Published online before print January 14, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Departamento de Bioquímica y Biología Molecular y Celular, Facultad de Ciencias, and
Servicio de Inmunología, Hospital Clínico Universitario, Universidad de Zaragoza, Spain
1 Correspondence: Dept. Bioquímica y Biología Molecular y Celular, Facultad de Ciencias, Universidad de Zaragoza, Zaragoza, E-50009, Spain. E-mail: anel{at}posta.unizar.es
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: T lymphocytes • growth arrest • apoptosis
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The regulated termination of T cell immune responses seems to be dependent in turn on several complex, possibly overlapping, cellular and molecular mechanisms. On one hand, T cell activation results in induction of the expression of the negative regulator cytotoxic T lymphocyte antigen 4, which competes with CD80/CD86 for the T cell costimulator CD28 [6 ]. Conversely, it is clear that interleukin 2 (IL-2) deprivation as a consequence of antigen exhaustion is one of the main causes of down-modulation of T cell responses [7 ]. It has been suggested that this is not just a passive cell death process but is related with the activated status of the expanded T cell population, re-termed as "activated T cell autonomous cell death" [8 ]. The BH3-only, proapoptotic member of the Bcl-2 family Bim has been demonstrated to play a main role in this process, and defects in their expression are associated with autoimmunity [9 , 10 ]. Finally, activation-induced cell death (AICD) of T cells generated by clonal expansion is also implicated in normal termination of immune responses. This process was first reported to be dependent on death receptor/death ligand interplay, especially on the Fas/Fas ligand (FasL) system [11 , 12 ]. In fact, lpr and gld mice, deficient, respectively, in functional Fas or FasL expression [13 , 14 ], or humans with similar defects [15 ] have systemic autoimmune disease characterized by lymphoproliferation. Our group suggested for the first time the participation of APO 2 ligand (APO2L)/tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and its receptors DR4 and DR5 in the down-regulation of T cell responses [16 ], something that has been confirmed recently in the APO2L/TRAIL knockout mice, which are more susceptible to autoimmune disease induction [17 ]. In addition, death receptor-independent mechanisms are implicated in AICD, such as the generation of free radicals after T cell receptor (TCR) engagement and/or Bim-dependent processes [10 , 18 ].
We characterized previously that FasL and APO2L/TRAIL are stored inside human T cell blasts in cytoplasmic compartments with characteristics of multivesicular bodies [19 ] and that they were rapidly released to the supernatant in their bioactive form associated with the internal microvesicles upon reactivation [19 , 20 ]. Upon phytohemagglutinin (PHA) or CD3 reactivation, microvesicles loaded with FasL and APO2L/TRAIL were secreted, and upon reactivation with anti-CD59 monoclonal antibodies (mAb), the secretion of microvesicles labeled preferentially with APO2L/TRAIL predominated [19 , 21 ]. In our previous studies, supernatants obtained from normal human T cell blasts were tested for apoptosis induction using a functional bioassay on tumoral Jurkat T cells, which are constitutively sensitive to FasL and to APO2L/TRAIL-induced cell death. However, naïve T lymphocytes are not sensitive to death receptor-induced apoptosis nor to AICD, and T cell blasts begin to be sensitive to AICD induction only at day 6 [16 , 22 ]. This sensitizing to AICD and to death receptor-induced apoptosis could be related with the expression or regulation of pro- or antiapoptotic proteins during the process of T cell blast generation. One of the most studied proteins in this respect is cellular Fas-associated death domain-like IL-1ß-converting enzyme-inhibitory protein (c-FLIP), of which two cellular isoforms are known, c-FLIP long and short (c-FLIPL and c-FLIPS, respectively) [23 ]. Initial data indicated that IL-2 reduced c-FLIPL levels in activated T cells, making them more sensitive to death receptors [23 , 24 ]. However, these results were not confirmed by other authors, which reported no c-FLIPL reduction during T cell blast generation [25 ]. It is now accepted that c-FLIPS acts as a potent inhibitor of Fas- and DR4/DR5-induced cell death through complete inhibition of caspase-8 activation [26 27 28 ], and the role of c-FLIPL seems to be more complex. Other proteins that have been implicated in apoptosis sensitivity regulation in T cells are Bcl-xL [22 , 29 ], Bcl-2, and Bim [10 ].
