Abstract
CD4+CD28null T cells are oligoclonal lymphocytes rarely found in healthy individuals younger than 40 yr, but are found in high frequencies in elderly individuals and in patients with chronic inflammatory diseases. Contrary to paradigm, they are functionally active and persist over many years. Such clonogenic potential and longevity suggest altered responses to apoptosis-inducing signals. In this study, we show that CD4+CD28null T cells are protected from undergoing activation-induced cell death. Whereas CD28+ T cells underwent Fas-mediated apoptosis upon cross-linking of CD3, CD28null T cells were highly resistant. CD28null T cells were found to progress through the cell cycle, and cells at all stages of the cell cycle were resistant to apoptosis, unlike their CD28+ counterparts. Neither the activation-induced up-regulation of the IL-2R α-chain (CD25) nor the addition of exogenous IL-2 renders them susceptible to Fas-mediated apoptosis. These properties of CD28null T cells were related to high levels of Fas-associated death domain-like IL-1-converting enzyme-like inhibitory protein, an inhibitor of Fas signaling that is normally degraded in T cells following activation in the presence of IL-2. Consistent with previous data showing protection of CD28null cells from spontaneous cell death, the present studies unequivocally show dysregulation of apoptotic pathways in CD4+CD28null T cells that favor their clonal outgrowth and maintenance in vivo.
The aging immune system and chronic inflammatory syndromes such as rheumatoid arthritis and acute coronary artery disease are characterized by high frequencies of CD4+ T cells that are deficient in CD28 expression (1, 2, 3). Compared with their CD28+ counterparts, they produce significantly higher levels of IFN-γ (3, 4), giving them the ability to function as proinflammatory cells. Moreover, CD4+CD28null T cell clones persist for years in circulation (5). Longevity of these cells appears to be related to their relative resistance to spontaneous cell death even in the absence of IL-2 (6). This phenomenon is associated with low levels of expression of the α-chain of the IL-2R (IL-2Rα), despite their ability to produce large amounts of IL-2, and an increased expression of the anti-apoptotic molecule Bcl-2 (4, 6). Inasmuch as CD4+CD28null T cells are highly oligoclonal (7), we examined their susceptibility to activation-induced cell death (AICD).5 In the normal immune system, AICD is a mechanism to delete activated T cells upon resolution of Ag-driven responses (8). Although the antigenic basis of T cell oligoclonality during aging and in chronic inflammatory diseases is not known, persistence of CD4+CD28null T cell clones in vivo suggests perturbation of apoptotic pathways.
AICD is one of the best-characterized systems of apoptosis. Subsequent to activation, T cells up-regulate Fas ligand (FasL), which then interacts with the Fas receptor on the same or on neighboring T cells. Fas-FasL interaction generates an apoptotic signal via the phosphorylation of the Fas receptor death domain, resulting in a cascade of activation events of caspases that ultimately lead to cell death (9). Molecules that are collectively referred to as inhibitors of apoptosis may, however, prevent Fas-mediated cell death. One of these is the Fas-associated death domain-like IL-1-converting enzyme inhibitory protein (FLIP), also known as Casper, CLARP, FLAME-1, or MRIT (10, 11, 12, 13, 14). FLIP inhibits apoptosis by directly interacting with Fas-associated death domain, or with caspases 8 and 10, resulting in the interruption of signal transduction from the Fas receptor. This regulator of apoptosis can also interrupt signaling of other death receptors, particularly members of the tumor necrosis receptor family (11, 15).
Although IL-2 is a T cell growth factor, it can also potentiate Fas-mediated AICD (16). Subsequent to Ag recognition by T cells, IL-2 production and IL-2Rα expression result in the progression of cells through the cell cycle as well as the up-regulation of FasL and a concomitant suppression of FLIP expression (17, 18). Consequently, activated T cells become susceptible to Fas-induced cell death as they proceed through the cell cycle (19, 20). Induction of FasL and the down-regulation of FLIP expression are IL-2-dependent processes (16, 17, 18). Because CD4+CD28null T cells unstably express IL-2Rα (4, 6), we examined whether there is differential susceptibility between CD4+CD28+ and CD4+CD28null T cells. Studies described in this work examined the interrelationship, if any, between the levels of IL-2Rα expression, cell cycle progression, and Fas-induced cell death. Inasmuch as CD4+CD28null T cells represent a highly oligoclonal subset of lymphocytes found during aging and in chronic disease states (1, 3, 7, 21, 22, 23), these studies permitted the evaluation of the hypothesis that T cell oligoclonality may be the result of persistent immune activation and the prevention of AICD. Alteration of cell death programs could account for the accumulation of CD4+CD28null T cells in vivo.
