| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
Departments of
* Immunology and
Oncology, University of Washington, Seattle, WA 98195; and
Program in Immunology, Fred Hutchinson Cancer Research Center, Seattle, WA 98109
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Establishment of the memory CD8+ T cell pool
following clearance of a viral infection is in part determined by the
presence of a concurrent CD4+ Th response during
the response to acute infection (2). Although in some
settings CD4+ Th cells are absolutely required
for induction of a CD8+ T cell response, based on
the necessity to activate and license APCs (8, 9, 10), many
viruses including LCMV can activate APC independent of
CD4+ Th cells and induce an antiviral
CD8+ T cell response (11, 12).
However, even with such viruses, CD4+ T cells
provide an additional helper function that promotes formation of a
memory CD8+ T cell response. IL-2 production is
an effector function of CD4+ Th cells not
alternatively provided by the virus. IL-2 delivers signals through the
heterotrimeric IL-2R, comprised of a unique 
-chain (CD25) required
for high affinity binding of IL-2, as well as the signaling 
- and
-chains shared with the IL-15R (13). Although both
naive CD8+ T cells and CD4+
Th cells can produce IL-2 following initial activation, only
CD4+ T cells retain this capacity after
differentiation into effector T cells (14, 15).
Defining the in vivo role of IL-2 in the regulation of CD8+ T cells responding to a virus has been complicated by the myriad effects of IL-2 in the immune system. In vitro studies have demonstrated that signals through the IL-2R can lead to enhanced CD8+ T cell proliferation and survival, or alternatively cell death following T cell activation (16, 17, 18), and data from some in vivo models have suggested that IL-2 may have a predominantly negative regulatory role on T cell responses by promoting activation-induced cell death (AICD). IL-2-/- mice display only a slight decrease in the magnitude of the peak CD8+ T cell response to viral infection (19, 20, 21), and suffer from a T cell-mediated lymphoproliferative autoimmune syndrome (22). Ab blockade of IL-2R signaling in vivo leads to a selective increase in the total population of CD8+ T cells with a memory phenotype, suggesting IL-2 may naturally function to induce cell death in memory CD8+ T cells (23). However, these global modulations of IL-2 biology may not reflect the direct role of IL-2R signaling in CD8+ T cells responding to a pathogen, but rather, the roles of IL-2R signals in T cell development and/or maintenance of regulatory CD25+ CD4+ T cells (24, 25, 26, 27).
To assess the direct role of IL-2R signals in the generation of
antiviral CD8+ T cell responses and the
establishment of memory CD8+ T cells, we designed
a strategy to deliver augmented IL-2R signals only to responding
CD8+ T cells during the period of the active
response to Ag. Therefore, transgenic (Tg) mice were generated with a
murine chimeric GM-CSF/IL-2R (GMIL2R) under the control of a CD8
enhancer/promoter, which results in transgene expression exclusively in
mature CD8+ T cells, thereby obviating any impact
on T cell development (28, 29). The GMIL2R is a two-chain
receptor consisting of the extracellular ligand-binding domains of the
GM-CSFR - and -chains fused to the intracellular signaling
domains of the IL-2R - and -chains, respectively. The GM-CSFR
ectodomains were selected for the chimeric receptor because the
receptor is not normally expressed on resting or activated murine
CD8+ T cells (30), and the GM-CSFR
shares nearly identical on/off rates and a dissociation constant for
ligand binding as the IL-2R (31, 32). Binding of GM-CSF
dimerizes the chimeric GMIL2R chains and delivers a signal to T cells
indistinguishable, either kinetically or biochemically, from the native
IL-2R (33, 34, 35). Moreover, GM-CSF, unlike IL-2, continues
to be produced following target recognition by
CD8+ T cells that have differentiated from naive
to effector T cells (14, 36). Thus, following Ag
stimulation, GMIL2R+ effector
CD8+ T cells retain via GM-CSF production an
autocrine mechanism to sustain proliferation without a requirement for
exogenous sources of IL-2 (29).
