Constitutive Caspase Activation and Impaired Death-Inducing
Signaling Complex Formation in CD95-Resistant, Long-Term
Activated, Antigen-Specific T Cells
Gudrun Strauss,* Ingrid Knape,* Ingo Melzner,†and Klaus-Michael Debatin1*
Elimination of T cells during an immune response is mediated by activation-induced cell death (AICD) and CD95-mediated
apoptosis. Chronic graft-vs-host disease and T cell-mediated autoimmune diseases are caused by the persistence of activated T cells
that escaped tolerance induction by deletion or silencing. To mimic the in vivo situation of long-term activated T cells, we
generated an in vitro system using HLA-A1-specific T cells, weekly restimulated by Ag. While short-term activated T cells (two
to five rounds of stimulation) were CD95 sensitive and susceptible to AICD, T cells stimulated more than eight times acquired
constitutive CD95 resistance and exhibited reduced AICD. Phenotypically, these long-term activated T cells could be identified as
effector/memory T cells. The expression of the proforms of the CD95 receptor initiator caspases, caspase-8 and -10, and the effector
caspase-3 was strongly decreased in these cells, and only active caspase fragments were detected. In contrast to short-term
activated T cells, constitutive CD95 receptor clustering was observed on the cell surface, and caspase-8 was bound to the CD95
receptor in the absence of receptor triggering. After further cross-linking of CD95, additional formation of the death-inducing
signaling complex (DISC) was strongly impaired. Reduced DISC formation in long-term activated T cells was associated with the
loss of PTEN expression and the increased phosphorylation of protein kinase B. Inhibitors of phosphoinositol 3-kinase restored
CD95 sensitivity and DISC formation in long-term activated T cells. These data suggest that defective CD95 signaling in effector/
memory T cells may contribute to the apoptosis resistance toward physiological stimuli in T cells mediating tissue destruction in
vivo. The Journal of Immunology, 2003, 171: 1172–1182.
phase, in which pathogens and infected cells are removed and fi-
nally the immune response is shut down by the deletion of acti-
vated cells. Only a few Ag-specific memory cells are maintained to
ensure an immediate and effective response after repeated Ag chal-
lenge. This balance between the expansion of effector cells and their
subsequent elimination is controlled by apoptosis mechanisms.
Apoptosis, as a main regulator of cellular homeostasis, mediates
the deletion of autoreactive T cells during thymic selection (1) and
eliminates activated T cells in the periphery during the termination
of an immune response (2). This is reflected by changes in sensi-
tivity to CD95-mediated cell death. While naive T cells are insen-
sitive to CD95-induced apoptosis, activated T cells become sen-
sitive after 3–6 days of stimulation (3). Upon repeated TCR
triggering, activated T cells undergo activation-induced cell death
(AICD),2which is mainly mediated via the CD95/CD95 ligand
ellular homeostasis is essential for the normal function of
the immune system. Ag exposure induces the expansion
of Ag-specific B and T cells, followed by an effector
(CD95L) system (4). This deletion process terminates the immune
response and mutations in either CD95 or CD95L lead to dysfunc-
tion of cellular homeostasis with autoimmunity (5, 6) and lym-
phoproliferation (7, 8).
Several recent reports reported that patients with different T cell-
mediated autoimmune diseases exhibit resistance toward CD95-
mediated cell death. CD95 resistance of T cell clones can be ex-
plained by either a defect in the CD95 signaling pathway or
changes in the expression of pro- and antiapoptotic molecules. In
CD95-sensitive cells, triggering of the CD95 receptor leads to tri-
merization of the receptor and recruitment of the adapter molecule
Fas-associated death domain (FADD) and of procaspase-8 and -10
into a complex called death-inducing signaling complex (DISC)
(9, 10). Subsequent self processing of the initiator caspases into
active caspase fragments induces cleavage of effector caspases
such as caspase-3 and -7, which finally execute the death program
(11). Impaired DISC assembly in CD95-resistant T cells can be
induced by either increased recruitment of c-FLICE-like inhibitory
protein (FLIP)shortto the DISC (12) or up-regulation of phospha-
tidylinositol 3?-kinase (PI3-K) activity (13). Also constitutive ac-
tivation of protein kinase B? (PKB?)/Akt prevents efficient DISC
formation (14). Studies in CD95-resistant T cell clones from pa-
tients with autoimmune diseases, however, showed that resistance
mostly correlates with the change in the expression of pro- and
antiapoptotic molecules. In CD95-resistant T cell lines from pa-
tients with multiple sclerosis, up-regulation of the apoptosis inhib-
itors FLIP and survivin was found (15, 16). Extensive studies were
also performed in T cell clones from patients with rheumatoid
arthritis. These patients have an expanded pool of CD4?CD28?T
cells, which are functionally active and persist over many years
(17). Altered responses to apoptosis-inducing signals such as anti-
CD95 ligation, induction of AICD, and growth factor withdrawal
*University Children’s Hospital and†Institute of Pathology, University of Ulm, Ulm,
Received for publication November 6, 2002. Accepted for publication May 29, 2003.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1Address correspondence and reprint requests to Dr. Klaus-Michael Debatin, Uni-
versity Children’s Hospital, Prittwitzstrasse 43, 89070 Ulm, Germany. E-mail ad-
2Abbreviations used in this paper: AICD, activation-induced cell death; CD95L,
CD95 ligand; DISC, death-inducing signaling complex; FADD, Fas-associated death
domain; FSC/SSC, forward scatter/side scatter; GrzB, granzyme B; PI3-K, phospha-
tidylinositol 3-kinase; IAP, inhibitors of apoptosis; PKB?/Akt, protein kinase B; RIP,
death receptor-interacting protein; FLIP, FLICE-like inhibitory protein; PARP, poly-
(ADP-ribose)-polymerase; PIP3, phosphatidylinositol-3,4,5,-triphosphate; XIAP, X-
The Journal of Immunology
Copyright © 2003 by The American Association of Immunologists, Inc. 0022-1767/03/$02.00
have been described in these cells and were associated with ele-
vated Bcl-2 levels (18–20).