Materials and Methods
Cell culture
Short-term human T cell lines and CD4+CD28null T cell clones were established from fresh PBMC, as described previously (1, 6). Short-term cell lines were established from PBMC that were initially stimulated with immobilized anti-CD3 (OKT3; American Type Culture Collection, Manassas, VA) for 12 h and maintained at densities of 0.5–2 × 106 cells/ml. Cells were passaged every 5–7 days in RPMI 1640 (BioWhittaker, Walkersville, MD) containing 10% FCS (Summit Biotechnology, Fort Collins, CO), 2 mM l-glutamine, 50 U/ml penicillin, 5 μg/ml streptomycin (Life Technologies, Grand Island, NY), and 10 U/ml human rIL-2 (Proleukin; Chiron, Emeryville, CA). Feeder cells consisting of γ-irradiated neuraminidase-treated EBV-transformed B lymphoblastoid cells were also added to the cultures. Cells were maintained in a humidified 7.5% CO2 tissue culture incubator.
T cell clones were established by limited dilution cloning of peripheral blood CD4+ T cells (7). Clonality was determined by standard seminested PCR for TCR BV-BJ elements, size fractionation, and sequencing. T cell clones were maintained on feeder cells of γ-irradiated, neuraminidase-treated EBV-transformed B lymphoblastoid cells in the presence of 20 U/ml IL-2.
The T cell lines and clones used in the present study were derived from patients with rheumatoid arthritis as well as from healthy donors. The phenotypic characteristics of CD4+CD28null T cells, such as the lack of CD40 ligand, elevated expression of IFN-γ, oligoclonality, etc. (1, 2, 3, 4, 5, 7), were generally very similar among the cells examined regardless of the source donor.
Jurkat cells (American Type Culture Collection) were maintained in RPMI 1640 medium (as indicated above) in the absence of IL-2. They were maintained at a density of 5 × 106 cells/ml in a humidified 5% CO2 tissue culture incubator.
Flow cytometry
Cell surface staining of T cells was performed using mAb to CD4, CD28, IL-2Rα (CD25), and IL-2Rβ (CD75) conjugated to the appropriate fluorochrome (Becton Dickinson, San Jose, CA). Briefly, 2 × 105 to 1 × 106 cells were incubated with mAb or Ig isotype control (Simultest; Becton Dickinson) for 25 min on ice, washed with cold PBS, and fixed with 1% paraformaldehyde for 60 min at 4°C.
For immunostaining of the Fas receptor, cells were incubated with unconjugated anti-CD95 (CH11; Beckman Coulter, Miami, FL), followed by FITC-conjugated anti-mouse Ig (BD-PharMingen, San Diego, CA). As controls, cells were also incubated in IgG instead of anti-CD95. Cells were subsequently stained with PE-conjugated anti-CD28 and peridinin chlorophyl protein (PerCP)-conjugated anti-CD4 (Becton Dickinson), washed, and fixed with 1% paraformaldehyde. Live cells were gated by forward and side scatter. Cells with reduced forward scatter were excluded, and Fas expression was determined as FITC fluorescence on either CD4+CD28+ or CD4+CD28null cells.
FasL expression was determined by intracellular immunostaining. Cells were activated with the appropriate Ab or pharmacologic agent in the presence of 10 μg/ml brefeldin A (Epicentre Technologies, Madison, WI). They were stained with PerCP-conjugated anti-CD4 and PE-conjugated anti-CD28 mAb, fixed with paraformaldehyde, and subsequently permeabilized with 0.05% Tween 20 (Sigma, St. Louis, MO) for 15 min at 37°C. Cells were washed, resuspended in biotin-conjugated anti-FasL mAb (NOK-2; BD-PharMingen) for 25 min on ice, followed by FITC-conjugated streptavidin. As control, cells were stained in a similar manner with anti-Bcl-2 mAb (Dako, Carpenteria, CA).
