| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Immunoregulation Unit, Autoimmunity Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892;
Emory Vaccine Center, Department of Microbiology, and University School of Medicine, Emory University, Atlanta, GA 30322;
Laboratory of Molecular Immunology, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD 20892;
Molecular Immunology and Inflammation Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD 20892; and
¶
Instituto de Biofísica Carlos Chagas Filho, Centro de Ciências da Saúde, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Activated T cells from humans and mice harboring CD95 mutations have defects in apoptosis due to TCR restimulation or direct stimulation through CD95 (6, 7, 8). However, recent experiments have indicated that FADD and caspase-8 may have other functions apart from mediating apoptosis. Both FADD- and caspase-8-deficient mice die in utero from similar developmental defects not clearly related to defective apoptosis (9, 10). T cells deficient in FADD display thymic developmental defects and an inability to proliferate in response to IL-2 (11). Mice with overexpression of a FADD death-domain only (FADD-DN) transgene show variable peripheral T cell proliferative defects (12, 13, 14). FADD can be found in the nucleus, is phosphorylated in a cell-cycle-specific manner, and interacts with a cell-cycle-regulated kinase (15). These results suggested that FADD influences T cell activation in a caspase-independent manner through interaction with components of the cell cycle machinery. Pharmacological inhibition of caspases during in vitro activation of human or mouse T cells causes partial inhibition of proliferation and IL-2 production (16, 17). More recently, an immunodeficiency syndrome with defective lymphocyte activation was described in association with a homozygous caspase-8 R248W mutation that abrogated enzymatic activity and destabilized the mutant protein (18). Whether this nonapoptotic caspase-dependent pathway is required for mature T or B cell differentiation or memory cell formation is not known.
In this study, we have expressed the MC159 viral FLIP (vFLIP) protein in the T cell lineage in transgenic mice to probe the function of death-receptor signaling in T cell development and homeostasis in vivo. Additionally, we were interested in how vFLIP expression may modify immune responses of T cells expressing this viral gene product. FLIP proteins are a family of cellular and viral proteins encoding death-effector domains that potently interfere with caspase-8 recruitment to death-domain-containing receptors and subsequent caspase-8 activation that initiates the apoptotic caspase cascade (19, 20). Viral FLIPs are nonessential viral genes present in the genomes of poxviruses and lymphotropic 
-herpesviruses including human herpes virus 8, which is associated with Kaposi sarcomas and certain non-Hodgkin B cell lymphomas in humans (21, 22). The MC159 vFLIP protein, encoded by the human Molluscum contagiosum poxvirus, blocks apoptotic signaling initiated through all known death-domain-containing TNFR family receptors by interfering with caspase-8 recruitment to FADD and subsequent activation of the caspase-8 proenzyme (23). Viral FLIP or its cellular homologue, cellular FLIPS (c-FLIPS), prevents recruitment and/or autoprocessing of caspase-8 in the signaling complex of CD95, and presumably other death receptors as well (24). The long form of c-FLIP, c-FLIPL, can inhibit death-receptor-induced apoptosis in cell lines, but also encodes a pseudo-caspase domain that allows partial processing of caspase-8 and may also stimulate the NF-
B and extracellular signal-regulated kinase nonapoptotic signaling pathways (25, 26). In transgenic mice overexpressing the equine herpesvirus E8 vFLIP in the thymus under the control of the proximal lck promoter, subtle developmental defects and resistance to CD95-induced apoptosis were seen, but peripheral T cell responses were not studied (27). Whether vFLIP has other effects on the peripheral immune system besides blocking apoptosis is not known.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The CD2-vFLIP transgene construct was generated by cloning the coding sequence of the MC159 vFLIP (23) with a COOH-terminal hemagglutinin (HA) tag into the EcoRI site of a CD2 enhancer cassette construct (28). Correct orientation of the insert was confirmed by restriction digestion and sequencing, and the 11.4-kb fragment for pronuclear injection was released by SalI/NotI digestion. Founder mice were generated on the C57BL/6 (B6) background, and DNA was screened by PCR for MC159 integration using the specific primers GACTACGCATCCGACTCCAAGGAGGTCCCTAGC (forward) and CGGAATTCTCAAGTCGTTTGCTCGGGGCT (reverse) and control mouse 
-globulin primers CCAATCTGCTCACACAGGATAGAGAGGGCAGG (forward) and CCTTGAGGCTGTCCAAGTGATTCAGGCCATCG (reverse). Integration was confirmed in positive founder lines using Southern blotting with the EcoRI fragment of the transgene. Repeated attempts to generate lines other than line 4817 that expressed detectable levels of the vFLIP transgene were unsuccessful. Transgenic mice were maintained in a heterozygous state on a B6 background or were crossed to the homozygous TCR transgenic line OT-I (29) or 5CC7 (30). All animal procedures were done under approved animal protocols. B6.SJL CD45.1 congenic mice were obtained from The Jackson Laboratory (Bar Harbor, ME). For Trypanasoma cruzi infections, mice (age 3 mo) were infected i.p. with 106 Vero cell-derived trypomastigotes of T. cruzi clone Dm28c, and parasitemia was monitored by microscopic examination of peripheral blood samples (31). For lymphocytic choriomeningitis virus (LCMV) infections, 2 x 105 PFU of the Armstrong strain of LCMV was injected i.p. as previously described (32).