Enhanced growth of CD8+ T cells expressing the GMIL2R in response to immunization with irradiated tumor cells was previously demonstrated both in vitro and in vivo (29). Although these data indicated a positive regulatory role for IL-2R signals in the generation of CD8+ T cell responses to immunization, as has been suggested by vaccine studies that incorporate IL-2 during immunization (37), the availability of Ag and subsequent TCR signaling are short-lived in these settings due to the inability of the immunogen to replicate. Because promotion of AICD by IL-2 occurs primarily following repetitive stimulation of effector CD8+ T cells in the presence of high Ag loads (18, 38, 39), the negative regulation of CD8+ T cell responses by IL-2R-mediated AICD may be more relevant following recognition of replicating or abundant Ags (40, 41), such as might occur in humans following EBV or HIV infections. Therefore, we examined infection of GMIL2R Tg mice with LCMV, a well-described murine virus that generates high viral burdens, and for which strategies have been developed to precisely evaluate the magnitude of the CD8+ T cell response and the contributions of CD4+ Th function. Our data demonstrate that following LCMV infection, the physiologically provided IL-2R signals limit the magnitude of acute CD8+ T cell expansion and the resultant pool of memory CD8+ T cells formed during the primary antiviral response. Additionally, although a concurrent CD4+ Th response to the virus increases the size of the memory CD8+ T cell pool established, these contributions by CD4+ Th can be supplanted by provision of increased IL-2R signals to responding CD8+ T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All mice were handled according to University of Washington and American Association for the Accreditation of Laboratory Animal Care guidelines. B6 and Thy-1.1 congenic mice were purchased from The Jackson Laboratory (Bar Harbor, ME). GMIL2R Tg mice were generated on a pure B6 background, and expression of the GMIL2R in CD8+ T cells did not lead to alterations in T cell development, the size of lymphoid T cell populations, nor expression of activation/memory markers on CD8+ T cells in the absence of Ag stimulation (29). For assessing primary infections, mice received 2.5 x 105 PFU of the Armstrong strain of LCMV i.p. For assessing recall responses, 5 x 106 PFU of the more virulent LCMV clone 13 was administered i.v.
In vivo CD4+ T cell depletion
Culture supernatants from the anti-CD4-producing GK1.5 hybridoma were isolated, and Ab precipitated with 50% saturated ammonium sulfate solution (42). Precipitates were resuspended and dialyzed 50-fold against 1x PBS before use. Mice were injected i.p. with 0.75 ml of purified GK1.5 Ab, which led to deletion of >99% of CD4+ T cells within 2 days of Ab depletion (42). The CD4-deficient state was maintained by weekly Ab administration (42).
Intracellular cytokine staining and flow cytometry
Spleens were harvested, homogenized, and passed through nylon
filters to create single cell suspensions, and RBC lysed via
resuspension in ACK (150 mM NH4Cl, 10 mM
KHCO3) lysis buffer. A total of 106
splenocytes was then cultured in 96-well flat-bottom plates with 10
U/ml IL-2, and 0.1 µg/ml of either NP396–404 or gp33–41 peptide in
RPMI supplemented with 10% FBS (10% RPMI). One hour later, 10 µg/ml
of brefeldin A (Sigma-Aldrich, St. Louis, MO) was added, and 4 h
later cells were harvested, stained with anti-CD8 FITC mAb, fixed,
permeabilized with cytofix/cytoperm solution, and stained with
anti-IFN- PE mAb (BD PharMingen, La Jolla, CA). Cells were
analyzed on a FACSCalibur, and data analysis was performed using
CellQuest software (BD Biosciences, San Jose, CA).
Isolation of lymph node, lung, and liver T cells
Mice were perfused with HBSS supplemented with 10 U/ml heparin. Inguinal and mesenteric lymph nodes were isolated and homogenized to form single cell suspensions. Lung and liver tissue were resected as intact exsanguinated organs and homogenized, and lymphocytes were isolated similar to published protocols (43).
Adoptive transfer for examination of memory T cell turnover
Single Tg mice expressing TCRs specific for the gp33 peptide of LCMV (TCR) and double Tg TCR x GMIL2R mice were infected with 2.5 x 105 PFU of LCMV i.p. 30 days after infection, and splenocytes were harvested and labeled with CFSE for 1 h at 37°C in 10% RPMI. A total of 3 x 106 T cells was then transferred i.v. into LCMV-immune Thy-1.1 congenic mice also infected 30 days prior with LCMV. Ten days after transfer, spleens from congenic hosts were isolated, and the transferred T cells were examined via flow cytometry.
