were labeled with the intracellular fluorescent dye CFSE (27), mixed with unlabeled, SEA-activated cells from uninfected mice, cultured, and assessed for survival.
Coculture had no effect on the survival of SEA-activated cells, whether
they came from VV infected mice or not (Fig. 6). That is, Vß3+ T cells
from VV-infected, SEA-treated mice had a slow rate of death, and
Vß3+ T cells from mice given SEA alone died rapidly,
regardless of whether or not they were cultured together. This was true
even when as few as 10% of the cells in the culture came from mice
given SEA alone (Fig. 6). Quantitation of the mean CFSE fluorescence of
labeled, activated T cells from VV-infected mice showed that they had
not undergone cell division while in culture. Because the amount of
CFSE per cell after 20 h of culture was the same in the
Vß3+ T cells as in the Vß3- T cells that
had not been affected by SEA (data not shown).
|
Measurement of cytokine transcripts in LN cells during viral infection
Because some cytokines rescue activated T cells from death in vivo
and in vitro (8, 28, 29, 30), cytokine gene expression in LN cells from
SEA-treated mice, with or without VV infection, was compared. The
previous experiments in this study were performed by giving SEA on day
3 of infection and removing LN cells 2 days later, on day 5 of
infection. The time course of infection shown in Fig. 4 indicated that
LN cells removed for analysis of survival at this time, day 5, could
also be analyzed meaningfully for expression of protective factors.
Therefore, RNA was isolated from whole LN 5 days after VV infection and
subjected to an RNase protection assay, which is quantitative and
allows for measurement of mRNA levels for many cytokines at once.
Correlation of gene transcription with virally induced protection
required comparison of transcript levels in mice inoculated with VV and
then treated with SEA vs levels in mice treated with SEA alone. Fig. 7 shows such a comparison for the 22
cytokine mRNAs out of 34 tested that produced a signal in the RNase
protection assay. mRNA levels for the housekeeping genes L32 and GAPDH
were used as controls. Of the 22 cytokine mRNAs compared, only 4 showed
an increase of >3-fold in VV-infected mice, those for IL-4, IL-10,
IFN-
, and lymphocyte inhibitory factor (LIF). Several others,
for IL-1ß, IL-1RA, IL-7, IL-12p35, IL-15, IFN--inducing factor,
granulocyte-macrophage CSF, and macrophage CSF showed relative
increases of <3-fold, and the rest showed no relative increase. The
type I IFNs, and ß, which constitute a subfamily of numerous gene
transcripts, were not tested in these assays. The expression of these
IFNs was presumably induced in VV-infected mice, however, because type
I IFN receptor signaling is required for infected mice to survive
vaccinia infection (31).
|
, and LIF further because transcription of these
cytokines correlated best with virally induced protection of activated
T cells. IL-4 rescues activated cells from death in culture
IL-4, IL-10, IFN-, and LIF were tested for rescuing activity by
adding them to cultures of cells that had been activated in vivo by
SEA. Of these cytokines, only IL-4 was able to prevent death of
activated CD4 T cells (data not shown, and Fig. 9). Addition of IL-10
or LIF to culture failed to prevent death, while IFN- showed minimal
activity that was not dose-responsive in whole LN cultures (data not
shown) and no activity in purified T cell cultures (1). IFN-,
moreover, was found not to be required for protection in experiments in
which IFN- was neutralized in vivo with mAb treatment during VV
infection, or in which IFN- -/- mice were tested for virally
induced protection in short-term assays (T.M., unpublished
observations). Of the 22 cytokines whose transcript levels were
measured, therefore, IL-4 was the only one to increase strongly during
VV infection and to protect activated T cells from death when tested in
culture.