Cell purification and labeling
CD4 and CD8 T cells were purified from red-cell-depleted single cell suspensions of spleen and/or lymph nodes with T cell enrichment columns (R&D Systems, Minneapolis, MN) and magnetic bead CD4 or CD8 negative selection reagents (Miltenyi Biotec, Auburn, CA) according to the manufacturer’s instructions. For CFSE labeling, cells were suspended in serum-free RPMI 1640 at a concentration of 2 x 107 cells/ml. CFSE (Molecular Probes, Eugene, OR) was added at a final concentration of 5 µM and the cells were incubated for 8 min in the dark at room temperature. Labeled cells were then washed once with FCS and three times with complete RPMI 1640. All other Abs were obtained from BD PharMingen (San Diego, CA) except for anti-HA-FITC (Covance, Princeton, NJ), anti-phospho-STAT 5A/B specific for Y964, and control anti-pan-STAT5 (Zymed Laboratories, San Francisco, CA).
In vitro T cell activation, growth, and cytokine assays
T cells were activated with either Con A (5 µg/ml; Sigma-Aldrich, St. Louis, MO) or anti-CD3
(145-2C11) and anti-CD28 mAbs (BD PharMingen) coated on tissue-culture plates at the indicated concentrations. For proliferation assays, cells were grown at 1 x 106 cells/well and pulsed with 1 µCi of [3H]thymidine for the last 24 h of growth. Supernatants were harvested at 24 and 48 h for measurement of secreted cytokines by ELISA (R&D Systems) or cytometric bead array (BD PharMingen) according to the manufacturers instructions. For extended proliferation assays, cells were placed in new wells at 1 x 106 cells/well with 10 U/ml recombinant human IL-2 (Teceleukin; National Cancer Institute Biological Resources Branch Preclinical Repository, Frederick, MD). Medium was changed and replaced with fresh IL-2 every 48 h. To calculate cell yield and viability, cells were counted by trypan blue exclusion on a hemocytometer. For calcium flux assays, lymphocytes were labeled with 4 µg/ml fluo-4-acetoxymethyl ester (fluo-4-AM) and 10 µg/ml fura red (Molecular Probes) at 37°C for 30 min and subsequently were labeled with Abs to surface markers. Cells were washed twice and kept at room temperature until directly before analysis, when they were warmed to 37°C. Abs were directly added and mixed with cell suspensions at the times indicated on the histograms, and intracellular calcium concentrations were estimated from the fluo-4-AM:fura red ratio.
Quantitative RT-PCR analysis
RNA was prepared using TRIzol (Life Technologies, Rockville, MD) and RNeasy columns (Qiagen, Valencia, CA) according to the manufacturer’s instructions. Fifty nanograms of RNA was reverse transcribed and amplified in triplicate using the Superscript One-Step RT-PCR kit (Invitrogen, San Diego, CA), according to the manufacturer’s instructions, with the addition of a 1/50 dilution of ROX reference dye (Invitrogen). Exon-spanning primer and probe sets were designed by us or were ordered from Applied Biosystems (Assays On-Demand; Foster City, CA). Primer efficiency was measured by serial dilution of template and was used to calculate the final relative abundance of specific mRNA.