In this study, we have undertaken a systematic study using normal human T cell blasts from at least 20 healthy donors to characterize the changes in the expression of antiapoptotic (c-FLIPL, c-FLIPS, Bcl-2, Bcl-xL) or proapoptotic proteins (Bcl-xS, caspase-8, Bim) during the process of T cell blast generation and their correlation with protection or sensitivity to apoptosis; the effect of cytotoxic anti-Fas mAb, of recombinant APO2L (rAPO2L)/TRAIL, or of supernatants containing bioactive FasL and/or APO2L/TRAIL associated with microvesicles on normal human T cell blasts; the expression pattern of pro- and antiapoptotic proteins in T cell blasts surviving PHA reactivation, which have a memory phenotype; the changes in the expression pattern of pro- and antiapoptotic proteins after sensitizing of T cell blasts to death receptor-induced apoptosis by an anti-CD3 or anti-CD59 pulse; and the contribution of FasL, APO2L/TRAIL, and free radical generation to human T cell blast AICD.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cytotoxicity assays
To analyze the sensitivity of human T cells blasts to FasL and rAPO2L/TRAIL, 2 x 106 cells/ml were incubated, respectively, with 100 ng/ml anti-human Fas mAb (CH11; Upstate Biotechnology, Lake Placid, NY) or 500 ng/ml rAPO2L during 24 h in the presence or absence of 1 µg/ml cycloheximide (CHX). The rAPO2L preparation used in this study corresponds to APO2L.0, which predominates the monomer form [30
] and different from previous preparations, does not contain a polyhistidine tag, which promoted aggregation [31
]. rAPO2L was kindly provided by Dr. Avi Ashkenazi (Genentech, South San Francisco, CA). These experiments were also performed in the presence of 100 ng/ml of the blocking anti-human Fas mAb SM1/23 (Bender Medsystems, Bascelona, Spain) or the anti-APO2L/TRAIL mAb 5C2 to prevent CH11 and rAPO2L from binding to their receptors, respectively [16
, 21
, 32
].
To obtain supernatants from day 6 human T cells blasts, 5 x 106 cells/ml were restimulated by incubation with 50 µg/ml PHA at 37°C for 5 min. PHA was then removed by brief centrifugation in a Beckman Minifuge (12,000 g, 5 s), and cells were washed three times with RPMI 1640, resuspended in complete medium, and incubated for 1 h at 37°C. Supernatants were collected after incubation and clarified by two sequential centrifugations at 18,000 g for 5 s each. For CD59 stimulation, the anti-human CD59 mAb VJ1/12.2, kindly provided by Dr. Francisco Sánchez-Madrid (Hospital de la Princesa, Madrid, Spain), was immobilized in the bottom of the wells of 24-well plates. Wells were washed with phosphate-buffered saline (PBS), cells (5x106 cells/ml) were added in fresh complete medium, and supernatants were collected after 3 h of incubation at 37°C. For CD3 stimulation, the anti-human CD3 mAb UCHT1, kindly provided by Dr. Marisa Toribio (Centro de Biología Molecular "Severo Ochoa," Madrid, Spain), was used, and 3-h supernatants were obtained as indicated for CD59 stimulation. Toxicity of supernatants from cells pulse-stimulated with PHA, anti-CD3, or anti-CD59 mAb was tested on 6-day human T cell blasts from the same donors by culturing them at 2 x 106 cells/ml in these supernatants during 24 h in the presence or absence of 1 µg/ml CHX. Supernatants from nonstimulated cells were used as controls. To analyze the involvement of secreted FasL or APO2L/TRAIL in the toxicity of these supernatants, bioassays were also performed in the presence of SM1/23 and/or 5C2 as described previously.
To analyze the sensitizing of human T cells blasts to FasL or APO2L/TRAIL after a pulse with anti-CD3 or anti-CD59 mAb, antibodies were immobilized in the bottom of the wells of 24-well plates. Wells were washed with PBS, and cells (5x106 cells/ml) were added in fresh complete medium and incubated during 3 h at 37°C. After this pulse, 5 x 106 cells were lysed, and protein expression was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Another fraction of pulsed cells was washed with RPMI and cultured in fresh medium in the presence of 100 ng/ml anti-human Fas mAb CH11 or 500 ng/ml rAPO2L during 24 h.
Sensitivity to AICD was tested by incubation of 2 x 106 cells/ml human T blasts with PHA concentrations ranging from 1 to 5 µg/ml for 24 h. Growth inhibition by CD3 or CD59 ligation was tested by incubation of 2 x 106 cells/ml human T blasts in wells of 24-wells plates coated with anti-CD3 or anti-CD59 mAb, as described previously, at 37°C for 24 h.
In all of the mentioned cytotoxicity assays, cell death was determined by Trypan blue staining and microscopic inspection and also by propidium iodide (PI) staining and flow cytometry, and cell growth was estimated by counting Trypan blue-negative (viable) cells and by a modification of the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) reduction method for microplates [32 ].
Apoptosis induction was also tested after 6 h of PHA treatment by determining loss of mitochondrial membrane potential (
m) and the production of reactive oxygen species (ROS). Briefly, control or treated cells (5x105 in 100 µl) were incubated with 5 nM 3,3'-dihexyloxacarbocyanine iodide [DiOC6(3)] and with 2 µM 2-hydroxyethidium at 37°C for 10 min in RPMI, and m (green fluorescence) and 2-hydroxyethidium oxidation to ethidium (red fluorescence) were determined by flow cytometry. To analyze the involvement of ROS, cells were preincubated with 400 µM of the porphyrin with superoxide dismutase-like activity manganese (III) tetrakis (5,10,15,20-benzoic acid; MnTBAP) [18
] for 1 h at 37°C. To analyze the involvement of FasL or APO2L/TRAIL, asays were also performed in the presence of 1 µg/ml human anti-FasL mAb NOK-1 (PharMingen, San Diego, CA) and 100 ng/ml anti-APO2L/TRAIL mAb 5C2, as described previously.