Flow cytometry was performed on either a FACSCalibur or FACSVantage flow cytometer (Becton Dickinson). Cell populations were analyzed using the WinMDI program (Joseph Trotter, The Scripps Research Institute, La Jolla, CA).
Apoptosis and cytotoxicity assays
About 1 × 106 T cells were cultured in 96-well plates coated with either anti-CD3 mAb (OKT3), anti-CD95/Fas mAb (CH11), or an IgG isotype at 50 μg/ml. After 18 h, cells were harvested and immunostained for CD4 and CD28, fixed in paraformaldehyde, and permeabilized with 0.05% Tween 20. Subsequently, 2.5 μg/ml 7-amino-actinomycin D (7-AAD; Calbiochem, San Diego, CA) was added and incubated for 30 min at room temperature. The proportion of 7-AAD+ subdiploid cells was determined by flow cytometry. IL-2R and FasL expression of parallel cultures were also examined. Cells were immunostained for IL-2Rα and IL-2Rβ, as described above. Apoptosis assays were conducted using T cell lines and clones 7–10 days after the last passage and stimulation with EBV-transformed B cell feeders.
The biological activity of soluble FasL was examined. CD28+CD28null and CD28+CD28+ T cells were activated with either plate-immobilized anti-CD3 or a cocktail of 1 μg/ml PMA and 10 nM ionomycin (Sigma) for 24 h. Culture supernatants were harvested and added to 1 × 106 Jurkat cells. Cells were cultured for 18 h in the presence or absence of 5 μg/ml of the anti-FasL mAb NOK-2. Viability of Jurkat cells was determined by trypan blue exclusion.
Immunoblotting
Western blots were performed as described previously (6). A total of 2 × 106 T cells was lysed in a hypotonic buffer and centrifuged at 12,000 × g for 10 min at 4°C, and protein concentrations were determined using the Bio-Rad protein assay reagent (Bio-Rad, Hercules, CA). Detergent-solubilized protein at 10 μg/lane was separated on 12% SDS-polyacrylamide gels using a mini-gel system and transferred onto Sequi-Blot polyvinylidene difluoride membranes (Bio-Rad). Membranes were incubated with 4% BSA in TBS, followed by a 1/500 dilution of a mAb to the short form of FLIP (F-20), and subsequently incubated in a 1/1000 dilution of HRP-conjugated anti-mouse Ig (Santa Cruz Biotechnology, Santa Cruz, CA). Blots were developed using ECL chemiluminescence reagents (Amersham Pharmacia Biotech, Piscataway, NJ) and exposed to radiographic films (BIOMAX-MS; Kodak, New Haven, CT). Equal protein loading was ascertained by staining the membrane with GELCODE Blue reagent (Pierce, Rockford, IL).
Cell cycle analysis
T cells were synchronized by incubation with aphidicolin (Sigma); 5 μM aphidicolin was empirically determined to synchronize ≥98% of both CD4+CD28null and CD4+CD28+ T cells at the G1-S boundary within 30 h of incubation (data not shown). Synchronization was verified by 5-bromo-3′-deoxyuridine labeling using an in situ proliferation kit (BrDU-FLUOS; Roche Molecular Biochemicals-Boehringer Mannheim, Indianapolis, IN) combined with standard propidium iodide staining for DNA content and flow cytometry.
Synchronized cells were washed extensively and cultured on plates coated with 50 μg/ml IgG, anti-CD3, or anti-Fas Ab for 18 h. As indicated, activation was conducted either in the presence or absence of 5 μg/ml exogenous IL-2. Cells were subsequently immunostained for CD4, CD28, IL-2Rα, and IL-2Rβ, as described above, and with propidium iodide staining for DNA content.
Statistical analysis
Quantitative analysis was conducted with Student’s t test, and, if appropriate, with the Wilcoxon signed rank test using the SigmaStat software (SPSS, Chicago, IL).