Plaque assay
Plaque assays were performed, as previously described (44). Briefly, Vero cell monolayers were infected with serial dilutions of splenic homogenate or serum from LCMV-infected mice at 37°C for 1 h. Infected monolayers were then overlayed with 5% media 199 in 0.5% agarose. Plaques were counterstained with neutral red 4 days later and enumerated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Following acute LCMV infection, mice develop splenomegaly due in
part to an increase in splenic CD8+ T cells
(45). Therefore, CD4-intact B6 and GMIL2R Tg mice were
infected with 2.5 x 105 PFU of the
Armstrong strain of LCMV i.p., and spleens were removed at the peak of
T cell expansion on day 8 (45). Infected B6 mice developed
enlarged spleens (Fig. 1
), which included
an increase in the number of splenic CD8+ T cells
from 6.5 x 106 in uninfected control mice
(n = 5) to 2.9 x 107 cells.
Uninfected B6 and GMIL2R Tg mice had equivalent numbers of splenic
CD8+ T cells (both 6.5 x
106 CD8+ T cells), but
following LCMV infection, GMIL2R Tg mice (n = 6)
consistently showed more marked splenomegaly, which contained 6.9
x 107 CD8+ T cells/spleen
compared with 2.9 x 107 in infected B6
mice. Similar expansion of the CD8+ population at
the peak on day 8 was observed in CD4-deficient Tg and B6 LCMV-infected
mice, with CD4-deficient GMIL2R Tg mice containing 7.8 x
107 splenic CD8+ T cells
compared with 2.8 x 107 in B6 mice. Mice
from all groups displayed nearly identical serum viral titers of LCMV
by plaque assay on day 3 (1.3–1.7 x 103
PFU/ml) and completely cleared LCMV infection by day 9
postinfection.
|
To determine whether the increased number of
CD8+ T cells in spleens of infected GMIL2R Tg
mice reflected a larger LCMV-specific CD8+ T cell
response, LCMV-specific T cells reactive to the two codominant
immunogenic epitopes of LCMV, NP396–404 (NP396) and gp33–41 (gp33),
were quantitated by intracellular staining for IFN- production
(45). On day 8 following infection of B6 mice
(n = 5), 9.2% of all splenic
CD8+ T cells produced IFN- in response to
NP396, representing 2.6 x 106
NP396-specific splenic CD8+ T cells (Fig. 2, A and B). B6
mice lacking CD4+ T cells (n = 6)
had similar frequencies (9.5%) and numbers (2.4 x
106) of NP396-specific T cells compared with the
CD4-intact B6 mice (Fig. 2B), consistent with reports that
CD4+ T cells do not significantly enhance the
peak effector response to LCMV (42). Analysis of splenic
gp33-specific CD8+ T cells revealed similar
results (Fig. 2, A and C). In CD4-intact B6 mice,
9.6% of CD8+ T cells were gp33 specific,
representing 2.8 x 106 gp33-specific T
cells, whereas in CD4-deficient mice, 7.8% of
CD8+ T cells were gp33-specific, representing
2.3 x 106 T cells/spleen.
|
, B and C).
These responses were >3-fold higher than the response observed in
CD4-intact B6 mice (p < 0.001). Similarly,
spleens from CD4-deficient GMIL2R Tg mice (n = 6)
contained 9.5 x 106 NP396-specific and
7.3 x 106 gp33-specific
CD8+ T cells/spleen at frequencies of 11.6 and
9.5% (Fig. 2, B and C), which were >3-fold
greater than responses in CD4-deficient B6 mice
(p < 0.001).
By day 14 after infection in both CD4-intact and deficient B6 mice, the
NP396-specific and gp33-specific CD8+ T cell
pools had contracted >80% (Fig. 2, B and C),
most likely due to apoptosis following viral clearance and removal of
the antigenic stimulus (1, 39, 46). The NP396- and
gp33-specific CD8+ T cell pools of CD4-intact and
deficient GMIL2R Tg mice similarly contracted >80% from the peak,
suggesting that enhanced signaling through the chimeric IL-2R during
the response to the virus did not lead to either improved survival or
proportionately greater cell death from cytokine withdrawal following
viral clearance.