Immunoblotting
Cells (5x106) were lysed at 4°C in 100 µl of a buffer containing 1% Triton X-100 and protease and phosphatase inhibitors, as described previously [33
]. Lysed cells (1x106) were analyzed by SDS-PAGE on 12% gels, and separated proteins were transferred to nitrocellulose membranes, which were blotted with anti-c-FLIP (1.5 µg/ml, Alexis Biochemicals, San Diego, CA; clone Dave-2), anti-BclXS/L (0.4 µg/ml, Santa Cruz Biotechnology, Santa Cruz, CA; sc-1041), anti-Bcl-2 (0.4 µg/ml, Santa Cruz Biotechnology; sc-783), anti-caspase-8 (0.5 µg/ml, Upstate Biotechnology; clone 5F7), anti-Bim (0.5 µg/ml, Calbiochem, San Diego, CA; B39874), or anti-ß-actin (1/5000, Sigma Chemical Co., St. Louis, MO; clone AC-15) antibodies in 10 mM Tris/HCl, pH 8.0, 0.12 M NaCl, 0.1% Tween-20, 0.05% sodium azide (TBS-T) containing 2% skimmed milk. Membranes were washed with TBS-T and incubated with 0.2 µg/ml of the corresponding phosphatase alkaline-labeled secondary antibody (Sigma Chemical Co.). Immunoblots were revealed with the 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium substrate, as described previously [16
]. The expression level of the different proteins analyzed was quantified in a densitometer (Bio-Rad, Barcelona) and normalized to the same amount of ß-actin.
Flow cytometry
To analyze the different cellular populations present in fresh PBMC preparations or during the generation of human T cells blasts, 5 x 105 cells were stained for fluorescein-activated cell sorter (FACS) analysis in cold PBS containing 0.2% bovine serum albumin with anti-CD3-fluorescein isothiocyanate (FITC), anti CD19-phycoerythrin (PE), anti CD4-FITC, or anti CD8-PE (all from Caltag Laboratories, Burlingame, CA). PBMC (60±10%) and 90 ± 5% of human T cells blasts were CD3+. During the activation process, the CD4/CD8 ratio changed consistently from 2:1 in fresh PBMC to 0.75:1 in day-6 T cell blasts. To analyze the generation of memory T cells, cells were stained for FACS analysis, as described previously, with anti CD44-FITC mAb (Caltag Laboratories).
Labeling of microvesicles secreted by reactivated human T cell blasts for detection of FasL and/or APO2L/TRAIL expression was performed as indicated in ref. [19 ], using the FITC-labeled, rat anti-human FasL mAb (H11, Bender Medsystems) and the mouse anti-human APO2L/TRAIL mAb 5C2 plus PE-labeled, goat anti-mouse immunoglobulin G (Caltag Laboratories). Labeling was detected by flow cytometry using the gating protocol described previously [19 ].
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
As shown in Figure 1A , variability existed in the expression pattern of c-FLIPL among different donors, exemplified by five donors in the figure. Although in some cases, c-FLIPL levels in freshly isolated PBL were almost undetectable (e.g., donor 6), in most cases, this level was detectable—high in some (e.g., donors 12 and 14) and lower in others (e.g., donors 15 and 16). After 1 day of activation, this level was clearly increased in some donors (e.g., donor 6), or in most of them, it did not change substantially (e.g., donors 14–16), and exceptionally, it was decreased (only observed in donor 12). In day-6 T cell blasts, c-FLIPL levels increased with respect to day 1 (e.g., donors 12, 15, and 16: 54% of increase of the c-FLIPL/ß-actin ratio as a mean), or they did not change (e.g., donor 6), only decreasing by 30% in the case of donor 14. Hence, the most frequent patterns of c-FLIPL expression do not correlate with the higher sensitivity of T cell blasts to AICD or to death receptor-induced apoptosis. This result does not agree with early reports [23 , 24 ] but does with later ones [25 ]. Conversely, the expression pattern of c-FLIPS was the same in all donors analyzed. This protein was not detected in fresh PBL but was readily induced after 1 day of activation, disappearing already at day 4 and its expression undetectable in day-6 T cell blasts (Fig. 1A , donors 6, 12, and 14). This expression pattern does correlate with the protection of recently activated T cells to death receptor-induced apoptosis and with the higher susceptibility of T cell blasts to deletion.