Apoptosis assays
For thymocyte apoptosis assays, freshly isolated thymocytes were incubated with 2 µg/ml biotinylated anti-CD95 Jo-2 mAb with 1 µg/ml streptavidin for 18–24 h, and viability was measured by annexin V and propidium iodide (PI) staining. One microgram per milliliter anti-CD3 and anti-CD28 was used in the TCR-induced thymocyte apoptosis assays. For Ag restimulation experiments, splenic T cells were activated with Ag or 5 µg/ml Con A for 3 days and then were cultured in 100 U/ml IL-2 for at least 48 h before restimulation with graded doses of plate-bound anti-CD3 mAb 2C11 or Ag and irradiated syngeneic APCs (2:1 ratio) for 18–24 h. Cell viability was measured by annexin and PI staining, and specific apoptosis was calculated as previously described (33). Cell death was induced by plate-bound anti-CD3 mAb or a preparation of CD95 ligand (CD95L) stably trimerized by an N-terminal isoleucine zipper motif. Staurosporine was added to a final concentration of 1 µg/ml.
T cell transfer and immunizations and in vivo bromodeoxyuridine (BrdU) labeling
Purified cells were labeled with CFSE and injected i.v. into nonirradiated B6.SJL CD45.1 recipients. A total of 4–6 x 106 cells were adoptively transferred into each recipient. Mice were immunized i.p. 24 h posttransfer with 2.5 mg of OVA protein (Sigma-Aldrich; A-5503) mixed with 150 µg of poly(I:C) (Sigma-Aldrich; P-0913) in 250 µl of PBS. Tail vein blood was sampled before immunization and at different time points postimmunization and was stained for surface markers as indicated. Spleen, lymph node, and liver were also examined for T cell subsets at the end of each experiment. Proliferating cells in unmanipulated mice were labeled by twice daily i.p. injection with 2 mg of BrdU for 7 days. Labeling was detected by surface staining followed by intracellular staining with anti-BrdU-FITC as previously described (34).
Intracellular cytokine staining and tetramer analysis
Cells were stimulated with PMA (10 ng/ml) plus ionomycin (1 µM), OVA (SIINFEKL, 
1–100 nM), or anti-CD3/anti-CD28 (5 µg/ml) for 6 h with monensin (Calbiochem, La Jolla, CA; final concentration 10 µM) added for the final 4 h. Cells were surface labeled and fixed with 3% paraformaldehyde overnight at 4°C. Cells were permeabilized with 100 µl of 1x PBS + 0.1% saponin + 0.1% BSA + 0.01 M HEPES for 10 min at room temperature and were stained with PE-conjugated anti-IL-2, IL-4, or IFN-. Staining with LCMV peptide loaded MHC tetramers, and cytokine production in response to LCMV was performed as previously described (35). For detection of influenza-specific T cells, a PE-labeled class I Db tetramer was loaded with the nucleoprotein (NP) (366–374) ASNENMETM peptide.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Five lines of CD2-vFLIP mice were generated on the B6 mouse background and were tested for transgene expression by Western blotting and FACS staining against the influenza HA epitope at the C-terminal end of the transgenic protein. One line (4817) had stably transmissible transgene expression at the protein level and was characterized further. Western blotting demonstrated expression of a 30-kDa protein consistent with the size of MC159-vFLIP in the thymus and spleen of line 4817 mice (Fig. 1A). Western blotting and intracellular staining revealed specific transgene expression in resting and activated T cells in the spleen and equivalent expression in peripheral CD4 and CD8 T cells (Fig. 1A and data not shown). Cell yields from transgenic thymi were lower than those of controls from mice < 3 mo of age, a difference that disappeared in older mice after the peak of thymopoeisis (Table I). In the thymus, patterns of expression of CD4, CD8, and TCR -chain were normal. However, examination of the CD4–CD8– double negative population revealed a significant decrease in percentages of the most mature CD44–CD25– DN4 cells, with a compensatory rise in less mature populations (Fig. 1B and Table I). Examination of spleen and lymph nodes revealed normal cellularity, with a consistent increase in the CD4:CD8 ratio in T cells, accounted for by a decrease in absolute numbers of CD8 T cells (Table I). The altered CD4:CD8 ratio was not observed in the thymus, suggesting a peripheral CD8+ cell deficiency. Unlike with CD95-deficient lpr mice, there was no expansion of double negative CD4–CD8– 
T cells in the periphery (Fig. 1C). Spleen and lymph node cells from vFLIP transgenic mice up to 2 years old had no expansion of these cells above levels of B6 littermate controls. Unlike with CD95-deficient mice, sera from vFLIP transgenic mice did not contain high-titer antinuclear and anti-dsDNA Abs on a B6 background or a more autoimmune-prone MRL x B6 F1 background at up to 2 years of age (data not shown).