GMIL2R Tg mice generate an increased number of LCMV-specific memory T cells
By day 25 postinfection, spleens of infected mice return nearly to
normal size, and a stable long-lived memory CD8+
T cell pool to LCMV is established (45). At this time,
1.8 x 105 NP396-specific and 2.2 x
105 gp33-specific CD8+ T
cells were present in spleens of CD4-intact B6 mice (n
= 10), representing 2.0 and 2.7% of CD8+ T cells
(Fig. 2, B and C). Although the absence of
CD4+ T cells did not impact the magnitude of the
peak antiviral CD8+ T cell response achieved in
response to acute infection, CD4-deficient mice exhibited a significant
decrease in the size of the memory pool established, consistent with
previous reports (2). In CD4-deficient mice
(n = 9), the NP396-specific CD8+
T cell memory pool of 0.83 x 105 cells and
frequency of 1.7% were >50% smaller (p <
0.025) than the pool in CD4-intact mice (Fig. 2B).
Similarly, there was a nearly 60% decline (p
< 0.025) in formation of the gp33-specific memory
CD8+ T cell pool of 0.93 x
105 cells at a frequency of 1.9% of
CD8+ T cells (Fig. 2C).
At day 25 postinfection, spleen sizes in GMIL2R Tg mice were only 1.3-
to 1.4-fold larger than B6 mice, as compared with the >2-fold
increases on day 8 postinfection. This in part reflected the
significantly increased memory CD8+ T cell pools
in infected GMIL2R Tg animals compared with infected B6 mice. On day 25
postinfection, spleens from CD4-intact GMIL2R Tg mice
(n = 11) infected with LCMV contained 4.7 x
105 NP396-specific and 4.9 x
105 gp33-specific CD8+ T
cells at frequencies of 3.5 and 3.9% of
CD8+ T cells (Fig. 2, B and
C), which were 2.6- and 2.2-fold greater than infected
CD4-intact B6 mice (p < 0.001). In
CD4-deficient GMIL2R Tg mice (n = 8), the
NP396-specific and gp33-specific CD8+ T cell
splenic pools of 3.6 x 105 and 3.0 x
105 at frequencies of 3.6 and 3.2% were 4.3- and
3.2-fold larger (p < 0.001) than the pools in
infected CD4-deficient B6 mice (Fig. 2, B and C).
As with B6 mice, there were fewer total memory NP396- and gp33-specific
CD8+ T cells in CD4-deficient GMIL2R Tg mice
compared with CD4-intact GMIL2R Tg mice, but the magnitude of these
differences was smaller. The NP396-specific and gp33-specific memory
CD8+ T cell pools in CD4-deficient GMIL2R Tg mice
were only 24 and 38% less than the pools in CD4-intact GMIL2R Tg mice,
as contrasted to the 50 and 60% decreases in these memory pools in
CD4-deficient B6 mice compared with intact B6 mice. Indeed, the sizes
of the memory populations established in CD4-deficient GMIL2R Tg mice
exceeded the sizes achieved in B6 mice with an intact
CD4+ Th response. The increase in the size of the
memory CD8+ T cell pools in GMIL2R Tg mice was
stable and maintained long-term, with no change from days 25 to 50 in
both B6 and GMIL2R Tg mice (Fig. 2, B and C).
The increased peak responses and memory pool size detected in GMIL2R Tg mice presumably reflected the autocrine proliferative response to GM-CSF produced during the period CD8+ T cells recognized LCMV-infected targets. CD8+ T cells isolated from mice either during the period of acute viral infection or following viral clearance only produced GM-CSF if specifically stimulated with Ag, and no GM-CSF from exogenous sources could be detected either spontaneously from spleen cells or in the serum of mice that had cleared infection (data not shown).
GMIL2R+ CD8+ memory T cells exhibit similar rates of turnover compared with wild-type memory CD8+ T cells
Despite constitutive expression of the GMIL2R on
CD8+ T cells, GM-CSF does not cause naive
GMIL2R+ T cells to proliferate in the absence of
Ag in vitro (29). GM-CSF is not measurable in the serum of
mice (47), but, because it is potentially produced by a
variety of cell types (48, 49), such local production of
GM-CSF might lead to activation and proliferation of resting memory
GMIL2R+ CD8+ T cells.
