| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* National Institute of Immunology, New Delhi, India;
Department of Biological Sciences, University of Arkansas, Fayetteville, AR 72701; and
Rosenstiel Research Center, Brandeis University, Waltham, MA 02254
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In contrast, there is an alternate model in which complete stimulation of T cells would lead to terminal effector differentiation, while partial but still above-threshold triggering would cause differentiation directly into a memory state (6). In such a model suggested by data implicating differential signals (7) and pathways (6) in generating memory vs effector T cells, alternate pathways may be responsible for the generation of short-lived effectors vs long-lived memory cells. The existence of long-lived effector cells (8) complicates these interpretations further.
On this uncertain background it is not surprising that, while adjuvants have been used for a century for generating long-lived immune responses, their use is largely empirical. They commonly evoke inflammation, which has been thought to induce the costimulatory signals on APCs required for inducing T cell memory (9). Recent data suggest that adjuvants may be critical for inhibiting T cell apoptosis (10). We have shown earlier that the presence of pentoxifylline (PF) during priming of human CD4 T cells in vitro against alloantigenic target APCs inhibited the primary proliferative T cell responses but enhanced secondary proliferative responses and reduced the frequency of apoptosis in activated T cells (11, 12). However, it was unclear whether modulation of secondary proliferative responses of T cells by PF, measured over a short time frame in vitro, reflected physiologically relevant modulation of the induction of immune T cell responses in vivo. We now show evidence addressing this potential use of PF as an immune adjuvant for both CD4 and CD8 T cell responses in vivo, probably through its effects on AICD.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6 splenocytes were used as responders and BALB/c splenocytes (irradiated at 1000 rad or irradiated and fixed with 0.003% paraformaldehyde) were used as stimulators. For primary proliferative alloresponses, stimulators in graded doses were used with responder spleen cells (3 x 105/well) in 200 µl of Click’s medium (Irvine, Santa Ana, CA) with L-glutamine (Life Technologies, Rockville, MD), antibiotics, and 10% FCS. After 5 days, cultures were pulsed with [3H]thymidine (Amersham, Little Chalfont, U.K.) for 16 h before being harvested and counted by scintillation spectroscopy (Betaplate; Pharmacia-Wallac, Turku, Finland). Data are shown as mean ± SE of triplicate cultures. For allo-priming in vitro, responder cells (2 x 106 cells/ml) were cultured either by themselves or with stimulators (1 x 106 cells/ml) in the presence or absence of PF (100 µg/ml) for 72 h, rested for 24–48 h, and restimulated with titrated numbers of fresh stimulator APCs for 48–72 h.
T cell immune responses to immunization in vivo
BALB/c or C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME) were immunized with the Ags indicated. All animal experiments were done under approval from the Institutional Animal Ethics Committee. Pure protein Ags were used in a single i.p. dose of 1 mg per mouse (native Ags) and 300 µg per mouse (maleylated Ags), while in CFA they were given as a single s.c. dose of 100 µg per mouse. Proteins were maleylated as previously described (13). For alloimmunization, C57BL/6 mice were immunized s.c. with a single dose of gamma-irradiated (500 rad) BALB/c spleen cells (10–30 million cells per mouse). For immunization with Salmonella typhimurium (Stm), mice were given a single dose of 106 CFUs of the auxotrophic aroA mutant of Stm (Stm-aroA) per mouse i.p. as previously described (14). Immunized mice were treated as appropriate with various pharmacological modulators i.p. daily from day -2 to day +5 of immunization. The doses used were arrived at after preliminary titration and were 1–2 mg per mouse per day for PF, 1 mg per mouse per day for dibutyryl cAMP (dbcAMP), 10 µg per mouse per day for Rolipram (Rm), and 10 nmol per mouse per day for Rp-8-Br-cAMPs (RpBrcAMPs).
For measuring immune response recall in vitro in bulk assays, splenic
cells taken on the days indicated after immunization were stimulated in
vitro with titrated Ag doses. At 3 days, induced levels of IFN-
were
estimated in the culture supernatants using commercial enzyme
immunoassays (R&D Systems, Minneapolis, MN). In replicate cultures,
proliferation was estimated as above at 72–96 h. For cells from
Stm-aroA-immunized mice, sonicated aqueous extract of Stm was used as
the recall Ag (14).