 View larger version (50K): [in a new window] |
Figure 1. Expression pattern of Bcl-2 and c-FLIP and Bcl-x isoforms during normal human T cell blast generation. Anti-c-FLIP (A), Bcl-x (B), or Bcl-2 (C) immunoblots were performed on extracts from fresh peripheral blood lymphocytes (PBL) obtained from healthy donors (day 0) or at different times (1–13 days, as indicated) after PHA stimulation in the presence of IL-2, as described in Materials and Methods. The extracts used correspond to 1 x 106 cells, and expression levels of the proteins analyzed were quantified in a densitometer and normalized to the same amount of ß-actin. Molecular weight markers are indicated on the left, and the positions of c-FLIPL, c-FLIPS (A), Bcl-xL, Bcl-xS (B), Bcl-2 (C), and ß-actin (A–C) are indicated with arrows. Results shown are representative of at least 10 donors for each protein analyzed.
|
One of the most remarkable changes observed in our analysis was related to the expression of the short splice form of Bcl-x. This protein exerts proapoptotic activity by competition with the antiapoptotic proteins Bcl-2 and Bcl-xL, making cells with high Bcl-xS expression susceptible to the mitochondrial apoptotic pathway [35 ]. As shown in Figure 1B and observed in all donors analyzed, Bcl-xS expression was undetectable in fresh PBL or after 1 day of activation. However, the levels of Bcl-xS were strongly induced in day-5 or -6 T cell blasts at the point when T cell blasts begin to be susceptible to AICD. This pattern of expression correlates with the sensitivity of human T cell blasts to AICD and has not been reported before. The level of Bcl-xS decreased at longer times of culture (Fig. 1B) , when T cells with a memory phenotype (CD44high) predominate (data not shown).
Caspase-8 was expressed in fresh PBL, as its expression level in day-6 T cell blasts was equivalent to the initial level in most donors (e.g., donors 14 and 15, Fig. 2A ). Only in donor 16 was caspase-8 expression increased upon activation and remained elevated in day-6 T cell blasts (Fig. 2A) .
 View larger version (53K): [in a new window] |
Figure 2. Expression pattern of caspase-8 and Bim isoforms during normal human T cell blast generation. Anticaspase-8 (A) or Bim (B) immunoblots were performed on extracts from fresh PBL obtained from healthy donors (day 0) or at different times (1–13 days, as indicated) after PHA stimulation in the presence of IL-2, as described in Materials and Methods. The extracts used correspond to 1 x 106 cells, and expression levels of the proteins analyzed were quantified in a densitometer and normalized to the same amount of ß-actin. Molecular weight markers are indicated on the left, and the positions of the procaspase-8 doublet (A), of BimEL, BimL and Bims (B), and of ß-actin (A, B) are indicated with arrows. Results shown are representative of at least 10 donors for each protein analyzed.
|
Fas/CD95 expression was detected by flow cytometry in the CD3+ population of fresh PBL, and its expression level increased during blast generation, following exactly the same pattern as previously reported [22 ] (data not shown). A similar increase was also observed after 1 day of activation, again in agreement with ref. [22 ], and as Fas engagement in fresh PBL induces costimulation and not cell death [36 ], AICD sensitivity should be rather a result of changes in intracellular signal transducers. Conversely, approximately 40% of CD3+ fresh PBL expressed APO2L/TRAIL receptors, which did not further increase during T cell blast generation (data not shown).
Effect of cytotoxic anti-Fas mAb and of rAPO2L/TRAIL on human T cell blasts: sensitizing to cell death by CHX
Our previous studies, in which we characterized the possible implication of APO2L/TRAIL in the down-regulation of human T cell responses [16
], and the release of bioactive FasL and APO2L/TRAIL associated with microvesicles during the AICD process of human T cell blasts [19
, 20
] were performed using bioassays on Jurkat tumoral T cells, which are constitutively sensitive to death receptor-induced apoptosis. We wanted to perform the same job using normal human T cell blasts as targets to give physiological validity to our previous observations.
First, we performed 24-h cytotoxicity tests on day-6 T cell blasts using the cytotoxic anti-Fas mAb CH11 or rAPO2L/TRAIL at the same concentrations that induce maximum levels of apoptosis on Jurkat cells (100 and 500 ng/ml, respectively). Assays were performed in the absence or in the presence of IL-2. In the absence of IL-2, cells did not grow, and in some cases, some level of basal cell death was observed at 24 h (Fig. 3A ). In these conditions, CH11 or rAPO2L did not induce significant levels of cell death over that observed in control cells. Only in the presence of the protein synthesis inhibitor CHX, which did not induce cell death by itself, did CH11 induce high levels of cell death. However, rAPO2L did not induce cell death even in the presence of CHX. The level of cell death increased substantially in control cells from all donors after 48 h of IL-2 deprivation (data not shown).