Resting memory CD8+ T cells do normally exhibit a
steady proliferative rate, with 
20–30% of memory
CD8+ T cells dividing during a 7-day period
(50), and the observed stability of the size of the
LCMV-specific memory CD8+ T cell pools in GMIL2R
Tg mice between days 25 and 50 suggested GMIL2R+
CD8+ T cells did not have a higher turnover rate.
However, we directly examined the turnover of
GMIL2R+ and GMIL2R- memory
CD8+ T cells in an adoptive transfer model
(50). To generate a large population of LCMV-specific
memory T cells, Tg mice expressing a TCR specific for gp33 (TCR) were
crossed to GMIL2R Tg mice (TCR x GMIL2R) and infected with 2.5 x
105 PFU of LCMV (51). Before
infection, CD8+ T cells from TCR and TCR x
GMIL2R mice had a uniformly naive phenotype, with fewer than 10% of
CD8+ T cells expressing CD44. However, 30 days
following infection, >85% of CD8+ T cells
displayed a CD44high memory phenotype (data not
shown). CD8+ T cells from TCR and TCR x GMIL2R
Tg mice were harvested and loaded with CFSE to monitor T cell
proliferation and turnover. To closely mimic the environment that the
transferred memory CD8+ T cell would normally be
exposed, the CFSE-labeled cells were transferred into nonirradiated
LCMV-immune Thy-1.1 congenic hosts that had also been infected 30 days
prior. Ten days after transfer, spleens were harvested and CFSE content
was examined in Thy-1.2+
CD8+ T cells (Fig. 3). At day 10 posttransfer, 31% of
TCR T cells had divided at least once, consistent with published
results (50). TCR x GMIL2R T cells exhibited a similar
turnover rate, with 30% of TCR x GMIL2R T cells having undergone at
least one cell division. Similar numbers of TCR and TCR x GMIL2R were
recovered from recipient mice (data not shown).
|
In mice, memory T cells have been divided into two subpopulations
based in part on distinct migratory patterns to either lymphoid or
nonlymphoid tissues (43, 52, 53). Lymphoid and nonlymphoid
memory T cells also display unique phenotypic and functional
characteristics, such as differences in cytokine expression or effector
function upon target recognition (43, 53). Because GMIL2R
Tg mice exhibited an increased CD8+ T cell
response in the spleen, we next examined whether this reflected altered
differentiation impacting on the CD8+ T cell
pools established in nonlymphoid organs. CD4-intact B6 and GMIL2R Tg
mice were infected with 2.5 x 105 PFU of
LCMV i.p., and 30 days later spleen, lymph nodes, liver, and lung were
removed and perfused to eliminate blood lymphocytes. Liver and lung
lymphocytes were isolated as previously described (43).
CD4-intact GMIL2R Tg mice contained 2.5-fold more NP396- and
gp33-specific CD8+ T cells
(p < 0.005) in spleen and lymph nodes compared
with CD4-intact B6 mice (Fig. 4A). The virus-specific
CD8+ T cell populations of CD4-intact GMIL2R Tg
mice residing in liver and lung were also expanded relative to
CD4-intact B6 mice, with 2.2- and 1.8-fold increases observed in liver
and lung, respectively (p < 0.05). In
CD4-deficient GMIL2R Tg mice, the virus-specific lymphoid
CD8+ T cell pool was 3.3-fold greater
(p < 0.001) than in CD4-deficient B6 mice and
1.7-fold greater (p < 0.02) in both lung and
liver (Fig. 4B). Thus, the enhanced memory
CD8+ T cell pools in GMIL2R Tg mice included both
lymphoid and nonlymphoid compartments.