The establishment of autoaggressive T cell clones comprises the
development from naive T cells, triggered by autoantigen, into
effector and memory T cells. These distinct T cell subgroups can
be efficiently characterized by the expression of surface markers,
e.g., CD45RA, CD45RO, CD28, CD27, CD62L, CCR7, or inte-
grin family members and the intracellular expression of cytotoxic
molecules such as perforin, granzymes, and CD95L (21, 22).
While naive T cells express costimulatory molecules CD27 and
CD28 and are CD45RAhigh, CCR7?, and CD62L?, effector cells
exhibit a decrease in CD27, CD28, and CD62L expression (23),
but dramatically increased the expression of perforin, granzymes,
and CD95L (21). Loss of the chemokine receptor CCR7 and L-
selectin (CD62L) as well as CD27 and CD28 appears to charac-
terize a subset of memory T cells with immediate effector func-
tions (effector/memory T cells) (22, 24).
In the present study we recapitulated the in vivo situation of
continuous Ag stimulation by an in vitro system in which HLA-
A1-specific T cells were weekly restimulated with Ag for ?12 wk.
As expected, CD95 sensitivity and AICD increased with each re-
stimulation up to the fifth round. Interestingly, after this time point
CD95 sensitivity decreased again, and T cells became constitutive
CD95 resistant and exhibited reduced AICD after eight rounds of
stimulation. In contrast to CD95-sensitive, short-term activated T
cells, these long-term activated T cells exhibited constitutive
caspase-8, -10, and -3 activation and expressed CD95 receptor in
clusters on the cell surface. In addition, caspase-8 was associated
with the CD95 receptor in the absence of receptor cross-linking.
Further cross-linking did not lead to the formation of a functional
CD95-DISC, possibly induced by down-regulation of PTEN, ac-
tivation of PI3-K and constitutive PKB?/Akt phosphorylation
These findings might help to identify novel approaches to sensitize
apoptosis-resistant T cells in vivo.
Materials and Methods
All cell lines were grown in RPMI 1640 medium (Life Technologies, Pais-
ley, U.K.) supplemented with 10% heat-inactivated FCS (Biochrom, Ber-
lin, Germany), 2 mM L-glutamine, and 1 mM sodium pyruvate at 37°C in
a humidified atmosphere containing 6.5% CO2. The HLA-A1-expressing
lymphoblastoid cell lines C1R.A1 (25) and 721 (26) and the HLA-A1?cell
line C1R (27) as well as the human T cell line H9 were used.
Generation of short- and long-term activated alloreactive T cell
Ficoll-Hypaque-separated PBMC from healthy HLA-A1?donors (1 ?
106/ml) were incubated with mitomycin C-treated HLA-A1?721 stimu-
lator cells (1 ? 105/ml) and human rIL-2 (Biochrom; 30 U/ml; first stim-
ulation). After 1 wk viable cells were separated by Ficoll-Hypaque gradient
and restimulated with mitomycin C-treated 721 cells at a ratio of 10/1 in
medium containing 30 U/ml rIL-2 (second stimulation). Separation of vi-
able cells and restimulation were repeated weekly. Short-term activated
CTL presented T cells between the second and fifth stimulations, while
long term-activated CTL were stimulated at least eight times.