|
|
|
Defective postactivation survival of vFLIP transgenic T cells
In performing the TCR-induced apoptosis assays described above, we observed that activated vFLIP transgenic (TG) T cells were consistently less viable during culture in IL-2 than were control T cell cultures. To quantitate this effect, we labeled cells with the tracking dye CFSE after initial T cell activation with mitogens or CD3 and then quantitated cell division and survival with PI during culture with IL-2. As shown in Fig. 3A, there were consistently fewer highly proliferated viable cells (lower left quadrants) at every time point in transgenic T cell cultures compared with controls, coupled with decreased percentages of viable cells at each timepoint examined. Transgene expression was maintained in surviving cells for up to 11 days after initial ex vivo T cell activation, ruling out loss of transgene expression as an explanation for the surviving cells (Fig. 3B). When CFSE was added to resting T cells before stimulation, overlay plots of activated normal and transgenic CD4+ and CD8+ T cells revealed a maximum of only a one-half cycle delay in average divisions of vFLIP transgenic vs control cells (Fig. 3C). In contrast, quantitation of live cell numbers over time revealed a dramatic reduction in the expansion of T cells in cultures of purified CD4 or CD8 T cells, with 10-fold fewer viable cells present after 1 wk of culture in all experiments performed (Fig. 3D). CD95-deficient lpr/lpr CD4 or CD8 T cells on the same genetic background were not as defective in expansion, suggesting that the vFLIP transgene was not acting solely by blocking CD95 signaling. Similar experiments in which IL-15 rather than IL-2 was used also yielded defective postactivation survival by vFLIP TG T cells (data not shown).
|
Decreased postactivation survival and proliferation can result from ineffective primary stimulation through the TCR or through defective up-regulation of cytokine receptor chains. To assess whether these signals were intact, we measured a number of early activation parameters in vFLIP and control T cells after activation through CD3 and CD28. Up-regulation of the activation markers CD69, CD71 (transferrin receptor), and CD25 (IL-2R -chain), as measured by FACS staining, was indistinguishable from controls at 1, 2, or 3 days postactivation (Fig. 4A and data not shown). Surface IL-7R -chain (CD127) was also normal in resting and activated vFLIP-expressing T cells (data not shown). Quantitative PCR assays showed that CD25, CD122, and IL-15-R mRNA were also normally up-regulated after TCR stimulation in vFLIP transgenic T cells, and the levels of common -chain CD132 mRNA were preserved (Fig. 4B). The suppressor of cytokine signaling1 (SOCS-1) regulatory factor can inhibit IL-2- and IL-15-induced cellular signaling (39). However, we found no differences in SOCS-1 RNA expression between vFLIP transgenic and control T cell blasts (Fig. 4B). SOCS-1 protein levels were comparable in vFLIP transgenic T cells and controls before and after IL-2 stimulation (data not shown). Additionally, FACS analysis showed normal levels of the TNFR family receptors CD30 and CD27 and their ligands in resting and activated vFLIP transgenic CD4+ and CD8+ T cells (data not shown). We also examined other parameters of early T cell activation in vFLIP-expressing T cells. IL-2 production was generally normal at 24 and 48 h after stimulation, whereas proliferation as assayed by [3H]thymidine incorporation from days 2–3 was decreased by 50%, suggesting a defect in IL-2 responsiveness rather than production (Fig. 4C). Addition of IL-2 to cultures during stimulation did not rescue this proliferative defect or the defective postactivation survival in vFLIP-expressing T cells (data not shown). As another measure of the integrity of TCR signaling, we measured calcium fluxes in response to TCR cross-linking (Fig. 4C). Again, responses were similar to controls. Taken together, these results suggest that vFLIP blocks postactivation T cell survival by decreasing cytokine responsiveness in a cell autonomous manner, whereas early TCR-driven activation processes and cell division are relatively unaffected.