Clonal responder frequency and burst size analysis of Ag-specific precursor cells in the various groups of immunized mice was estimated by limiting dilution analysis (LDA) as reported earlier (12). Briefly, C57BL/6 mice were immunized s.c. with maleyl-OVA in CFA (100 µg per mouse) with or without PF cover as above. After 8 wk, titrating numbers of lymph node cells from various groups of mice were added to a constant number of gamma-irradiated spleen cells from unimmunized syngeneic mice as APCs, with or without maleyl-OVA (100 µg/ml). Of the wells plated per responder dose, those wells showing a response at least 3-fold that of the background value without maleyl-OVA were scored positive and the responder frequency was calculated as described (12). The response was considered clonal at the responder cell number where the proportion of positively responding wells was <37% (12). At responder cell concentrations showing clonal frequencies, the average cpm values of responding (vs nonresponding) wells were used as an estimate of the average clonal burst size of responding single T cells.
For in vitro tracking of alloresponsive CD4 and CD8 T cells, B6.Thy-1.1 congenic mice were immunized s.c. with irradiated BALB/c spleen cells with or without PF cover, and 7 wk later lymph node cells from individual immunized or unimmunized mice were labeled with 10 µM CFSE (Molecular Probes, Eugene, OR) for 15 min at 37°C and restimulated with gamma-irradiated BALB/c spleen cells as APCs. Four days later, live cells were stained for Thy-1.1 vs CD4 or CD8, and the degree of CFSE dilution on gated Thy-1.1-bearing responder CD4 or CD8 cells in unstimulated and stimulated cultures from individual mice was determined flow cytometrically. The frequencies of CD4 or CD8 cells undergoing various numbers of divisions were estimated using the proliferation platform of FlowJo software (TreeStar, San Jose, CA).
Bacterial clearance assays
T cells purified from splenic cells by passage over nylon-wool columns were transferred i.p. into naive syngeneic recipients (107 T cells per mouse), which were challenged with 100 CFUs i.p. of a virulent strain of Stm, Stm-754 (15), and 24 h later the bacterial load in their spleens was estimated by titrating splenic lysates on Salmonella-Shigella-agar plates for bacterial CFU counting.
Induction of superantigen-mediated T cell activation and loss in vivo
MHC-matched (H-2k) mouse strains CBA/J and
CBA/CaJ, which differ in their mls superantigen status, were
used. CBA/CaJ mice given 1 x 107 CBA/J
spleen cells i.p. were treated with either PBS or PF as above. Splenic
cells from these recipients, as well as from control untreated CBA/J
and CBA/CaJ mice, were stained with anti-TCR-V
6 (BD Biosciences,
Franklin Lakes, NJ) for flow cytometric analysis. Also, these spleen
cells were stimulated in vitro on day 7 postinjection with titrated
doses of gamma-irradiated CBA/J spleen cell APCs for 3 days and
proliferative responses were measured.
In cell transfers in the converse direction, CBA/J mice were given
CBA/CaJ spleen cells i.v. (3 x 107 cells
per mouse) and were treated with either PBS or PF daily from day -2 to
the day of analysis. At indicated time points postinjection, splenic
cells from recipient mice as well as from control CBA/J and CBA/CaJ
mice were stained with either anti-CD4 or anti-CD8 together
with anti-TCR-V6 (BD Biosciences) for two-color flow cytometric
analysis.
Flow cytometric analyses were done on a BD LSR (BD Biosciences), an
Elite ESP (Beckman Coulter, Fullerton, CA), or a Bryte (Bio-Rad, Hemel
Hampstead, U.K.) flow cytometer. Data were analyzed using either WinMDI
shareware or FlowJo software (TreeStar). In some experiments, the
CBA/CaJ spleen cells were labeled with CFSE before transfer, and
staining was done with anti-CD4, anti-CD8, and
anti-TCR-V6 for multicolor flow cytometry. For detection of
induction of activation markers on injected responder T cells,
recipient spleen cells were taken at 48 h after injection and
stained for TCR-V6 vs CD44 or CD95L (BD Biosciences) for two-color
flow cytometry.