VV protects activated T cells from rapid death in culture without increasing Bcl-2 expression
Increased Bcl-2 expression is one attribute of cytokine-induced
survival of lymphoid cells (8 , 32 , and 33 , and Fig. 9). If IL-4
secretion were to explain how infection prolongs the survival of
activated T cells, the amounts of Bcl-2 per cell might be expected to
increase relative to those in cells activated in the absence of
infection. This hypothesis was tested by measuring Bcl-2 expression in
SEA-activated cells after they were removed from uninfected or
VV-infected mice and after they were cultured with IL-4 or with IL-15
as positive control. Flow cytometric stains for Bcl-2 under such
conditions showed that CD4+Vß3+ T cells
activated 2 days earlier by SEA had decreased their expression of Bcl-2
(Fig. 8). VV infection did not restore
Bcl-2 levels, but culture with 100 ng/ml IL-4 or IL-15 did. Hence, IL-4
and IL-15 rescued activated T cells from death in vitro by a mechanism
that seemed distinct from that induced by infection in vivo.
|
The RNase protection assay described in Fig. 7 might have failed
to detect other cytokines as candidate survival factors due to
insufficient sensitivity of the assay. So, selected cytokines were
analyzed for their effects on Bcl-2 expression. IL-2, IL-4, IL-7, and
IL-15 have all been reported to rescue activated T cells from death in
vitro, and some of these are active in vivo (8, 32, 33). All induced
Bcl-2 expression in activated cells when tested in culture (Fig. 9). Each of the responses was
dose-dependent and correlative inasmuch as culture doses that caused
rescue also induced Bcl-2, while doses insufficient to cause rescue did
not induce Bcl-2. Since VV did not induce Bcl-2 in activated
CD4+ T cells, exposure of these cells to IL-2, IL-4, IL7,
and IL-15 after the cells had been activated seemed unlikely to explain
the adjuvant properties of VV infection.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The means by which infectious agents increase the life spans of activated T cells is not clear. Unlike LPS, VV infection did not make activated cells insensitive to Fas killing, since activated T cells from VV-infected mice still succumbed to cytopathic signaling via Fas. This result raised the question of the importance of Fas to the death of superantigen-activated T cells. The data in the literature is contradictory, with some laboratories indicating that Fas is obligatory for superantigen-induced death (39, 40), others showing that Fas is not needed (41), and yet others demonstrating that Fas plays a decisive role at low but not high doses of superantigen (42, 43, 44), or with membrane-bound but not soluble Ag (45). Our own experiments with Fas- and TNF receptor-deficient mice indicate that superantigen-activated cells from mutant mice die at least as fast in culture and in mice as do cells from wild-type or heterozygous mice (D. Hildeman, unpublished observation). A failure of viral infection to confer resistance to Fas-killing is consistent with the view that Fas signaling is not as important as other mechanisms in controlling T cell survival, at least during and after priming in LN.
Bcl-x expression was increased to the same extent in activated cells whether activation occurred in uninfected or in VV-infected mice. Hence, the amount of costimulation available to SEA-responsive cells seemed not to be limiting in uninfected mice, and was not increased further in mice undergoing VV infection. This conclusion was supported by our observations in multiple experiments that VV infection did not increase the expansion of superantigen-specific T cells over that observed after treatment with superantigen alone.
In attempts to discover which cytokines might account for VV’s ability to enhance T cell survival, we found that surprisingly few cytokine transcripts were induced in the LN of VV-infected and SEA-treated mice, relative to mice that had received SEA alone. Since the IL-2 family of cytokines is known to affect T cell survival (8, 28, 29, 30, 33) and since a member of the family, IL-15, has recently been reported to affect CD8+ T cells in vivo (46, 47), we paid particular attention to mRNA transcripts for this family. Vaccinia infection caused small increases in the levels of mRNAs for IL-2, IL-7, and IL-15 and a large increase in the level of mRNA for IL-4. These cytokines were unlikely to account for the longer life expectancy of activated T cells in VV-infected mice, however, because these cytokines raised the level of Bcl-2 in T cells (32, 33), and the rescued T cells in VV-infected mice did not have increased amounts of this protein. Moreover, culture of short-lived, activated T cells from mice given SEA alone with T cells from mice given SEA and VV did not rescue the former cells from death. Thus, T cell survival was not affected by a factor produced in vitro by the cells from VV-infected mice. IL-2 family members must be continually available in culture for them to protect T cells from death (T. Kent Teague, unpublished observations), indicating that members of this family must not have been responsible for the increased life expectancy in vitro of activated T cells from VV-infected mice. These conclusions rely, however, on the assumption that T cells respond to cytokines in vivo in the same way they do in vitro, an assumption that underlies much of immunologic research, but that may inappropriately discount the complexity of immune responses in animals.