|
A functional memory response must be capable of responding to
rechallenge with the Ag. Because excessive proliferation of
CD8+ T cells has been associated with exhaustion
of the responding cells (41), we examined whether the
increased T cell proliferative response of GMIL2R Tg mice during
primary LCMV infection affected the capacity to mount a recall response
to a secondary LCMV infection. To eliminate the potentially confounding
effects of anti-LCMV-neutralizing Abs on the viral load achieved
during secondary infection that could alter the magnitude of the
CD8+ T cell proliferative response
(54), we examined the recall response in B6 and GMIL2R Tg
mice lacking CD4+ Th function, because such mice
are unable to generate neutralizing Abs following LCMV infection
(44). Thus, cohorts of CD4-deficient B6 and GMIL2R Tg mice
initially infected with 2.5 x 105 PFU of
LCMV were rechallenged 30 days later, after the memory responses had
been established, with 5 x 106 PFU of a
more virulent LCMV strain, clone 13, i.v. (45). Following
rechallenge, mice rapidly clear even the more virulent virus and
display a peak recall response at 5 days after rechallenge (45, 55). Both groups of mice had no detectable virus in spleens by
plaque-forming assay at day 5 postrechallenge. The number of
virus-specific CD8+ T cells in CD4-deficient B6
mice increased 5.6-fold during these 5 days (Fig. 5). CD4-deficient GMIL2R Tg mice
increased the anti-LCMV CD8+ T cell pool
13.7-fold at 5 days following reinfection compared with preinfection
levels. Thus, the GMIL2R+ memory
CD8+ T cells were not only capable of mounting a
functional response to viral rechallenge, but the size of the recall
response was proportionally larger than the response of wild-type
memory CD8+ T cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Following acute LCMV infection, CD4+ Th cells have little impact on the magnitude of the peak CD8+ T cell response achieved. The lack of a requirement for CD4+ Th to license APC for effective stimulation of CD8+ T cells is most likely attributable to the direct maturation/activation of APC by LCMV infection (11). However, the basis for the apparent lack of a contribution by IL-2 produced by responding CD4+ Th cells to the peak of CD8+ T cell expansion is less clear, because our data demonstrate that the availability of IL-2 limits CD8+ T cell expansion. Several reasons may account for this. First, the peak magnitude of the CD4+ Th response to LCMV infection is >20-fold lower than the CD8+ T cell response (56). Thus, CD4+ Th cells may only be able to provide paracrine IL-2 to a minor population of responding CD8+ T cells. Second, because viral infection with LCMV directly infects and matures APC, CD8+ T cells can produce IL-2 and proliferate at sites in which CD4+ Th function might be scarce or absent, particularly because the responding CD8+ T cells can lyse the APC. Finally, the outcome of IL-2R signaling in CD8+ T cells may be tied to the timing of IL-2R signals (18). Because CD4+ Th production of IL-2 is not synchronized to activation of individual responding CD8+ T cells, exposure to paracrine IL-2 could cause both enhanced CD8+ T cell proliferation and cell death, leading to no net effect on the T cell response.
Although CD4+ Th function does not appear to contribute to the magnitude of the acute phase of CD8+ T cell expansion in B6 mice, CD4+ Th cells do play a role in determining the size of the memory CD8+ T cell pool formed following viral clearance (2). The contributions of CD4+ Th function to memory CD8+ T cell formation were much less evident in GMIL2R Tg because the number of memory CD8+ T cells formed after viral infection in CD4-deficient GMIL2R Tg mice was only slightly smaller than in intact GMIL2R Tg mice, whereas in wild-type mice, the presence of CD4+ T cells more than doubled the number of memory CD8+ T cells formed. Augmented IL-2R signals delivered via the chimeric GMIL2R could have several effects on CD8+ T cells, which might directly enhance the generation of memory CD8+ T cells and account for the diminished CD4 dependence in GMIL2R Tg mice. If memory CD8+ T cells are directly derived from the predominant population represented by the peak of expansion, enhanced signaling through an IL-2R during the expansion phase could result in increased expression of antiapoptotic factors such as bcl-2 (57), making the cells better able to survive the contraction phase. Alternatively, if memory CD8+ T cells are derived from a distinct subset of cells responding to the Ag, delivery of signals through an IL-2R might directly promote their proliferation and survival, and/or facilitate differentiation into memory cells.
Increased CD8+ T cell proliferation in the context of a persistent viral infection has been associated with exhaustion and deletion of Ag-specific CD8+ T cell responses (41). This exhaustion of CD8+ T cells has been thought to at least in part be attributable to excessive proliferation of Ag-specific CD8+ T cells Although GMIL2R Tg mice mounted an enhanced acute phase CD8+ T cell response to primary LCMV infection, the increased proliferative response did not limit the capacity of the memory CD8+ T cell pool in GMIL2R Tg mice to mount a secondary response to reinfection. Indeed, rather than exhibiting diminished responses, CD8+ T cells from GMIL2R Tg mice demonstrated stronger responses to rechallenge than wild-type CD8+ T cells, consistent with a positive effect from augmentation of IL-2R signals in secondary responses of CD8+ T cells, as was observed in primary responses. Thus, our data suggest that increased proliferation of CD8+ T cells does not predispose these cells to exhaustion, and are consistent with preliminary observations that exogenous IL-2 can promote the survival and effector function of CD8+ T cells that might normally be deleted in the face of persistent viral infection (41).