Induction of T cell AICD in vitro
Spleen cells from C57BL/6 mice were incubated in the presence or
absence of the anti-CD3
mAb, C363.29B (0.1 µg/ml; gift of Dr.
C. A. Janeway, Yale Medical School, New Haven, CT), with or without the
addition of various pharmacological modulators. Between 24 and 96
h later, the cells were stained for either CD4 or CD8 vs annexin V (BD
Biosciences) to detect apoptotic cells. The induction of caspases in
these cells was detected using a cell-permeable fluorescent caspase
substrate, VAD-fluoromethylketone-fluorescein (VAD-fmk-Flu; Promega,
Madison, WI).
CD8 T cell responses
Female C57BL/6 mice were immunized with 1 x 107 male spleen cells i.p. and treated with either PBS or PF as above. At day 60 postimmunization (p.i.), splenic cells were stimulated in vitro for 6 days with male APCs and then were used as effector cells in various E:T ratios to lyse 51Cr-loaded target RMA-S cells (H-2b) incubated with either an irrelevant H-2Db-binding lymphocytic choriomeningitis virus (LCMV)-derived peptide (KAVYNFATM; pLCMV) or with the H-2Db-binding H-Y peptide (KCSRNRQYL (16); pHY) in a 4-h cytotoxicity assay to determine the percentage of cytotoxicity.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have previously observed that PF inhibits primary allospecific
proliferation of human CD4 T cells in vitro (11, 12).
Primary allostimulated proliferation of mouse T cells is also similarly
affected by PF (Fig. 1
a). In
comparison to unprimed cells, primed cells show enhanced secondary
proliferative responses when restimulated with fresh stimulators.
Significantly, the presence of PF during priming enhanced this
secondary alloproliferative response (Fig. 1b). The
stimulators used were irradiated and lightly fixed to ensure that the
effect of PF was directly on the responding T cells. The cells
proliferating in these assays were confirmed to be CD4 by inhibition
with anti-CD4 mAb (data not shown).
|
c). In
contrast, if the mice were treated daily with PF from day -2 to day +5
of immunization, T cell responses could be recalled very efficiently at
this time (Fig. 1c). PF exhibited this effect in OVA
immunization with adjuvants such as alum or CFA as well (data not
shown). Immune priming in vivo by PF is mediated through protein kinase A (PKA)
Mice were immunized with maleylated conalbumin (CA), which is
delivered specifically to scavenger receptors on APCs and function as
better immunogens than native CA (13, 17), and treated as
above with PBS, PF, Rm, an inhibitor of the PDE4 isoform prominently
found in T cells (18), or the cell-permeable cAMP agonist
analog dbcAMP. When the recall T cell responses of these mice were
analyzed at 42 days p.i., transient treatment with PF, Rm, or dbcAMP
resulted in significantly higher proliferative responses than PBS
treatment yielded (Fig. 2a).
We next tested the effects of PKA inhibition by a diastereoisomer of
cAMP, RpBrcAMPs (19), on PF-mediated enhancement of T cell
priming. When T cell proliferative responses of maleyl-OVA-immunized
mice were analyzed at 42 days p.i., it was observed that, while
transient PF treatment led to enhanced recall responses, simultaneous
treatment with the PKA inhibitor RpBrcAMPs blocked this enhanced
commitment (Fig. 2b). However, treatment with the PKA
inhibitor alone did not modify T cell recall responses.
|
The effect of PF on immunization with a bacterial immunogen was
examined next. We immunized mice with live Stm-aroA bacteria as
described previously (14), along with PBS or PF treatment
as above. Recall proliferative and cytokine responses were evoked from
the spleen cells of these mice in vitro with a sonicated extract of Stm
45 days later. Mice immunized under PF cover showed a significant
increase in the proliferative responses (Fig. 3a), which we have confirmed
to be from CD4 T cells by blocking with anti-CD4 Ab (data not
shown). Similarly, IFN- levels were significantly higher in recall
cultures from mice immunized under PF cover (Fig. 3b). The
levels of other T cell cytokines such as IL-10 were also enhanced in
these cultures (data not shown).
|
c). Inhibition of postactivation T cell loss in vivo by PF
We next began testing the effects of PF on T cell activation and
AICD in vivo, using the well-characterized model of
superantigen-mediated activation and deletion of T cells bearing
certain TCR-V elements (20, 21), in the MHC-matched
(H-2k) mls-disparate mouse strains CBA/J
(mls-stimulator) and CBA/CaJ (mls-responder). We transferred
mls-responder T cells into mls-stimulator recipients with or without PF
treatment. The frequency of TCR-V6-bearing cells was higher in
PF-treated mice than in untreated mice on day 7, in both CD4 and CD8 T
cell populations (Fig. 4a).