Results
CD4+CD28null T cells are protected from undergoing AICD
Differential sensitivity of CD4+CD28+ and CD4+CD28null T cells to AICD was examined by incubation on plate-immobilized anti-CD3. Levels of apoptosis were determined by flow cytometric analysis of 7-AAD-stained cells. Apoptotic cells were identified as sub-G0 cells characterized by reduced DNA-activated fluorescence, which was indicative of fractional or subdiploid DNA content. Depicted in Fig. 1⇓A is a comparison of DNA staining between representative CD28+ and CD28null T cell clones before and after incubation with anti-CD3. These results show that a higher proportion of CD28+ cells compared with CD28null cells was sub-G0 even in absence of activation. Among the CD28+ cells, the percentage of subdiploid cells markedly increased after incubation with anti-CD3. This was unlike the situation in CD28null cells, which showed little or no detectable increase in the number of subdiploid cells in the presence of anti-CD3. Curiously, anti-CD3 elicited a noticeable decrease in the proportion of CD28null cells in G0-G1 and an increase in the proportion of cells in S-G2-M. In contrast, there was little difference in the proportion of CD28+ cells in G0-G1 and S-G2-M between anti-CD3-treated cells and controls.
CD4+CD28null T cells are resistant to AICD. Short-term T cell lines and T cell clones that were either CD28+ or CD28null were cultured for 24 h on plastic-immobilized anti-CD3 (aCD3) or IgG. Cells were washed and immunostained for CD4 and CD28, followed by 7-AAD, and analyzed for DNA staining by flow cytometry. Experiments were conducted on cells 7–10 days after the last stimulation. A, Histograms are representative for CD4+CD28+ and CD4+CD28null T cells, as indicated. Apoptotic cells were recognized by their reduced DNA-associated fluorescence with sub-G0 or subdiploid DNA content in the region marked M1 (indicated as percentage of total cells). Regions M2 and M3 represent the percentage of cells in S-G2-M and G0-G1, respectively. B, Results from six CD4+ T cell lines containing a mixture of CD28+ and CD28null cells and from five T cell clones for each of the CD28 phenotypes are summarized as box plots depicting the median (bar), the 10th and 90th percentiles (whiskers), and the 25th and 75th percentiles (box).
Depicted in Fig. 1⇑B are results from all T cell lines and clones that we examined. These show that there was a significantly higher proportion of CD28+ cells that underwent apoptosis compared with CD28null cells (p < 0.001). Corroborating previous data (6), the present data demonstrate the levels of spontaneous apoptosis, i.e., cells incubated in IgG, were also higher for the CD28+ cells (p < 0.01). Such difference in the susceptibility to AICD between CD4+CD28+ and CD4+CD28null was consistently observed in all cell lines and clones examined.
Induction of AICD in CD28+, but not CD28null, T cells was most likely Fas mediated. As shown in Fig. 2⇓, incubation of cells on either immobilized anti-CD3 or anti-Fas receptor mAb resulted in a significant increase of subdiploid CD28+ cells compared with those incubated with IgG isotype control (p < 0.001). There were no significant differences in the levels of apoptosis of CD28+ cells incubated on either anti-CD3 or anti-Fas Ab. Incubation of cells in a mixture of both Ab also showed similar high levels of apoptosis. In contrast, the CD28null cells did not show any appreciable increase of subdiploid cells following incubation with anti-CD3, anti-Fas, or both Ab over those cells incubated with IgG. Compared with the CD28+ cells, they also maintained significantly lower levels of spontaneous apoptosis (p < 0.01) as in Fig. 1⇑. The resistance of CD28null cells to Fas-induced apoptosis was observed in all cell lines and clones examined.
Cross-linking of CD3 and Fas does not induce apoptosis of CD4+CD28null T cells. CD4+ T cell lines and clones were incubated for 24 h on plastic-immobilized IgG, anti-CD3 (aCD3), and/or anti-Fas (aFas). The number of subdiploid cells was examined by flow cytometry, as in Fig. 1⇑. Data shown are representative of four CD4+ T cell lines containing a mixture of CD28+ and CD28null cells and five clones of each of the CD28 phenotypes. Features of box plots are as in Fig. 1⇑.
CD4+CD28+ and CD4+CD28null T cells have equivalent levels of Fas and soluble FasL expression
The resistance of CD4+CD28null T cells to Fas-induced apoptosis might be argued to be due to a lack of the Fas receptor. Various cell lines and clones were, therefore, examined for Fas expression. As shown in Table I⇓, CD28+ and CD28null T cells have equivalent levels of Fas expression. In fact, the densities of Fas on the cell surface are nearly identical between the two cell types (data not shown). Fas-expressing cells were distinguished by FITC fluorescence shifts of one-half to two log intensities over the isotype control. At these levels of expression, the CD4+CD28null, but not CD4+CD28+, T cells were consistently unresponsive to Ab cross-linking of Fas (Fig. 2⇑).