|
A specific deficit of functional CD8+ memory T cells in CD2-vFLIP transgenic mice
To determine which T cells in vivo are affected by vFLIP expression, we examined mature T cell subsets from vFLIP transgenic mice and controls. Flow cytometric analysis of peripheral blood, lymph node, and splenic T cells with surface markers for memory cells revealed a consistent defect in CD44highCD122high CD8+ T cell numbers, a subset previously identified as containing functional CD8+ memory T cells (Fig. 5A). CD8+ memory phenotype T cells are known to expand with age. However, at every age examined, the absolute numbers of vFLIP CD8+CD44high T cells per spleen were significantly lower than in age-matched controls (Fig. 5B). This was independent of the CD8+ T cell deficiency seen in vFLIP transgenic mice, in that the percentage of CD44high cells within the CD8+ T cell pool was also significantly reduced at all ages (data not shown). Strikingly, there was no significant CD4+ memory T cell defect measured by CD44, CD122, and/or CD62L. Multiparameter analysis of CD44highCD8+ T cells from vFLIP transgenic mice revealed that CD62Lhigh central memory phenotype cells were most severely affected (Fig. 5C).
|
-deficient mice, in which memory CD8+ T cells are depleted due to decreased self-renewal (32, 40, 41, 42). To determine whether vFLIP transgenic CD8+ memory T cells are similarly impaired in cell turnover, we labeled dividing cells in cohorts of matched transgenic and control mice with BrdU. As has been reported, we found that the CD44highCD122high CD8+ T cell subset was enriched in proliferating cells compared with naive CD8+ T cells in normal B6 mice (Fig. 6A). In vFLIP transgenic mice, although the CD44high subset was depleted, the remaining CD44high cells incorporated BrdU as much as or more than did control cells (Fig. 6A). CD4 memory phenotype T cells from vFLIP transgenic mice incorporated BrdU to a similar extent as did controls (data not shown). To examine the function of residual memory phenotype cells in vFLIP transgenic mice, we stimulated vFLIP and B6 control T cells and measured their ability to rapidly secrete cytokines, a hallmark of memory T cells (Fig. 6B). As expected, only CD44high cells were able to produce IFN- in this time frame. However, the residual vFLIP transgenic CD8+CD44high cells were not capable of producing IFN- in response to CD3/CD28 signals and were also partially impaired in producing IFN- in response to PMA/ionomycin. In contrast, memory phenotype vFLIP-transgenic CD4+ cells produced amounts of IFN- equivalent to those of wild-type memory T cells (Fig. 6B, right panels). The defect in IFN- secretion was not due to skewing to a type-2 cytotoxic T cell phenotype, because no detectable IL-4 was made by transgenic or wild-type CD8+ T cells restimulated in vitro (data not shown). Thus, the memory phenotype CD8+ T cells that are spontaneously produced in vFLIP transgenic mice can self-renew, but are not fully functional.
|
STAT5A and STAT5B are major mediators of signaling by -chain binding cytokines thought to be important for the generation and maintenance of CD8+ memory T cells. To determine whether vFLIP can block STAT function, we performed transactivation experiments with an IFN- activation site-element luciferase vector as a probe for STAT transactivation in cells transiently transfected with vFLIP or vector controls. Additionally, we examined STAT5 phosphorylation in response to IL-2 in vFLIP transgenic and control T cells. In neither case did vFLIP block STAT activation (Fig. 7, A and B). To approach this problem genetically, we crossed vFLIP transgenic mice with STAT5B transgenic mice that exhibit enhanced cytokine-driven STAT activation and dramatically increased numbers of functional CD8+ memory T cells (43). When peripheral T cells from double-transgenic STAT5B x vFLIP mice were analyzed compared with age-matched littermate controls, the STAT5B transgene could only partially rescue memory phenotype CD8+ cells in the periphery. Although STAT5B x vFLIP double transgenic mice had an increased percentage of CD44high cells within the CD8+ T cell population above that of the vFLIP single transgenic mice, double transgenic mice lacked the overall expansion of CD8+ cells seen in STAT5B transgenic mice (Fig. 7C). Absolute numbers of CD44high cells in peripheral blood or spleen in the double transgenic mice were 4-fold lower in STAT5B x vFLIP double transgenics compared with STAT5B transgenic mice, indicating that the deficiency in these cells could not be overcome by enhanced STAT5-dependent cytokine signaling. Similar numbers were found in the spleens of these mice, with 4-fold higher yields of CD8+CD44high cells from STAT5B compared with STAT5B x vFLIP transgenic mice. We also examined the memory phenotype CD8+ T cells in the double transgenic mice for their ability to secrete IFN- and found that like the vFLIP transgenic cells, compared with nontransgenic littermates, STAT5B x vFLIP double transgenic cells were deficient in IFN- production when compared with STAT5B single transgenic littermates (Fig. 7D). Taken together, these data strongly support that vFLIP transgenic memory phenotype CD8+ cells are diminished in number and are functionally impaired via a STAT5-independent mechanism.