Superantigen-mediated T cell apoptosis has been reported to be
dependent on initial proliferation (22), and the very high
frequencies of TCR-V6-bearing cells in PF-treated recipients (Fig. 4a) indicated that PF may not inhibit T cell proliferation
in vivo and may instead enhance survival.
|
6-bearing cells
was further examined by marking the input cells from mls-responder mice
with the fluorescent dye CFSE before being transferred into
mls-stimulator recipients as before. By 72 h posttransfer,
CFSE-labeled TCR-V6-bearing cells had proliferated to the point of
losing the fluorescein label irrespective of PF treatment (Fig. 4b). By 7 days posttransfer, there was significant reduction
in the frequency of TCR-V6-bearing cells in the PBS-treated
recipients, while these cells persisted at virtually unchanged levels
in PF-treated mice (Fig. 4b). Thus, while PF inhibited
TCR-V6-bearing T cell loss in vivo, it did not appear to slow down
their activation-induced proliferation to any significant degree.
Because TCR-V6-bearing cells lost CFSE label by 72 h, we
examined the effect of PF on their rate of proliferation by sampling
shorter time points. At 24 h posttransfer, CFSE-labeled
TCR-V6-bearing cells could be detected in mls-responder CBA/CaJ
recipients as well as in mls-stimulator CBA/J recipients with or
without PF treatment (Fig. 5a). However, while no
significant proliferation had as yet begun at this time, both CD4 and
CD8 TCR-V6-bearing cells had already become larger than their
counterparts in mls-responder mice in both PF-treated and
saline-treated mice (Fig. 5a). By 48 h posttransfer,
TCR-V6-bearing CD4 and CD8 cells had begun to proliferate, and PF
treatment made no difference in the number of cell divisions they went
through by this time (Fig. 5b). We also examined whether PF
modified the induction of cell surface activation markers on
TCR-V6-bearing T cells upon injection into mls-stimulator mice. The
enhancement of CD95L and CD44 levels on TCR-V6-bearing cells 2 days
after injection into mls-stimulator recipients was not inhibited by PF
(Fig. 5c). Activation markers were not induced if
mls-responder cells were transferred into mls-responder syngeneic
recipients, and PF did not affect the induction of two other T cell
activation markers, CD69 and CD122 (data not shown). Thus, mls-driven T
cell activation appeared unaffected by PF in vivo.
|
6-bearing T
cells in these groups of mice were followed flow cytometrically at days
0, 3, 7, or 14 postinjection. Significant loss of TCR-V6-bearing
cells was seen in mls-responder mice after injection of mls-stimulator
APCs, and this loss was prevented by PF in vivo (Fig. 6a). However,
TCR-V6-bearing cells in mice given PF in vivo did show an increase
in size at early time points (Fig. 6a). One week after cell
injection, the responses of the splenic T cells from the mls-responder
recipients to mls-stimulator APCs in vitro were determined. While
normal mls-responder cells proliferated well, cells from mls-responder
mice that had been given mls-stimulator APCs in vivo did not (Fig. 6b). However, mls-responder mice given PF in addition to
mls-stimulator APCs in vivo showed an excellent maintenance of mls-1
reactivity (Fig. 6b). Thus, PF-mediated protection of
responding T cells is not accompanied by induction of anergy.