Our data suggest different conclusions than those drawn from recent work examining the effects of the loss of IL-2 function on CD8+ T cells during an antiviral CD8+ T cell response (58). Following adoptive transfer of TCR Tg IL-2-/- CD8+ T cells into wild-type or IL-2-/- mice and infection of the recipient mice with a recombinant vesicular stomatitis virus, IL-2 availability did not appear to be a limiting factor for the splenic CD8+ T cell response. By contrast, our data, which demonstrate significantly enhanced anti-LCMV responses in mice with augmented autocrine IL-2R signals, suggest IL-2 availability is limiting. This disparity may be attributable to differences in the immunobiology of LCMV and vesicular stomatitis virus infection, but alternatively the lack of requirement for autocrine IL-2 observed by D’Souza et al. could be due to either altered development of IL-2-/- donor T cells (20), paracrine IL-2 production by CD4+ T cells in wild-type recipient mice, or the loss of contributions by IL-2 during T cell development in the recipient IL-2-/- mice, leading to modified peripheral T cell function, such as from failure to develop CD4+ regulatory T cells (24, 59).
The GMIL2R transduces a signal that is biochemically indistinguishable
from the native IL-2R (33). However, the IL-2R- and
common -chains are activated by binding of IL-15 as well as IL-2,
with both cytokines capable of enhancing CD8+ T
cell proliferation in vitro, making it possible the GMIL2R is mimicking
the effects of either or both IL-2 and IL-15. Although the two
cytokines share these two signaling receptor chains, IL-2 and IL-15
receptors have distinct -chains required for high affinity ligand
binding, and the biologic effects of the cytokines are not entirely
overlapping. For example, T cells become susceptible to AICD following
repetitive TCR stimulation in vitro. Although AICD can be enhanced in
vitro by the presence of pharmacologic doses of IL-2 during TCR
activation, exposure of activated T cells to pharmacologic doses of
IL-15 does not appear to increase apoptosis (60), and
may actually enhance T cell survival (61). Given the
common signaling pathways engaged by IL-2 and IL-15 and the fact that
the cytokines were added at the initiation of culture, such findings
are difficult to explain. The disparate effects might reflect distinct
kinetics of signaling, such as could occur if IL-15R expression is
delayed compared with IL-2R, making the IL-15 signal more prominent
when T cells have completed an initial proliferative burst and have
become less susceptible to AICD (62). Alternatively, the
distinct effects of IL-15 might be mediated by a unique signaling
capability of the IL-15R chain, as compared with the nonsignaling
IL-2R chain (63, 64, 65). Such unique signaling events
would make the chimeric GMIL2R more like the IL-2R, because it lacks a
signaling -chain. The patterns of cytokine expression also differ
between IL-2 and IL-15, and again might make the signals delivered
through the GMIL2R more similar to IL-2R signals. IL-2, similar to
GM-CSF, is produced by activated T cells and is rapidly extinguished
following Ag withdrawal. IL-15, in contrast, is produced primarily by
stromal cells and APC, but not T cells (66), and may not
be abundantly available to rapidly proliferating
CD8+ T cells, particularly in the context of
viral infection of APC (67). However, it should be noted
that IL-15R signaling may contribute to the acute phase of the
antiviral CD8+ T cell response (68),
although it remains unclear whether this reflects effects of IL-15 on T
cells or APC (69). Thus, we cannot exclude the possibility
that some of the effects of the GMIL2R may reflect IL-15R-like rather
than IL-2R-like signals.