|
To study CD8 memory T cell formation and maintenance in response to a defined Ag, we crossed vFLIP transgenic mice with the TCR-TG OT-I transgenic line, in which >90% of T cells are specific for the OVA SIINFEKL peptide/Kb complex (29). Transfer of purified CFSE-labeled CD8 peripheral T cells from these mice into nonirradiated CD45 congenic hosts, followed by immunization with OVA and poly(I:C) adjuvant, resulted in the rapid expansion (20- to 50-fold) of OT-I CD45.2 donor cells with the concomitant loss of CFSE on the transferred cells in both OT-I and OT-I x vFLIP donors (Figs. 8, A and B). At the peak of the response, levels of -IFN production in vitro by OT-I x vFLIP donor cells in response to the antigenic peptide were comparable with those of controls (Fig. 8C). As has been previously observed, after a decline to baseline levels at 1 wk after immunization, a cohort of OT-I cells survived and could be measured in peripheral blood up to 1 mo later. These long-term surviving donor cells have been previously shown to have properties of memory cells. Unlike OT-I cells, vFLIP x OT-I donor T cells declined to undetectable levels in peripheral blood by 14 days after immunization (Fig. 8A). OT-I x vFLIP double transgenic cells could not be detected in spleen, lymph nodes, or liver lymphocyte preparations after day 14 postimmunization, and these cells could not be rescued after reboosting with OVA at >30 days after initial immunization (data not shown). To determine whether the failure to produce memory cells in this system was due to a failure of T cell differentiation or survival, we studied the expression of a number of cell surface markers associated with memory T cells. Levels of CD122, CD25, CD44, CD69, and CD71 on donor OT-I x vFLIP cells 4–7 days after OVA immunization were all indistinguishable from those of OT-I controls (data not shown).
|
Defective CD8+ T cell-mediated responses to infections by vFLIP transgenic mice
To determine whether the defects we observed in vFLIP transgenic mice affect CD8 T cell responses to defined pathogens, we tested responses against two pathogens known to require sustained CD8+ T cell responses for clearance of infection. Infection of B6 mice with the intracellular protozoan T. cruzi causes an acute parasitemia and expansion of CD8 T cells with an inversion of the CD4:CD8 ratio in animal models and human myocarditis (31, 44). Parasite clearance is dependent on CD8+ T cells (45, 46), and vaccines inducing CD8+ T cell immunity are protective (47). Viral FLIP transgenic and control mice were infected with T. cruzi, and parasitemia was measured at regular intervals. As shown in Fig. 9A, in contrast with B6 mice, which are normally resistant to T. cruzi, B6.vFLIP transgenic mice failed to control infection, with increased parasite counts in the blood during most of the infection. Mortality in the vFLIP group was 40% in two independent experiments, whereas all of the control mice survived. Analysis of splenic T cell subsets after 28 days of infection revealed that vFLIP transgenic mice did not exhibit the CD8+ T cell expansion or inversion of the CD4:CD8 ratio that normally accompany a successful T cell response to T. cruzi.