|
The mls-based system used above established that PF inhibited the
loss of T cells in vivo consequent to activation. Therefore, a direct
measurement of AICD was next undertaken. We stimulated normal mouse
lymph node cells in vitro with an anti-CD3 mAb in the presence
or absence of PF. The cultured cells were stained for either CD4 or CD8
vs annexin V to detect apoptotic cells 4 days later. While anti-CD3
mAb induced substantial apoptosis in both CD4 and CD8 T cells, the
presence of PF significantly inhibited the apoptosis (Fig. 7). Thus, PF protected not only CD4 but
also CD8 T cells against AICD. PF also inhibited the
anti-CD3-induced activation of caspases, the cysteine aspartyl
proteases that mediate many AICD pathways, in these T cells at 48
h (Fig. 7).
|
Finally, we tested for the potential enhancement of CD8 T cell
memory by PF in vivo using immunization for the male-specific Ag H-Y
(16). Female mice were primed with a single dose of
syngeneic male spleen cells either with or without PF cover, and
H-Y-specific cytotoxic T cell CTL precursors were assayed 60 days p.i.
Mice immunized under PBS cover showed poor CD8 T cell responsiveness by
day 60 p.i. (Fig. 8). In contrast,
if the mice were transiently treated with PF, H-Y-specific CTL
responses could be recalled very efficiently even at day 60 p.i.
(Fig. 8).
|
We next examined whether PF cover in immunized mice enhanced the
frequency and/or the clonal burst size of Ag-specific T cells. C57BL/6
mice (three mice per group) were immunized s.c. with maleyl-OVA with or
without PF cover from day -1 to day +5 of immunization. Seven weeks
later, LDA was done with maleyl-OVA stimulation of lymph node cells to
estimate the frequencies and clonal burst sizes of the memory CD4 T
cells. PF treatment significantly increased the clonal frequency of
maleyl-OVA-specific T cells (Fig. 9a). In contrast, clonal burst
size as indicated by the proliferation seen in responding
wells at clonal frequency (<37%) showed no difference between
cultures from mice treated with saline or PF (Fig. 9a).
|
b). While the frequency of large cells undergoing division
is greater in the PF-treated immune mice than in the saline-treated
immune mice, the extent of proliferation of individual responding cells
does not appear to be different. When the frequencies of CD4 and CD8
cells undergoing various numbers of divisions were estimated, it could
be observed that, while the frequencies of responding cells were higher
in PF-treated mice at every cell division, neither CD4 nor CD8 T cells
from mice immunized under PF cover have undergone clonal division at
any more rapid rate than cells from mice immunized without PF cover
(Fig. 9c). Thus, the frequency, but not the clonal burst
size, of Ag-specific proliferation-competent T cells increases if the
mice are treated with PF at the time of immunization. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To begin with, we have confirmed that PF has the same effect on allo-priming of mouse T cells in vitro that we have reported it has on human T cell priming (11, 12). The presence of PF during allo-priming in vitro enhances the resultant secondary alloresponsiveness of T cells, an effect likely to be directly on T cell signaling, because the allostimulators used were fixed and show no evidence of APC activation or cytokine secretion from them (data not shown).
In vivo, we have used PF treatment in a clinically used dose range for 5 days into primary immunization, starting from just before immunization. Thus, exposure to PF is transient and restricted to only the first few days while T cell priming in vivo is taking place. Thereafter, mice were maintained for varying periods of time before being tested for persistence of recall responses. The initial experiments used single doses of pure protein Ags such as OVA or CA without any adjuvant at all because persistence of immune responses is poor in the absence of adjuvant, to see whether PF could function on its own to enhance the persistence of recall T cell responses. In some experiments, we have used the maleylated forms of these protein Ags as immunogens, because they show enhanced immunogenicity without providing any Ag-nonspecific adjuvant activity (13, 17). We have also observed this adjuvant effect of PF during immunization with adjuvants such as alum or CFA. In all these situations, transient PF treatment early during immunization resulted in enhanced persistence of recall CD4 T cell proliferative responses in the immunized mice, establishing that PF could indeed function in vivo as an immune adjuvant.
PF is an inhibitor of PDEs, and most of its functions are thought to be mediated by enhanced cAMP levels as a result of reduced degradation of cAMP via PDE (23). This would predict that other modalities of enhancing cAMP levels would have effects similar to that of PF. PDE4 is a prominent PDE isoform in T cells and Rm is a specific inhibitor of PDE4 (18) that is structurally unrelated to PF. Transient treatment during immunization with either Rm or the cell-permeable analog of cAMP, dbcAMP, showed the same kind of enhancement in T cell recall responses as that shown by PF treatment. Thus, enhanced cAMP levels for a few days during T cell priming in vivo is sufficient to enhance the persistence of T cell recall responses.