Provision of augmented IL-2R signals to CD8+ T cells enhanced T cell responses both during the acute and recall responses, which could potentially be useful clinically. For example, administration of IL-2 to HIV patients during bursts of viral replication such as following structured treatment interruption might be an effective means of enhancing CD8+ T cell immunity to HIV (70, 71, 72). However, regimens for delivering supplemental IL-2 will require adequate penetration of secondary lymphoid organs and other sites at which targets are recognized. Administration of high doses of IL-2 may be able to accomplish this, but can be associated with toxicity that could limit clinical utility (73, 74). For therapeutic settings in which adoptive T cell transfer with Ag-specific CD8+ T cells is an option, it might be possible to overcome these obstacles to effective IL-2 therapy by genetically modifying memory/effector CD8+ T cells with the chimeric GMIL2R before transfer. Such an approach might also obviate the requirement for long-term IL-2 administration to sustain specific T cell responses, and could permit establishment of long-term functional immunity in patients in which a concurrent CD4+ T cell response against either viruses or tumors is chronically deficient or absent.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Philip D. Greenberg. University of Washington, Box 356527, Seattle, WA 98195. E-mail address: pgreen{at}u.washington.edu
3 Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; AICD, activation-induced cell death; Tg, transgenic; GMIL2R, GM-CSF/IL-2R.
Received for publication July 8, 2002. Accepted for publication August 30, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
are two necessary lymphokines in the development of cytolytic T cells. J. Immunol. 143:1210.[Abstract]
production during lymphocytic choriomeningitis virus infection. J. Immunol. 155:5690.[Abstract]
chain, resistance to clonal deletion and IL-2 dependency. Int. Immunol. 10:371. chain. Eur. J. Immunol. 23:2472.[Medline]
and chains mediate the signal for T-cell proliferation. Nature 369:333.[Medline]
c chain sufficient for Jak kinase activation and induction of proliferation in T cells. Mol. Cell. Biol. 16:309.c chain for Janus kinase activation leading to T cell proliferation. Proc. Natl. Acad. Sci. USA 94:1878.-chain in regulating IL-2-dependent, activation-induced CD8+ T cell death. J. Immunol. 163:3131.-mediated apoptosis in fibroblasts by TRAF2 recruitment to the IL-15R chain. FASEB J. 13:1575. chain signals through association with Syk in human B cells. J. Immunol. 167:6292. intracellular trafficking in human melanoma cells and signal transduction through the IL-15R. Oncogene 19:5153.[Medline]
This article has been cited by other articles:
|
|
 |
|
  S. Mostbock, M. E. C. Lutsiak, D. E. Milenic, K. Baidoo, J. Schlom, and H. Sabzevari IL-2/Anti-IL-2 Antibody Complex Enhances Vaccine-Mediated Antigen-Specific CD8+ T Cell Responses and Increases the Ratio of Effector/Memory CD8+ T Cells to Regulatory T Cells J. Immunol., April 1, 2008; 180(7): 5118 - 5129. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Jin, D. Gong, D. Adeegbe, A. L. Bayer, C. Rolle, A. Yu, and T. R. Malek Quantitative assessment concerning the contribution of IL-2R{beta} for superantigen-mediated T cell responses in vivo Int. Immunol., April 1, 2006; 18(4): 565 - 572. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Moore, A. Gallimore, S. J. Draper, K. R. Watkins, S. C. Gilbert, and A. V. S. Hill Anti-CD25 Antibody Enhancement of Vaccine-Induced Immunogenicity: Increased Durable Cellular Immunity with Reduced Immunodominance J. Immunol., December 1, 2005; 175(11): 7264 - 7273. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Teague, R. M. Tempero, S. Thomas, K. Murali-Krishna, and B. H. Nelson Proliferation and Differentiation of CD8+ T Cells in the Absence of IL-2/15 Receptor {beta}-Chain Expression or STAT5 Activation J. Immunol., September 1, 2004; 173(5): 3131 - 3139. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. N. D'Souza and L. Lefrancois IL-2 Is Not Required for the Initiation of CD8 T Cell Cycling but Sustains Expansion J. Immunol., December 1, 2003; 171(11): 5727 - 5735. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Deepe Jr. and R. S. Gibbons Protective and Memory Immunity to Histoplasma capsulatum in the Absence of IL-10 J. Immunol., November 15, 2003; 171(10): 5353 - 5362. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
)
and GMIL2R Tg mice (•), and CD4-deficient B6 (
) and GMIL2R Tg
(
) mice were infected with 2.5 x 105 PFU of LCMV
i.p., and spleens were harvested on days 8, 14, 25, and 50
postinfection. At each time point, the number and frequency of
CD8+ T cells specific for NP396 (B) or gp33
(C) were quantitated by intracellular staining for
IFN-