Fas expression on CD4+CD28+ and CD4+CD28null T cells
The level of FasL expression was also examined. Results of flow-cytometric studies, however, showed low and widely variable expression of either cytoplasmic or membrane FasL (data not shown). Inasmuch as the reason for this variability is not known, we examined the ability of either cell type itself to induce Fas-mediated cell death, presumably through the production of biologically active soluble FasL. CD28+ and CD28null T cells were activated either by incubation with anti-CD3 or with mitogenic concentrations of PMA and ionomycin. The culture supernatants were harvested and assayed for cytotoxic activity on Jurkat cells. As shown in Fig. 3⇓, there was no significant spontaneous death of Jurkat cells as indicated by the high survival rate of the cells in the medium controls. Addition of supernatants from either activated CD28+ or CD28null cells was cytotoxic to Jurkat cells. The levels of cytotoxicity were equivalent for all of the supernatants tested. The addition of anti-FasL mAb resulted in the neutralization of the cytotoxic activity of the supernatants. The levels of biological activity of soluble FasL-containing supernatants from either CD28+ or CD28null T cells were similar for anti-CD3-activated or PMA/ionomycin-treated cells.
CD4+CD28+ and CD4+CD28null T cells produce equivalent amounts of cytotoxic soluble FasL. T cell clones were incubated with either anti-CD3 (circles) or a combination of PMA and ionomycin (triangles) for 18 h. Culture supernatants were collected and added to triplicate cultures of Jurkat cells. Survival of Jurkat cells was monitored after 24 h. Cytotoxicity of supernatants due to soluble FasL was examined by the addition of anti-FasL Ab during the bioassay (gray symbols). Data shown are representative of three clones for each of the CD28 phenotypes.
Resistance of CD4+CD28null T cells to AICD is associated with the induction of IL-2Rα expression
In addition to IL-2 being the major growth factor for both naive and activated T cells (24, 25), it can also potentiate AICD (16, 17). Because in vivo expanded CD4+CD28+ and CD4+CD28null T cell clones generally have an activated phenotype (26), we examined whether their differential sensitivity to AICD could be related to the level of IL-2Rα expression. As depicted in Fig. 4⇓, the constitutive levels of IL-2Rα expression on CD4+CD28+ T cells were significantly higher than on CD4+CD28null cells (p < 0.001), as we reported previously (4). Moreover, such high levels of expression on CD28+ cells were also found to further increase upon incubation with anti-CD3 (p < 0.01). Such induction of IL-2Rα with anti-CD3 was accompanied by a significant increase in the number of subdiploid cells over the IgG controls (p < 0.001). Neither the combination of anti-Fas and anti-CD3 nor anti-Fas by itself significantly changed the observed levels of apoptosis and IL-2Rα expression in CD28+ cells.
Resistance of CD4+CD28null T cells to Fas-mediated apoptosis is independent of IL-2Rα expression. Duplicate cultures of T cells were incubated overnight on plastic-immobilized IgG, anti-CD3 (aCD3), or anti-Fas (aFas). Cells were washed and immunostained for CD4 and CD28, followed by 7-AAD staining, and analyzed by flow cytometry. Parallel cultures were analyzed for IL-2Rα expression. Data shown were based on the analysis of four short-term CD4+ T cell lines containing both CD28+ and CD28null subsets and six sublines sorted for CD28. Features of box plots are as in Fig. 1⇑.
Incubation of CD4+CD28null T cells with anti-CD3 also resulted in significant up-regulation of IL-2Rα expression compared with unstimulated cells (p < 0.001). In fact, the levels of IL-2Rα induction were equivalent to those seen on their CD28+ counterparts. However, there was no perceptible change in the number of subdiploid CD28null cells despite such increases in IL-2Rα expression during activation with anti-CD3. The addition of anti-Fas to the CD28null T cell cultures did not lead to further increases in IL-2Rα expression nor did it increase the level of apoptosis. Moreover, anti-Fas by itself did not induce IL-2Rα expression and had no perceptible effect on apoptosis over that of the unstimulated controls.