 View larger version (11K): [in a new window] |
Figure 3. Effect of anti-Fas mAb or rAPO2L/TRAIL on normal human T cell blasts. (A) Day-6 T cell blasts were incubated at 2 x 106 cells/ml with control medium (Cont), 100 ng/ml cytotoxic anti-human Fas mAb CH11, or 500 ng/ml rAPO2L/TRAIL (APO2L) during 24 h in the presence or absence of 1 µg/ml CHX, as indicated, and in the absence of IL-2. (B) Day-6 T cell blasts were incubated at 2 x 106 cells/ml with 1 µg/ml CHX, 100 ng/ml CH11, or 500 ng/ml rAPO2L/TRAIL, as indicated, during 24 h in the presence of 30 UI/ml IL-2. (C) Day-6 T cell blasts were incubated at 2 x 106 cells/ml with control medium (–), 100 ng/ml CH11, or 500 ng/ml rAPO2L/TRAIL during 24 h in the presence of 1 µg/ml CHX and 30 UI/ml IL-2, as indicated. Assays were also performed in the presence of 100 ng/ml of the anti-Fas-blocking mAb SM1/23 or the anti-APO2L/TRAIL-blocking mAb 5C2, as indicated. (A and C) Cell death was analyzed by Trypan blue and PI staining with identical results. (B) Cell growth was estimated by counting Trypan blue-negative cells, with similar results using the MTT reduction method, and results were expressed as percentage of cell growth inhibition with respect to control cells. Results are the mean ± SD of experiments performed in cells from six different donors.
|
Susceptibility of normal human T cell blasts to death receptor-induced apoptosis by CHX treatment correlated in all cases with a drastic reduction in c-FLIPL levels (55% of reduction in the blot shown in Fig. 4 ) if compared with ß-actin or other proteins, such as Bim or Bcl-xL, which expression was more slightly affected (15% of reduction for ß-actin in the blot shown in Fig. 4 ). It also correlated with the loss of residual c-FLIPS levels (Fig. 4) , as previously reported in tumoral cells [37 ].
 View larger version (79K): [in a new window] |
Figure 4. Sensitizing human T cell blast to death receptor-induced apoptosis by CHX correlates with reduction in the expression levels of c-FLIPL and c-FLIPS. Anti-c-FLIP immunoblots were performed on extracts from fresh PBL obtained from healthy donors (day 0) or after 1 or 6 days of PHA stimulation in the presence of IL-2, as indicated. Day-6 T cell blasts were additionally incubated for 24 h in the presence or absence of 1 µg/ml CHX, as indicated. The extracts used correspond to 1 x 106 cells, and expression levels of c-FLIPL and c-FLIPS were quantified in a densitometer and normalized to the same amount of ß-actin. The positions of c-FLIPL, c-FLIPS, and ß-actin are indicated with arrows. Results shown are representative of five donors.
|
 View larger version (26K): [in a new window] |
Figure 5. Effect of supernatants obtained from T cell blasts pulse-stimulated with PHA, anti-CD3, or anti-CD59 mAb on normal human T cell blasts. (A) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 24 h with control medium (Control) or supernatants obtained from T cell blasts from the same donors pulse-stimulated with PHA (ssn PHA), anti-CD3 (ssn  -CD3), or anti-CD59 mAb (ssn -CD59) in the presence or absence of 1 µg/ml CHX, as indicated, and in the absence of IL-2. Assays using supernatants plus CHX were also performed in the presence of 100 ng/ml of the anti-Fas-blocking mAb SM1/23, the anti-APO2L/TRAIL-blocking mAb 5C2, or a combination of both, as indicated. (B) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 24 h with 1 µg/ml CHX or with supernatants from T cell blasts from the same donors obtained as in A, as indicated, in the presence of 30 UI/ml IL-2. (C) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 24 h with control medium (–) or with supernatants from T cell blasts from the same donors obtained as in A, as indicated, in the presence of 1 µg/ml CHX and 30 UI/ml IL-2. Assays were also performed in the presence of 100 ng/ml SM1/23 or 5C2, as indicated. (A and C) Cell death was analyzed by Trypan blue and PI staining with identical results. (B) Cell growth was estimated by counting Trypan blue-negative cells with similar results using the MTT reduction method, and results were expressed as percentage of cell growth inhibition with respect to control cells. In these experiments, supernatants were obtained from pulsed cells, as indicated in Materials and Methods, and were control media supernatants from the same T cell blasts not pulsed with any stimuli. Results are the mean ± SD of experiments performed in cells from six different donors.