The experiments reported here identify several properties of the factor(s) induced by virus infection that protect activated T cells from death. The factor must affect activated but not resting T cells, it must protect cells without inducing proliferation or Bcl-2 or Bcl-x expression, and must not be dependent on continuous secretion for efficacy.
We report elsewhere that type I IFNs satisfy many of these criteria
inasmuch as their addition to cultures of activated, but not resting, T
cells promotes survival without inducing Bcl-2 or Bcl-x expression, and
without inducing proliferation. However, CFSE-transfer experiments with
cells from IFNß receptor-deficient mice showed that virally
induced survival was equivalent whether wild-type or receptor-deficient
cells were transferred to infected recipients (our unpublished
observation). Thus, type I IFNs can influence T cell survival in vitro,
but are apparently not required for prolonged survival in vivo.
Finally, virus infections might be acting to inhibit death pathways, rather than induce rescuing factors. Factors or pathways that are elicited globally in an animal and that induce activated T cell death unless innate immune responses are activated, for example, may be responsible. Examples of mechanisms that might behave in this manner are glucocorticoid-induced death (46), and monocyte-induced cytotoxicity of T lymphocytes (47, 48). Whichever factor or pathway is found to support activated T cell survival will likely explain an important part of the mechanism by which viral infections result in dramatically increased numbers of Ag-specific T cells after successful immune responses, and by which self-antigens can become targets in autoimmunity.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Philippa Marrack, Howard Hughes Medical Institute, Department of Medicine, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206. E-mail address:
3 Abbreviations used in this paper: VV, vaccinia virus; SEA, staphylococcal enterotoxin A; BSS, balanced saline solution; pfu, plaque-forming units; CFSE, carboxyfluorescein diacetate succinimidyl ester; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; LN, lymph node; LIF, lymphocyte inhibitory factor.
Received for publication August 26, 1998. Accepted for publication January 28, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-chain signaling cytokines regulate activated T cell apoptosis in response to growth factor withdrawal: selective induction of anti-apoptotic (bcl-2, bcl-xL) but not pro-apoptotic (bax, bcl-xS) gene expression. Eur. J. Immunol 26:294.[Medline]
This article has been cited by other articles:
|
|
 |
|
  S. J. Ramos, J. B. Hernandez, M. Gatzka, and C. M. Walsh Enhanced T Cell Apoptosis within Drak2-Deficient Mice Promotes Resistance to Autoimmunity J. Immunol., December 1, 2008; 181(11): 7606 - 7616. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Reckling, S. Divanovic, C. L. Karp, S. Wojciechowski, Y. Belkaid, and D. Hildeman Proapoptotic Bcl-2 Family Member Bim Promotes Persistent Infection and Limits Protective Immunity Infect. Immun., March 1, 2008; 76(3): 1179 - 1185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Wojciechowski, P. Tripathi, T. Bourdeau, L. Acero, H. L. Grimes, J. D. Katz, F. D. Finkelman, and D. A. Hildeman Bim/Bcl-2 balance is critical for maintaining naive and memory T cell homeostasis J. Exp. Med., July 9, 2007; 204(7): 1665 - 1675. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Sengupta, P. Jayaraman, P. M. Chilton, C. R. Casella, and T. C. Mitchell Unrestrained Glycogen Synthase Kinase-3beta Activity Leads to Activated T Cell Death and Can Be Inhibited by Natural Adjuvant J. Immunol., May 15, 2007; 178(10): 6083 - 6091. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Tripathi, T. C. Mitchell, F. Finkelman, and D. A. Hildeman Cutting Edge: Limiting Amounts of IL-7 Do Not Control Contraction of CD4+ T Cell Responses J. Immunol., April 1, 2007; 178(7): 4027 - 4031. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Mittal, S. Papa, G. Franzoso, and R. Sen NF-{kappa}B-Dependent Regulation of the Timing of Activation-Induced Cell Death of T Lymphocytes J. Immunol., February 15, 2006; 176(4): 2183 - 2189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Barouch-Bentov, E. E. Lemmens, J. Hu, E. M. Janssen, N. M. Droin, J. Song, S. P. Schoenberger, and A. Altman Protein Kinase C-{theta} Is an Early Survival Factor Required for Differentiation of Effector CD8+ T Cells J. Immunol., October 15, 2005; 175(8): 5126 - 5134. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Takahashi, M. G. V. Hanson, H. R. Norell, A. M. Havelka, K. Kono, K.-J. Malmberg, and R. V. R. Kiessling Preferential Cell Death of CD8+ Effector Memory (CCR7-CD45RA-) T Cells by Hydrogen Peroxide-Induced Oxidative Stress J. Immunol., May 15, 2005; 174(10): 6080 - 6087. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. O. Valenzuela, C. D. Hammerbeck, and M. F. Mescher Cutting Edge: Bcl-3 Up-Regulation by Signal 3 Cytokine (IL-12) Prolongs Survival of Antigen-Activated CD8 T Cells J. Immunol., January 15, 2005; 174(2): 600 - 604. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ichii, A. Sakamoto, Y. Kuroda, and T. Tokuhisa Bcl6 Acts as an Amplifier for the Generation and Proliferative Capacity of Central Memory CD8+ T Cells J. Immunol., July 15, 2004; 173(2): 883 - 891. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhu, B. J. Swanson, M. Wang, D. A. Hildeman, B. C. Schaefer, X. Liu, H. Suzuki, K. Mihara, J. Kappler, and P. Marrack Constitutive association of the proapoptotic protein Bim with Bcl-2-related proteins on mitochondria in T cells PNAS, May 18, 2004; 101(20): 7681 - 7686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Hildeman, T. Mitchell, B. Aronow, S. Wojciechowski, J. Kappler, and P. Marrack Control of Bcl-2 expression by reactive oxygen species PNAS, December 9, 2003; 100(25): 15035 - 15040. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Rubinstein, A. N. Kadima, M. L. Salem, C. L. Nguyen, W. E. Gillanders, and D. J. Cole Systemic Administration of IL-15 Augments the Antigen-Specific Primary CD8+ T Cell Response Following Vaccination with Peptide-Pulsed Dendritic Cells J. Immunol., November 1, 2002; 169(9): 4928 - 4935. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Turner, R. Cross, W. Xie, and P. C. Doherty Concurrent Naive and Memory CD8+ T Cell Responses to an Influenza A Virus J. Immunol., September 1, 2001; 167(5): 2753 - 2758. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-C. Ku, J. Kappler, and P. Marrack The Growth of the Very Large CD8+ T Cell Clones in Older Mice Is Controlled by Cytokines J. Immunol., February 15, 2001; 166(4): 2186 - 2193. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Kedl, W. A. Rees, D. A. Hildeman, B. Schaefer, T. Mitchell, J. Kappler, and P. Marrack T Cells Compete for Access to Antigen-Bearing Antigen-Presenting Cells J. Exp. Med., October 16, 2000; 192(8): 1105 - 1114. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Teague, B. C. Schaefer, D. Hildeman, J. Bender, T. Mitchell, J. W. Kappler, and P. Marrack Activation-Induced Inhibition of Interleukin 6-Mediated T Cell Survival and Signal Transducer and Activator of Transcription 1 Signaling J. Exp. Med., March 20, 2000; 191(6): 915 - 926. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. S. Liwski and T. D. G. Lee Nematode Infection Enhances Survival of Activated T Cells by Modulating Accessory Cell Function J. Immunol., November 1, 1999; 163(9): 5005 - 5012. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Teague, D. Hildeman, R. M. Kedl, T. Mitchell, W. Rees, B. C. Schaefer, J. Bender, J. Kappler, and P. Marrack Activation changes the spectrum but not the diversity of genes expressed by T cells PNAS, October 26, 1999; 96(22): 12691 - 12696. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zhang, S. F. Schlossman, R. A. Edwards, C.-N. Ou, J. Gu, and M. X. Wu Impaired apoptosis, extended duration of immune responses, and a lupus-like autoimmune disease in IEX-1-transgenic mice PNAS, January 22, 2002; 99(2): 878 - 883. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Kedl, M. Jordan, T. Potter, J. Kappler, P. Marrack, and S. Dow CD40 stimulation accelerates deletion of tumor-specific CD8+ T cells in the absence of tumor-antigen vaccination PNAS, September 11, 2001; 98(19): 10811 - 10816. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