There were equivalently high levels of IL-2Rβ expression on both cell types (data not shown). In contrast to the results with IL-2Rα, incubation of cells with anti-CD3 did not affect the levels of IL-2Rβ expression.
Resistance of CD4+CD28null T cells to AICD is independent of cell cycle progression
The resistance of CD28null T cells to apoptosis despite induction of IL-2Rα following activation raised the question whether this was due to their inability to progress through the cell cycle. In normal T cells, IL-2/IL-2R signaling commits cells to enter the cell cycle (24, 25). Subsequently, they become susceptible to Fas-mediated apoptosis (18), which peaks during the S phase of the cell cycle (20). Thus, we examined the relative proportions of T cells at the different stages of the cell cycle following activation. CD28+ and CD28null T cells were synchronized with aphidicolin and incubated in anti-CD3 with or without anti-Fas, and the numbers of cells in G0-G1 and S-G2-M were determined. As shown in Figs. 1⇑A and 5, a significant proportion of CD4+CD28null T cells was in S-G2-M following incubation with anti-CD3 compared with those incubated with IgG (p < 0.03). Furthermore, the anti-CD3-induced response was significantly higher than that seen with their CD28+ counterparts (p < 0.001).
As indicated in Fig. 4⇑, even under conditions of increased IL-2Rα expression due to activation by anti-CD3, there was no perceptible increase in apoptosis among CD28null cells, unlike CD28+ cells, which underwent high levels of Fas-mediated AICD. The addition of anti-Fas to the cultures did not elicit any significant change in relative proportion of cells in S-G2-M over those in G0-G1 phase of the cell cycle (Fig. 5⇓). Also, anti-Fas did not affect the overall levels of anti-CD3-induced apoptosis in either CD4+CD28+ or CD4+CD28null T cells (Fig. 4⇑).
Cross-linking of Fas does not interfere with cell cycle progression of CD4+CD28null T cells. Aphidicolin-synchronized cells were incubated overnight with plastic-immobilized IgG or anti-CD3 (aCD3) in the presence or absence of anti-Fas (aFas). All cultures contained exogenous IL-2. Cells were washed and immunostained for CD4 and CD28, followed by propidium iodide staining for DNA content. Cell cycle analysis was examined by flow cytometry. Data shown are based on the analysis of 10 T cell lines and clones as in Fig. 4⇑. Features of box plots are as in Fig. 1⇑.
High levels of FLIP expression are maintained in CD4+CD28null T cells following activation
The similar levels of cell surface expression of Fas as well as the production of biologically active FasL by CD28+ and CD28null T cells suggested that the resistance of the latter to Fas-mediated death may be a perturbation in Fas signaling. The role of the apoptosis inhibitor FLIP was evaluated because it has been shown to inhibit proximal signals emanating from Fas-FasL interaction, and its down-regulation subsequent to activation has been associated with AICD (17, 18). Indeed, Western blotting experiments revealed that incubation of CD4+CD28+ T cell clones with anti-CD3 resulted in the loss of FLIP expression (Fig. 6⇓). In contrast, CD4+CD28null T cell clones maintained high levels of FLIP expression in the presence of anti-CD3. Neither the presence nor absence of exogenous IL-2 in CD28null T cell cultures affected the levels of FLIP expression (data not shown). In all CD28null clones examined, there was a negligible difference between activated and unstimulated cells.
CD4+CD28null T cells maintain high levels of FLIP expression following activation. T cells were incubated on plastic-immobilized IgG (−) or anti-CD3 (+) for 24 h. Total cell lysates were prepared and subjected to SDS-PAGE and Western blotting for FLIP. Addition of exogenous IL-2 did not alter FLIP expression of CD28null cells (data not shown). Immunoblots shown are representative of eight clones for each of the CD28 phenotypes examined.