|
A PHA pulse, which causes T cell blast death, induces c-FLIPS and Bcl-xL in the surviving population, which has a memory phenotype
Data presented in the previous sections indicate that human T cell blasts, although sensitive to AICD, are refractory to death receptor-induced apoptosis unless CHX is present. However, it could be possible that reactivation of T cell blasts would also make them sensitive to death receptor-induced cell death. To analyze this, we performed experiments giving 6-day T cell blasts a pulse with PHA, anti-CD3, or anti-CD59 mAb and after that pulse, treating them overnight with anti-Fas cytotoxic antibody or with rAPO2L. We also took cells just after the pulse to analyze by immunoblot the possible changes in anti- or proapoptotic proteins. Results for anti-CD3 or anti-CD59 pulses will be shown in the next subsection. In the case of a pulse with PHA, approximately 70% of the cells were dead even in the absence of anti-Fas or rAPO2L treatment, and in fact, death receptor ligation did not increase the level of cell death induced by the PHA pulse alone (data not shown). However, the anti-c-FLIP immunoblots made on pulsed T cell blasts gave a paradoxical result. As shown in Figure 6A
, the 5-min PHA pulse that induced massive cell death later on also clearly induced, already after 1 h and peaking 3 h after the pulse, a net increase in c-FLIPS expression. This is remarkable, as c-FLIPS is the most efficient protector against death receptor-induced cell death [26
27
28
]. We reasoned that the induction of c-FLIPS expression could be restricted to cells surviving the AICD process, which could be the precursors of memory cells. To analyze this hypothesis, we generated T cell blasts, gave a 5-min PHA pulse, recovered surviving cells 24 h after (no more than 30% of the initial T cell blast population), and analyzed CD44 and c-FLIP expression by flow cytometry (Fig. 6B) and by immunoblot, respectively (Fig. 6C)
. During the process of T cell blast generation, as expected, CD44 expression was induced in all T cells. In day-6 T cell blasts, the CD44high population reached 63%. However, almost all cells surviving after the PHA pulse (92%) were CD44high, a marker of T cell memory phenotype. This surviving population was enriched in c-FLIPS expression if compared with the initial T cell blast population (Fig. 6C
, upper panel). Regarding other pro- or antiapoptotic proteins, it was also observed that although the level of Bcl-xS expression was the same in the T cell blast population before PHA reactivation and in the surviving population, Bcl-xL expression was selectively enriched in the surviving population (Fig. 6C
, lower panel). No significant changes were observed in the expression of Bim, caspase-8, Bcl-2, or Mcl-1 in the surviving population (data not shown).
 View larger version (37K): [in a new window] |
Figure 6. Human T cell blasts surviving PHA-induced AICD are memory T cells with increased c-FLIPS and Bcl-xL expression. (A) Anti-c-FLIP immunoblots were performed on extracts from day-6 T cell blasts pulse-stimulated with 50 µg/ml PHA during 5 min and were cells further incubated in fresh medium during 1 h or 3 h or overnight (ON), as indicated. The extracts used correspond to 1 x 106 cells, and expression levels of c-FLIPL and c-FLIPS were quantified in a densitometer and normalized to the same amount of ß-actin. (B) CD44 expression, as determined by flow cytometry of fresh PBL (top panel), day-6 T cells blasts (middle panel), or day-6 T cell blasts surviving overnight reactivation with PHA (bottom panel), following the same protocol as in A. Numbers in the upper-right correspond to the percentage of CD44high cells, which are from the same donor as in C. (C) Anti-c-FLIP (upper panel) or anti-Bcl-x (lower panel) immunoblots were performed on extracts from fresh PBL (day 0), from day-6 T cell blasts, or from day-6 T cell blasts surviving (Surv.) overnight reactivation with PHA following the same protocol as in A, as indicated. In this case, the extracts used correspond to the same amount of total protein, as determined by the Bradford method, and expression levels of the proteins analyzed were quantified in a densitometer and normalized to the same amount of ß-actin. The positions of c-FLIPL, c-FLIPS, and ß-actin (A, C) and of Bcl-xL and Bcl-xS (C) are indicated with arrows. Results shown are representative of three donors.
|
 View larger version (32K): [in a new window] |
Figure 7. Normal human T cell blasts are sensitized to cytotoxic anti-Fas mAb or to rAPO2L/TRAIL-induced cell death by a pulse with anti-CD3 or anti-CD59 mAb. (A) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 24 h in the presence of IL-2 with control medium, 100 ng/ml cytotoxic anti-human Fas mAb CH11, or 500 ng/ml rAPO2L/TRAIL (rAPO2L), as indicated, only after receiving no pulse (Control) or a 3-h pulse with immobilized anti-CD3 (-CD3 pulse) or anti-CD59 mAb (-CD59 pulse). Cell death was analyzed by Trypan blue and by PI staining with identical results. Results shown are the mean ± SD of four different donors. (B) Anti-c-FLIP immunoblots were performed on extracts from day-6 T cell blasts before or after the 3-h pulse with immobilized anti-CD3 or anti-CD59 mAb, as indicated. The extracts used correspond to 1 x 106 cells, and expression levels of c-FLIPL and c-FLIPS were quantified in a densitometer and normalized to the same amount of ß-actin. Results shown are representative of the same four donors used in A.
|
m and were positive for ROS generation. Although MnTBAP reduced ROS production approximately 50%, it did not affect m loss. However, the combination of anti-FasL and anti-APO2L/TRAIL-blocking mAb partially reduced m loss and also inhibited free radical generation to 50%. The combination of the blocking mAb with MnTBAP further prevented m loss, reaching a maximum of 30% of inhibition. The incomplete inhibition of these early apoptotic processes correlates with the lack of inhibition of cell death induction after 24 h (data not shown).