The major pathway of signal transduction by cAMP is mediated through the cAMP-dependent PKA, although there are PKA-independent pathways suggested as well. The PKA inhibitor we have used, RpBrcAMPs, inhibits release of the catalytic subunits of PKA from the regulatory subunits, preventing their activation (19). Inhibition of the PF-mediated immune enhancement by simultaneous treatment with the PKA inhibitor provides evidence that the immune adjuvant effect of PF is indeed mediated through cAMP-dependent PKA activation.
The data discussed so far deal with adjuvant-free immunization with pure proteins and with immune responses read out in vitro as Ag-induced proliferation. However, in vitro recall responses do not necessarily indicate functional capabilities in vivo. Therefore, it was essential to examine whether the PF-mediated immune enhancement observed extended to immune protection in vivo against infections. For this purpose, we have tested the effect of PF on immunization with a Salmonella strain Stm-aroA, a vaccine vector (24, 25) we have worked with previously (14).
The results with Stm immunization and PF usage make three points.
First, there is a clear PF-mediated enhancement of the proliferative
recall responses induced by Stm-aroA immunization. Second, this
enhancement extends to the generation of T cell cytokines as well,
specifically IFN-, the cytokine crucially responsible for providing
protective T cell immunity against facultative intracellular pathogens
such as Stm (26). This is significant in light of
conflicting indirect reports about whether PF affects the induction of
a Th1 cytokine, IFN-, in CD4 T cell responses (27, 28).
In fact, we observe no major PF-mediated alteration in the balance of
cytokine profiles of T cells primed in vivo under PF cover (data not
shown). Third, adoptive transfer experiments show that the anti-Stm
T cells generated under PF cover also show enhanced functionality in
vivo in clearing Stm infection. Thus, PF is clearly useful as an agent
to enhance CD4 T cell responses to vaccination in vivo.
While our data show that this effect of PF is mediated through PKA, it
was not clear what downstream event in T cell activation was being
targeted. Our preliminary data in the system using human T cell
allo-priming in vitro had indicated that the induction of
activation-mediated T cell death could be reduced by PF
(12). To examine whether this was also the case in vivo,
it was necessary to track responding T cells in vivo. Because T cells
responding to protein immunogens in normal mice are difficult to
identify and to track phenotypically, we used a well-established model
of inducing activation and death in T cells bearing specific TCR-V
segments via superantigens (20, 29).
In the mls-based model we have used, a superantigenic protein, v-sag-7,
coded by the integrated retroviral gene Mtv-7, leads to activation of
TCR-V-6-expressing T cells. In a mouse strain that does express
Mtv-7 endogenously, such as CBA/J, the TCR-V-6-bearing T cells are
deleted in the thymus during development and are not seen in the
peripheral lymphoid tissues (29). However, while a related
strain, CBA/CaJ, is syngeneic to CBA/J in other respects, it does not
have the Mtv-7 integration and therefore shows a significant component
of TCR-V-6-expressing T cells in peripheral lymphoid organs. When
mature CBA/CaJ TCR-V-6-bearing T cells were injected into
mls-responder CBA/J mice, PF had no effect on the resultant activation
(as indicated by the expression of activation markers such as CD44 and
CD95) and extensive proliferation (as indicated by dilution of the
fluorescent cytoplasm marker, carboxyfluorescein) of these cells, but
it prevented the cell loss that followed.
An interesting point here is the divergence of the effect of PF on T cell proliferation in vitro and in vivo. While PF is inhibitory for T cell proliferation in vitro, it does not appear to have the same consequence in vivo. Although we do not have direct evidence of an explanation of this difference, it may be noted that PF specifically inhibits the induction of the transcription factor c-rel in T cells, which is crucial for the induction of IL-2 transcription (30), and the absence of IL-2 is known to lead to a far greater compromising of T cell proliferation in vitro than in vivo (31, 32).