|
10-fold reduction in cell numbers over the next week, virus-specific cells persist for many months after infection at relatively stable levels, becoming functional memory cells (48). When vFLIP transgenic mice were infected with the LCMV Armstrong strain, primary expansion still occurred, but numbers of CD8 T cells responding to the five major LCMV epitopes were 4- to 10-fold lower at day 8 after infection than were controls (Fig. 9B). Kinetic analysis of the frequencies of T cells reactive to the immunodominant Db/GP33 epitope by tetramer staining revealed a similar deficit at day 8 in the vFLIP transgenic mice. By day 50, no GP33-reactive T cells could be detected in the vFLIP mice, whereas control mice had easily detectable numbers of these cells (Fig. 9B). Examination of T cells from spleen, lymph nodes, liver, and lung revealed similar trends, showing that the absence of LCMV-reactive cells was not the result of abnormal migration or redistribution. At day 9 postinfection, two-thirds of the vFLIP mice tested still had detectable serum LCMV titers (>200 PFU/ml), whereas none of the control mice had detectable viremia at this time point. This suggests that the blunted anti-LCMV CD8 T cell response in vFLIP transgenic mice was unable to control viral infection as quickly as was the immune response in control mice. The kinetics of anti-LCMV responses are similar to the responses to OVA shown in Fig. 8, with the exception that peak responses were somewhat lower. Decreased cell survival by day 8, when responses were first measured in the LCMV infections, rather than day 4 in the OT-I transfer experiments may account for this difference. To study the ability of Ag-specific CD8+ T cells to mount a recall response, we infected vFLIP transgenic and control mice with the influenza PR8 strain (H1N1), for which the immunodominant NP(366–374) epitope has been determined on the B6 background (58). As shown in Fig. 9C, NP(366–374)-specific T cells detected with MHC peptide tetramers from both vFLIP and control mice expanded to 12–16% of the total CD8 peripheral blood pool by day 8 after infection. Mice were rechallenged on day 28 with the H3N2 X.31 strain, which shares the same NP(366–374) sequence but cannot be neutralized by Abs generated against PR8, allowing a productive reinfection. In control mice, NP(366–374)-specific CD8+ T re-expanded with the expected kinetics of memory cells, reaching peak frequencies similar to the primary response that was sustained for >50 days after rechallenge. Strikingly, tetramer-positive CD8 cells from vFLIP transgenic mice were barely detectable after secondary X.31 infection, with a peak of 4% frequency in the CD8 pool 1 wk after infection. Subsequently, these cells disappeared below the limit of detection. Similar results were seen with the subdominant epitope PA(224-33) (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our results show that expression of vFLIP, whose only known function is to block recruitment of caspase-8 into death receptor complexes, results in a phenotype similar to that of caspase-8 rather than FADD deficiency, whereas at the same time potently blocking CD95-induced apoptosis. After in vitro acitvation, vFLIP transgenic T cells exhibited near-normal cell cycling, but increased postactivation cell death and decreased cell yields (Fig. 3). The decreased thymidine incorporation (Fig. 4C) likely results from the increased cell death rather than decreased cycling. Our findings are similar in some respects to the in vitro responses of T cells from transgenic mice overexpressing cellular FLIP (c-FLIP), in which decreased proliferation to strong TCR stimuli is seen, although the relative contributions of cell cycling vs survival were not examined in those mice (51, 52). We do not see the increased responsiveness to low-dose TCR stimulation seen by Lens et al. (52) in one line of c-FLIP transgenic mice, perhaps because vFLIP is not known to activate the NF-B or extracellular signal-regulated kinase pathways, as has been shown for c-FLIP.
In vivo, vFLIP transgenic mice exhibited a mild CD8 T cell deficiency but a profound derangement of memory CD8+ T cell generation and function. The developmental defect in vFLIP transgenic mice was less severe than in FADD-deficient or FADD-DN transgenics. Expression of the M. contagiosum MC159 vFLIP used in this study resulted in a milder drop in total thymocyte yield compared with transgenic mice overexpressing the equine herpesvirus E8 vFLIP in the thymus (27) (2-fold vs 5-fold in young animals). Fewer studies have examined the effects of death receptor blockade on in vivo immune responses, but our results appear to be distinct from those obtained with FADD-DN transgenic mice, in which naive CD8 cells are significantly depleted and expansion of LCMV-specific CD8+ T cells after infection was essentially absent. Mice with T cell-specific caspase-8 deficiency had lower LCMV-specific CTL responses 8 days after infection, but kinetic studies of the frequency of LCMV-specific T cells were not performed.
The mechanism by which vFLIP impairs memory CD8+ T cell generation does not appear to involve defects in primary T cell activation; rather, vFLIP appears to block postactivation survival signals. CD8+ memory T cells are known to require cytokine signals delivered through the common -chain (c) family of cytokine receptors, particularly IL-7 and IL-15 for their generation and maintenance (53). Thus, direct or indirect blockade of c signaling would be an attractive mechanism for how vFLIP impairs T cell survival and CD8+ memory cell generation. Our results showing normal STAT transactivation in the presence of vFLIP, normal STAT5 phosphorylation after cytokine challenge of vFLIP-expressing T cells, and the lack of rescue of memory cells by STAT5B overexpression indicate that any c signaling defect in vFLIP transgenic mice is downstream of or parallel to STAT5 activation.