Discussion
CD4+CD28null T cells are characterized by oligoclonality and longevity (26, 27, 28). Although their high frequencies of up to 50% of the total CD4 compartment (1, 2, 3) suggest chronic activation by persistent, but yet undefined Ag(s), it may be argued that this phenomenon may also be explained by perturbation of apoptotic pathways. Indeed, the present data demonstrate their resistance to AICD. Incubation of CD4+CD28null T cells with anti-CD3, a condition that renders normal T cells susceptible to Fas-mediated cell death (29), did not result in apoptosis (Figs. 1⇑ and 2⇑). Furthermore, neither the cross-linking of Fas by itself nor the cocross-linking of CD3 and Fas promoted apoptosis in these cells. This is unlike the situation in CD4+CD28+ T cells, which are highly susceptible to apoptosis following incubation with Ab to CD3 or Fas. These data support the notion that CD4+CD28null T cells have a survival advantage over their CD28+ counterparts (27). Aberrations in AICD could confer such advantage and contribute to the phenomenal restriction of the T cell repertoire found during aging and in chronic inflammatory diseases (28, 29, 30, 31, 32, 33, 34, 35).
The insensitivity of the CD4+CD28null T cells to Fas-mediated cell death is not due to defective FasL expression. CD28null cells produce soluble FasL that is as functionally active as that secreted by CD28+ cells (Fig. 3⇑). Because CD28null and CD28+ cells express equivalent levels of Fas (Table I⇑), the resistance of CD28null, but not CD28+, cells to AICD suggests a perturbation of Fas signaling. Additionally, CD4+CD28null T cells did not display the proapoptotic effects of IL-2, as have been reported for normal activated T cells (16, 17). Cross-linking of CD3 on CD28null cells induces IL-2Rα expression with concomitant cell cycle progression without perceptible increase in Fas-mediated apoptosis (Figs. 4⇑ and 5⇑).
The notion of defective Fas signaling in CD4+CD28null T cells is supported by the finding that these cells maintain high levels of the anti-apoptotic molecule FLIP subsequent to activation (Fig. 6⇑). Normal resting T cells express large amounts of FLIP, which binds to either the death domain of Fas or to one of the effector caspases, thereby effectively interrupting transduction of death signals (36). Activating signals through the TCR-CD3 complex, however, result in the degradation of FLIP (17). Thus, activated T cells become sensitive to Fas-mediated apoptosis, as seen with CD4+CD28+ T cells, but not with CD4+CD28null T cells (Figs. 1⇑, 2⇑, and 4⇑). This protection of CD28null T cells from Fas-mediated AICD is clearly related to the accompanying high levels of FLIP. Such relationship between FLIP expression and resistance to apoptosis has been reported for various cell types, including T lymphocytes (17, 18, 37, 38).
The down-regulation of FLIP in T cells is IL-2 dependent (17) and coincides with cell cycle progression (18). These findings indicate that FLIP is a convergence point of Fas and IL-2 signaling pathways. They also suggest that the metabolic fate of FLIP is a critical factor in determining whether the transduction of IL-2/IL-2R signals will ultimately lead to apoptosis (16, 17) or the completion of cell division (25, 39). In the present study, CD4+CD28null, but not CD4+CD28+, T cells have high levels of FLIP expression (Fig. 6⇑), which are accompanied by the up-regulation of IL-2Rα subsequent to activation (Fig. 4⇑). Instead of apoptosis, such induction of IL-2Rα expression on CD4+CD28null T cells following incubation with anti-CD3 is associated with the progression of cells through the cell cycle (Fig. 5⇑). In fact, the number of CD28null cells that are in S-G2-M is not diminished by the cocross-linking of CD3 and Fas. These data are in marked contrast to previous studies showing that normal T cells undergo apoptosis as they progress through the cell cycle (19, 40), with predominance of apoptosis during the S phase (20). Collectively, these results indicate that the molecular machinery responsible for FLIP down-regulation is disconnected from IL-2 signaling in CD4+CD28null T cells. Rather than apoptosis, the growth-promoting effects of IL-2 signal transduction appear to be a default pathway in these cells.
Although the pathway that directly links IL-2 to FLIP remains to be elucidated, previous studies have demonstrated that FLIP degradation can be prevented by cyclosporin A and rapamycin, which selectively inhibit IL-2 production and IL-2 signal transduction, respectively (18). Inasmuch as IL-2 production and the induction of IL-2Rα expression are intact in CD4+C28null T cells (3, 4, 6, 23) (Fig. 4⇑), the maintenance of FLIP in these cells suggests a defect in the IL-2-dependent regulation of this anti-apoptotic molecule. This is reinforced by the observation that the addition of exogenous IL-2 altered FLIP expression following activation of CD28+, but not of CD28null, T cells (data not shown). Although earlier studies have suggested a role for transcriptional repression in mouse T cells (17), recent studies with human T cells have shown that IL-2 does not affect the steady state levels of FLIP transcripts (18). IL-2, therefore, appears to influence the translational control of FLIP and/or targeting of the FLIP protein to degradation pathways. Whether or not such pathways regulating FLIP expression are defective in CD4+CD28null T cells remains to be examined.