An additional issue was the reactivity status of T cells that were activated under PF cover so that their loss was prevented. This is a particularly significant issue in light of previous reports that elevating cAMP leads to T cell anergy rather than to enhanced memory (33, 34). It was necessary to confirm that T cells protected from deletion by PF remained Ag responsive rather than becoming anergic. We find that, while injection of mls-stimulator APCs into mls-responder mice caused tolerance in the responder mice, in part by deletion and in part by anergy as reported earlier (20, 35, 36), concomitant PF treatment not only increased the frequency of mls-reactive T cells surviving deletion but also maintained their superantigen responsiveness. Previous reports regarding the induction of anergy by increased cAMP have used secondary T cells, most commonly as stable T cell clones (33), and while primary T cells have been used in one of these reports, the anergy induced in them has been the result of blockade of CD40-CD40 ligand interaction rather than by enhancement of cAMP levels (34). Long-term T cell clones may show quite different signaling pathways in comparison to naive T cells, particularly naive T cells in vivo, and these data together may indicate an interesting difference between priming vs restimulation of Ag-specific T cells.
However, our data show that T cell activation and proliferation in vivo are not affected by PF, nor is there any evidence for induction of T cell anergy due to PF cover. Instead, our data so far suggested that PF could be directly inhibiting AICD of T cells. In vitro experiments demonstrate that anti-CD3 mAb-mediated induction of apoptosis, as detected both by the activation of caspases in the T cells and by the membrane reorganization characteristic of apoptosis (i.e., the exposure of the phosphatidylserine polar head groups detected by binding of annexin V), is inhibited by PF, establishing the ability of PF to inhibit AICD in T cells.
Are these two properties of PF (i.e., enhancing the persistence of recall T cell responses and inhibiting T cell apoptosis) causally related? Interestingly, while CD4 and CD8 T cells may differ in their requirements for memory commitment and in the AICD pathways they follow (37, 38), PF nonetheless inhibited the loss in vivo of activated mls-reactive CD8 T cells and inhibited anti-CD3-induced AICD in them in vitro. If inhibition of apoptosis is sufficient for causing enhanced T cell priming, CD8 T cells should also show PF-mediated increases in the persistence of recall responses. This hypothesis was borne out when the effect of PF on persistence of CD8 T cell responses was tested, reinforcing the probability of a causal relationship between inhibition of T cell AICD and enhancement of persistence of T cell responsiveness.
One prediction of our explanation is that PF treatment during
immunization would lead to an increase in the memory T cell frequency
rather than in the efficiency of their response. We have shown
previously that, in human CD4 T cells allo-primed in vitro, PF cover
during priming leads to an increase in the frequency but not
the clonal burst size of Ag-specific secondary T cells. We have now
also shown that memory T cells, both CD4 and CD8, generated in vivo by
immunization under PF cover, show an increase in frequency but not in
clonal burst size. We have used LDA assays with maleyl-OVA immunization
for this purpose, as well as tracking proliferating primed
alloresponder cells in vitro by CFSE dilution. In preliminary
experiments, we have also examined the frequencies and number of cell
divisions undergone by maleyl-OVA-specific CD4 T cells from mice
immunized >6 wk earlier with or without PF cover, and we find that,
while the frequencies of responding cells are increased by PF cover,
the modal number of cell divisions responding T cells undergo is not
different. In other experiments, we have also estimated the frequency
of Ag-specific T cells surviving in vivo without taking recourse to in
vitro challenge, by using pigeon cytochrome c as an
immunogen in H-2k mice, where the
I-Ek-restricted CD4 T cells responding to the
immunodominant peptide mainly use TCR-V3 and TCR-V
11. In pigeon
cytochrome c-immunized H-2k mice,
there is an increase in the frequency of TCR-V3-V11-expressing
CD4 T cells with a secondary phenotype
(CD44highCD62Llow), and PF
treatment significantly increases this frequency as early as 1 wk after
immunization (data not shown). Together, all these data confirm that PF
causes an increase in the frequency of surviving Ag-specific T cells,
rather than any increase in their burst size, supporting the
probability that PF leads to an enhancement of memory T cell frequency
through inhibition of AICD during priming.