 View larger version (28K): [in a new window] |
Figure 8. FasL, APO2L/TRAIL, and ROS generation are partially implicated in PHA-induced apoptosis and in anti-CD3 or anti-CD59 inhibition of IL-2-dependent growth of normal human T cell blasts. (A) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 6 h with control medium (Control) or with 2.5 µg/ml PHA in the presence or absence of 400 µM MnTBAP, a combination of 1 µg/ml anti-FasL-blocking mAb NOK-1 and 100 ng/ml anti-APO2L/TRAIL-blocking mAb 5C2, or a combination of MnTBAP, NOK-1, and 5C2, as indicated. The loss of m was analyzed by staining with DiOC6(3) and flow cytometry, and results corresponded to the percentage of cells showing a low-staining DiOC6(3). ROS generation was analyzed by staining with 2-hydroxyethidium and flow cytometry, and results corresponded to the percentage of cells positive for ROS generation. Results shown are the mean ± SD of three different donors. (B) Day-6 T cell blasts were incubated at 2 x 106 cells/ml during 24 h with immobilized anti-CD3 or anti-CD59 mAb in the presence or absence (–) of 400 µM MnTBAP, a combination of 1 µg/ml NOK-1 and 100 ng/ml of 5C2, or a combination of MnTBAP, NOK-1, and 5C2, as indicated. Cell growth was estimated by counting Trypan blue-negative cells with similar results using the MTT reduction method, and results were expressed as a percentage of cell growth inhibition with respect to control cells. Results are the mean ± SD of experiments performed in cells from the same three donors as in A.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the absence of IL-2, T cell blasts begin to die at 24 h, and cell death is prominent after 48 h. In these conditions, no additional effect of death receptor ligation could be observed (Figs. 3A and 5A) , in agreement with Hieronymus et al. [41 ]. Hence, the mechanism of down-modulation of T cell responses dependent on IL-2 deprivation seems to predominate on death receptor-induced apoptosis. The increased Bcl-xS and Bim levels in T cell blasts reported here could be more related to this down-regulation mechanism, which is dependent on the mitochondrial apoptotic pathway, as proposed previously [8 , 41 ].
In the presence of IL-2 and in the absence of any pharmacological treatment, cytotoxic anti-Fas mAb or rAPO2L/TRAIL rather induces inhibition of IL-2-dependent growth and not cell death on normal human T cell blasts (Fig. 3B) . This observation is new and rather unexpected. However, T cell clonal expansion shutdown by FasL and/or APO2L/TRAIL would be enough to justify their immunoregulatory role in a process that seems independent of apoptosis induction. In fact, the data presented here extend to normal human T cell blast regulation. Similar effects of APO2L/TRAIL, observed previously on autoantigen-specific T lymphocytes [42 43 44 ], associated in one study with a decrease in calcium signaling and CDK4 expression [43 ]. Our data also extend this type of apoptosis-independent regulation of normal T cell expansion for the first time to Fas ligation.
The physiological validity of this observation is supported by the fact that supernatants from T cell blasts pulse-stimulated with PHA or through CD3 or CD59 ligation, containing bioactive FasL and/or APO2L/TRAIL on microvesicles, had the same effect on normal human T cell blasts from the same donors (Fig. 5B) . Finally, CD3 or CD59 ligation rather resulted in the inhibition of IL-2-dependent T cell growth and not in cell death, which was partially dependent on FasL and/or APO2L/TRAIL release (Fig. 8B) .
In these experiments, cell death was only observed in the presence of CHX (Figs. 3 , A and C, and 5 , A and C), correlating with a net reduction in c-FLIPL and c-FLIPS expression (Fig. 4) . Cell death induced by the above-mentioned supernatants in the presence of CHX was prevented by blocking anti-Fas and/or anti-APO2L/TRAIL mAb (Fig. 5 A and C) , similarly to results obtained before using Jurkat cells as targets [16 , 19 , 20 ]. The use of CHX, a pharmacological agent, to sensitize T cell blasts to death receptor-induced apoptosis, would preclude the interpretation of these experiments in a physiological context. However, a pulse through CD3 or CD59 in the presence of IL-2 sensitizes human T cell blasts to Fas or APO2L-induced cell death, also inducing a reduction in c-FLIPL and c-FLIPS levels (Fig. 7) . This observation is in agreement with a recent report about the sensitizing to Fas by CD3 ligation. In that study, the sensitizing is attributed to Fas recruitment to lipid rafts [45 ]. In the present study, we show that this sensitizing effect is also associated with a reduction in c-FLIP levels and extend these results to CD59 triggering and sensitizing to APO2L/TRAIL-induced apoptosis.