Recent studies have implicated STAT-5 in the proapoptotic effects of IL-2 (41). Gene reconstitution experiments with IL-2Rβ and STAT-5 knockout mice demonstrated that IL-2 conferred susceptibility to apoptosis in activated mouse T cells through IL-2Rβ-coupled activation of STAT-5, which led to the up-regulation of FasL expression. This confirms that activation invariably results in FasL induction (16, 17, 18), which subsequently triggers cell death upon its interaction with the Fas receptor. In the present work, it is clear that CD28null T cells do not significantly differ from their CD28+ counterparts in the levels of Fas expression (Table I⇑) and the production of biologically active FasL (Fig. 3⇑). Moreover, both cell types express equivalent levels of IL-2Rβ (data not shown). Thus, it is unlikely that protection of CD28null T cells from AICD could be due to a dissociation of STAT-5-dependent events from FasL expression.
It should also be mentioned that STAT-5 has been reported to be an inducer of IL-2Rα (42). Thus, activation-induced up-regulation of IL-2Rα on T cells, as shown in Fig. 4⇑, would be predicted to sustain IL-2 signaling, which would consequently promote FLIP down-regulation and susceptibility to Fas-mediated apoptosis. As already discussed, this is clearly not the case for CD4+CD28null T cells, which up-regulate IL-2Rα following activation and yet maintain FLIP expression. A peculiar feature of these cells, however, is that IL-2Rα is particularly unstable (4). Although clearly inducible (Fig. 4⇑), the high level of IL-2Rα drops precipitously after 24 h (4 and data not shown). Although the molecular basis for IL-2Rα instability on CD4+CD28null T cells is not yet known, their resistance to Fas-mediated cell death is curiously reminiscent of mouse T cells that are deficient in IL-2Rα, which also do not undergo AICD (43, 44).
In conclusion, the present data provide strong evidence for the dysregulation of apoptotic pathways in CD4+CD28null T cells. Unlike their CD28+ counterparts, they maintain high levels of FLIP following activation. Such persistence of FLIP occurs despite the induction of IL-2Rα, and neither the cross-linking of CD3 nor Fas affects cell cycle progression, clearly indicating the absence of proapoptotic effects of IL-2 in these cells. Although these FLIP-expressing CD4+CD28null T cells are a major component of the T cell repertoire in several human autoimmune diseases (2, 28, 45, 46), it is important to note that overexpression of FLIP in mice results in lymphoproliferative autoimmune disorders (47). Because many CD4+CD28null T cells have autoreactive properties (31, 46, 48), the role of FLIP in the differentiation of T cell effector functions is a provocative proposition.
Acknowledgments
We thank Dr. Alicia Algeciras-Schimnich for technical advice and discussions, and James Fulbright for assistance in the preparation of this manuscript.
Footnotes
-
↵1 This work was supported by the Mayo Foundation, the Austrian Research Fund (Schroedinger Grant J01194), and grants from the National Institutes of Health (RO1-AR41974, RO1-AR42527, and RO3-AR45830).
-
↵2 M.S. and A.N.V. contributed equally to this work and must be regarded as co-first authors.
-
↵3 Address correspondence and reprint requests to Dr. Abbe Vallejo (vallejo.abbe@mayo.edu) or Dr. Jörg Goronzy (goronzy.jorg@mayo.edu), Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
-
↵4 Current address: Department of Medicine, University of Innsbruck, Austria.
-
↵5 Abbreviations used in this paper: AICD, activation-induced cell death; 7-AAD, 7-amino actinomycin D; FasL, Fas ligand; FLIP, Fas-associated death domain-like IL-1-converting enzyme inhibitory protein; PerCP, peridinin chlorophyl protein.
- Received May 24, 2000.
- Accepted September 7, 2000.
- Copyright © 2000 by The American Association of Immunologists