While our data appear to support a linear model of memory T cell generation in which all activated T cells first differentiate into effectors and some of these persist as memory by escaping AICD (4, 5), it is not clear whether all activated T cells rescued from apoptosis would necessarily enter the memory pathway, or whether qualitatively distinct signaling pathways are needed for memory differentiation (6, 7). Further analysis of the phenotype of the Ag-specific T cells triggered under cover of PF would allow this issue to be addressed. Also, the characterization of the T cell death pathway affected by PF would allow a new approach to the relationship between death and memory in T cells.
Thus, these data show that transient PF treatment early during immunization generates longer-lasting T cell memory in vivo via a cAMP-PKA-mediated pathway by inhibiting AICD-mediated T cell loss rather than inhibiting T cell activation per se. In passing, it is curious that the cAMP-dependent signaling pathway, which is critical for long-term neural memory (39), can also apparently contribute to T cell memory. Therefore, our observations identify a novel class of immunological adjuvants for vaccination, especially because PF enhances immune memory in both CD4 and CD8 T cells without altering the T cell cytokine balance. PF is commonly used clinically, and its doses used here are within the clinical range (23, 40). The cAMP-dependent pathway has been extensively analyzed, making it possible to find other potential drugs even more efficient at enhancing immune memory.
|
Footnotes |
|---|
2 R.S. and M.V. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Satyajit Rath or Dr. Vineeta Bal, National Institute of Immunology, Aruna Asaf Ali Road, New Delhi 110067, India. E-mail addresses: satyajit@nii.res.in or vineeta{at}nii.res.in, or Dr. Jeannine M. Durdik, Department of Biological Sciences, University of Arkansas, Fayetteville, AR 72701. E-mail address: jdurdik{at}uark.edu
4 Abbreviations used in this paper: AICD, activation-induced cell death; CA, conalbumin; dbcAMP, dibutyryl cAMP; LCMV, lymphocytic choriomeningitis virus; LDA, limiting dilution analysis; PDE, phosphodiesterase; PF, pentoxifylline; p.i., postimmunization; PKA, protein kinase A; Rm, Rolipram; RpBRcAMP, Rp-8-Br-cAMP; Stm, Salmonella typhimurium; VAD-fmk-Flu, VAD-fluoromethylketone-fluorescein.
Received for publication November 13, 2001. Accepted for publication August 14, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1 and interleukin 2 synergize to prevent apoptosis and promote effector expansion. J. Exp. Med. 182:699.-dominant versus an interleukin-4-dominant immune response. Int. Immunol. 5:1431. action. J. Biol. Chem. 270:20599. in host defense against mycobacteria and Salmonella in mice and men. Curr. Opin. Immunol. 11:346.[Medline]
This article has been cited by other articles:
|
|
 |
|
  K. Sliwa, O. Forster, E. Libhaber, J. D. Fett, J. B. Sundstrom, D. Hilfiker-Kleiner, and A. A. Ansari Peripartum cardiomyopathy: inflammatory markers as predictors of outcome in 100 prospectively studied patients Eur. Heart J., February 2, 2006; 27(4): 441 - 446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Sliwa, A. Damasceno, and B. M. Mayosi Epidemiology and Etiology of Cardiomyopathy in Africa Circulation, December 6, 2005; 112(23): 3577 - 3583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Sliwa, A. Woodiwiss, V. N. Kone, G. Candy, D. Badenhorst, G. Norton, C. Zambakides, F. Peters, and R. Essop Therapy of Ischemic Cardiomyopathy With the Immunomodulating Agent Pentoxifylline: Results of a Randomized Study Circulation, February 17, 2004; 109(6): 750 - 755. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Yang, K. W. McIntyre, R. M. Townsend, H. H. Shen, W. J. Pitts, J. H. Dodd, S. G. Nadler, M. McKinnon, and A. J. Watson Phosphodiesterase 7A-Deficient Mice Have Functional T Cells J. Immunol., December 15, 2003; 171(12): 6414 - 6420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Vig, A. George, R. Sen, J. Durdik, S. Rath, and V. Bal Commitment of Activated T Cells to Secondary Responsiveness Is Enhanced by Signals Mediated by cAMP-Dependent Protein Kinase A-I Mol. Pharmacol., December 1, 2002; 62(6): 1471 - 1481. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