We have also observed that a PHA pulse directly induces cell death on most of normal human T cell blasts but that all T cells surviving AICD are memory CD44high T cells, in agreement with previous proposals [46 ]. In addition, our data offer a molecular explanation for the resistance of surviving T cells to AICD. On one hand, the specific induction of c-FLIPS expression in this population makes it more resistant to death receptor-induced apoptosis. Conversely, the specific enrichment in Bcl-xL levels makes it more resistant to the mitochondrial apoptotic pathway (Fig. 6) . Taking into account that Bim and Bcl-xS expression is already high in day-6 T cell blasts before reactivation (see Figs. 1B and 2B ), an additional increase in Bcl-xL expression is necessary for surviving reactivation. An increased Bcl-2 expression in mouse memory CD8+ but not CD4+ T cells has been described [47 ]. We could not confirm this result, but this could be a result of the use of a mixed population in our experiments. The increase in Bcl-xL expression in human memory T cells has been described previously but only in long-term human T cell lines [48 ]. The increase in Bcl-xL and c-FLIPS expression upon TCR/CD28 ligation in human T cell blasts has been described, although not in the context of AICD survival [27 ].
Our data also indicate that death receptor and free radicals are involved, at least partially, in PHA-induced AICD (Fig. 8A) . However, the incomplete inhibition of early PHA-induced m loss by a combination of MnTBAP and FasL and APO2L/TRAIL-blocking mAb, which correlate with the lack of inhibition of cell death induction, indicates that other cell death mechanisms are involved in the AICD process. Conversely, our results indicate that the inhibition of IL-2-dependent cell growth by CD59 ligation is mainly mediated by free radical generation and by death receptors (preferentially through bioactive APO2L/TRAIL secretion; see also refs. [19
, 21
]). However, inhibition of IL-2-dependent cell growth by CD3 ligation, although also involving free radical generation, FasL, and APO2L/TRAIL, seems to be dependent on additional mechanisms (Fig. 8B)
.
Together with the recent data about the higher susceptibility of APO2L/TRAIL knockout mice to autoimmunity [17 ], the present data further demonstrate the role of APO2L/TRAIL as a physiological regulator of normal human T cell responses. This role is achieved by cell- or microvesicle-associated APO2L/TRAIL, controlling T cell blast proliferation if they are not reactivated and inducing cell death if they are reactivated. APO2L/TRAIL is always less potent than FasL but predominates in the regulation mediated by CD59 ligation.
Finally, the role of CD59 as a possible physiological regulator of normal human T cell responses is also demonstrated. This observation opens the intriguing possibility that CD59, although inhibiting complement attack, also contributes to T cell activation and to down-modulation of T cell responses. In fact, it has been reported that mice deficient in glycosylphosphatidylinositol-linked proteins have autoimmunity problems [49 ].
Received September 15, 2004; revised November 15, 2004; accepted December 9, 2004.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 Biochem. J. 284,189-199
B upregulates cFLIP, a cycloheximide-sensitive inhibitor of death receptor signaling Mol. Cell. Biol. 21,3964-3973This article has been cited by other articles:
 |
|
 |
|
  A. Bosque, J. I. Aguilo, M. del Rey, E. Paz-Artal, L. M. Allende, J. Naval, and A. Anel Cell cycle regulation by FasL and Apo2L/TRAIL in human T-cell blasts. Implications for autoimmune lymphoproliferative syndromes J. Leukoc. Biol., August 1, 2008; 84(2): 488 - 498. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Meinander, T. S. Soderstrom, A. Kaunisto, M. Poukkula, L. Sistonen, and J. E. Eriksson Fever-Like Hyperthermia Controls T Lymphocyte Persistence by Inducing Degradation of Cellular FLIPshort J. Immunol., March 15, 2007; 178(6): 3944 - 3953. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Bosque, J. I. Aguilo, M. A. Alava, E. Paz-Artal, J. Naval, L. M. Allende, and A. Anel The induction of Bim expression in human T-cell blasts is dependent on nonapoptotic Fas/CD95 signaling Blood, February 15, 2007; 109(4): 1627 - 1635. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. Rautajoki, E. M. Marttila, T. A. Nyman, and R. Lahesmaa Interleukin-4 Inhibits Caspase-3 by Regulating Several Proteins in the Fas Pathway during Initial Stages of Human T Helper 2 Cell Differentiation Mol. Cell. Proteomics, February 1, 2007; 6(2): 238 - 251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Laudanski, C. Miller-Graziano, W. Xiao, M. N. Mindrinos, D. R. Richards, A. De, L. L. Moldawer, R. V. Maier, P. Bankey, H. V. Baker, et al. Cell-specific expression and pathway analyses reveal alterations in trauma-related human T cell and monocyte pathways PNAS, October 17, 2006; 103(42): 15564 - 15569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Del-Rey, J. Ruiz-Contreras, A. Bosque, S. Calleja, J. Gomez-Rial, E. Roldan, P. Morales, A. Serrano, A. Anel, E. Paz-Artal, et al. A homozygous Fas ligand gene mutation in a patient causes a new type of autoimmune lymphoproliferative syndrome Blood, August 15, 2006; 108(4): 1306 - 1312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Neujahr, C. Chen, X. Huang, J. F. Markmann, S. Cobbold, H. Waldmann, M. H. Sayegh, W. W. Hancock, and L. A. Turka Accelerated Memory Cell Homeostasis during T Cell Depletion and Approaches to Overcome It. J. Immunol., April 15, 2006; 176(8): 4632 - 4639. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