Like FADD-DN transgenic and caspase-8-deficient T cells, vFLIP transgenic T cells are resistant to CD95-mediated apoptosis, although they do not develop any of the hallmarks of CD95 deficiency. We suggest that this is because the defective proliferation and survival in vFLIP transgenic T cells may be dominant over the impaired apoptosis seen in restimulation assays, not allowing abnormal help to be given to autoreactive B cells. Although CD95 has been implicated in T cell costimulation (54), it is unlikely that CD95 itself triggers the survival pathway blocked by vFLIP. As shown in Fig. 3, CD95-deficient T cells did not exhibit decreased postactivation survival and CD95-deficient mice have no defect in CD8 memory cell generation. Other members of the death receptor family may be good candidates for such a receptor. It is also possible that vFLIP is blocking signaling by other receptor families that share common adapter proteins with death receptors, such as the Toll-like receptor/IL-1 family.
These results are the first description of specific T cell memory defects induced by blocking components of death receptor signaling and have implications for understanding the signals that govern the generation and maintenance of memory T cells. Memory cells may arise as a separate lineage of long-lived cells during T cell expansion or through progressive differentiation steps from effector T cells (48). Our results suggest that vFLIP selectively impairs the generation of functional long-lived memory CD8+ T cells, but does not impair turnover of residual memory phenotype cells, and only slightly shortens the half-life of naive CD8+ T cells. In the influenza system, the contrast between near-normal primary effector T cell expansion and defective T cell memory responses is most apparent (Fig. 9C). It has been suggested that CD4+ T cell help is necessary for the generation of CD8 T cell memory (55, 56). In vFLIP transgenic mice, normal numbers and cytokine secretion by CD4 memory phenotype cells argues against a defect in CD4+ T cell help. Additionally, in the OT-I transfer experiments, long-term survival of CD8+ OT-I T cells was compromised despite normal nontransgenic CD4+ T cells in the recipients. Understanding how vFLIP impairs the production of memory CD8 T cells would allow new insights into the pathways governing memory cell generation. Because CD8 memory cell responses are believed to be critical for effective immunity to many intracellular pathogens, enhancing the signaling pathways blocked by vFLIP may be a novel strategy for boosting the effectiveness of vaccines against viral pathogens. Interfering with CD8 memory cell generation by exogenous delivery of vFLIP may be a novel means to elicit cell tolerance in CD8+ T cell-mediated autoimmunity, graft rejection, or graft-vs-host disease.
Our results also suggest that rather than just being a viral blocker of apoptosis, vFLIP may be conserved in and expressed by viruses to disrupt host immune cell homeostasis. The immune deficiency induced by vFLIP expression significantly impaired CD8 T cell responses to LCMV and the pathogenic protozoan T. cruzi, increasing susceptibility to these infectious agents. Although the Molluscum contagiousum virus containing the MC159 vFLIP gene does not infect T cells, a number of FLIP-encoding -herpesviruses do (21). The one report of in vivo infection with a vFLIP mutant virus did not study differences in the antiviral immune responses in the animals infected with vFLIP-deleted vs wild-type virus (22). If antiviral T cells become preferentially infected with their target virus, as has been shown for HIV (57), viruses containing vFLIP may function through the mechanism described here to specifically disable antiviral CD8 T cell responses.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Z.W. and M.R. contributed equally to this work.
3 Current address: Department of Microbiology & Immunobiology, Queen’s University Belfast, BT9 7BL, Ireland.
4 Address correspondence and reprint requests to Dr. Richard M. Siegel, National Institutes of Health, Building 10, Room 9N238, Bethesda, MD 20892. E-mail address: siegelr{at}mail.nih.gov
5 Abbreviations used in this paper: FADD, Fas-associated death domain protein; FADD-DN, FADD death-domain only; vFLIP, viral FLIP; c-FLIP, cellular FLIP; HA, hemagglutinin; LCMV, lymphocytic choriomeningitis virus; fluo-4-AM, fluo-4-acetoxymethyl ester; PI, propidium iodide; CD95L, CD95 ligand; BrdU, bromodeoxyuridine; NP, nucleoprotein; TG, transgenic; SOCS-1, suppressor of cytokine signaling 1; c, common -chain; GFP, green fluorescent protein.
Received for publication January 5, 2004. Accepted for publication March 11, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1) death signal. J. Biol. Chem. 273:4345.B and Erk signaling pathways. Curr. Biol. 10:640.[Medline]
), B6.vFLIP transgenic (
), and B6.lpr/lpr (*) mice stimulated for 3 days with CD3/CD28 and cultured with 10 U/ml IL-2 from day 3 onward. E, Mixing experiment in which purified CD8+ (top) and CD4+ (bottom) T cells from control B6 (CD45.2) (
) or DMSO control (