Lymphocyte isolation and biomagnetic separation
PBMC were obtained from peripheral blood of healthy donors. Isolation
was done by density centrifugation of blood on Ficoll (Amersham Phar-
macia Biotech, Uppsala, Sweden). CD4?and CD8?T cells were isolated
from PBMC by depletion of monocytes by plastic adherence for 2 h at
37°C. Nonadherent cells were stained with supernatants from hybridomas
A9 (anti-CD16; provided by M. Pfreundschuh, Homburg/Saar, Germany),
HD37 (anti-CD19) (28), and HP2/6 (anti-CD4) (29) to obtain CD8?cells
or OKT8 (anti-CD8) (30) to isolate CD4?T cells. Depletion of T cells was
performed with BioMag goat anti-mouse IgG Beads (PAESEL?LOREI,
Hanau, Germany). The purity of the population was determined by flow
cytometry with CD4-FITC and CD8-FITC mAb on a FACScan cytometer
(BD Biosciences, Heidelberg, Germany).
Target cells (2 ? 106) were labeled with 200 ?Ci of Na51CrO4(Amer-
sham-Buchler, Braunschweig, Germany) for 1 h. Increasing numbers of
effector cells were titrated to 5 ? 103target cells and incubated for 4 h at
37°C. Fifty microliters of supernatant was assayed for51Cr release in a Top
CountNXT counter (Packard BioScience, Dreieich, Germany). Maximum
release was determined by incubation of target cells in 100 ?l of 10% SDS,
and spontaneous release was determined by addition of medium. The per-
centage of specific release was calculated as % specific release ? (exper-
imental release ? spontaneous release)/(maximum release ? spontaneous
release) ? 100. Assays were set up in triplicate. Cytotoxicity assays were
always performed on day 6 after the last T cell stimulation.
Induction and analysis of apoptosis
To determine CD95 sensitivity, cells were treated with either Apo-1 IgG3
Ab (anti-CD95) (31) or soluble recombinant human CD95L (Alexis, Ger-
many). Cell death was determined by measuring forward/side scatter (FSC/
SSC) or by annexin V-FITC staining to externalized phosphatidylserine
(Annexin V-FITC Kit; Bender Med Systems, Vienna, Austria) on a
FACScan cytometer (BD Biosciences). To analyze AICD, T cells (2 ?
105/ml) were cultured in triplicate for 24 h on 48-well plates coated with
OKT3 (10 ?g/ml; anti-CD3 ?-chain, obtained from American Type Culture
Collection, Manassas, VA) in the presence of rIL-2 (30 U/ml), and cell
death was determined by FSC/SSC measurement. ZVAD-fmk (Bachem,
Heidelberg, Germany) and APO-1 Fab were used to inhibit AICD. Specific
apoptosis was calculated according to the formula: 100 ? [experimental
cell death (%) ? spontaneous cell death (%)]/[100 ? spontaneous cell
death (%)]. Data always represent the mean of triplicate determinations.
To determine CD95 surface expression, 5 ? 105T cells were double
stained with CD95-FITC (31) and CD4-PE or CD8-PE (BD Bioscience,
Heidelberg, Germany). To analyze surface marker expression of short- and
long-term activated T cells, cells were double stained with CD27-PE,
CD56-PE, CD62L-PE (BD Bioscience, Heidelberg, Germany), CD28-PE,
CD69-PE (DAKO, Hamburg, Germany), CD45RO-PE (Immunotech, Mar-
seilles, France), and CD4-FITC (29) or CD8-FITC (30). For the expression
of CCR7, cells were stained with CCR7 (R&D Systems, Wiesbaden, Ger-
many), followed by goat anti-mouse IgG (H?L) F(ab?)2FITC (Dianova,
Hamburg, Germany) and were counterstained with either CD4-PE or
CD8-PE (BD Bioscience). To exclude dead cells, cells were always coun-
terstained with 1 ?g/ml propidium iodide (Sigma-Aldrich, Steinheim, Ger-
many), and surface marker expression was only measured on the propidium
iodide-negative population. Flow cytometry was performed on a FACScan
Western blot analysis and DISC assays
For Western blot analysis, 5 ? 107cells were lysed for 15 min at 4°C in
lysis buffer (30 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100,
10% glycerol, 1 mM PMSF, and 1 ?M DTT), followed by high speed
centrifugation. Twenty micrograms of lysate was separated on a 10–20%
gradient SDS page and electroblotted onto Hybond ECL nitrocellulose
membrane (Amersham-Buchler). Membranes were blocked for 1 h in PBS
supplemented with 5% milk powder and 0.1% Tween 20. Membranes were
stained for 2 h (mouse mAb) or overnight (rabbit and goat polyclonal IgG)
with the first Ab, followed by 1-h incubation with the HRP-conjugated
second Ab, and detection was performed by ECL (Amersham Bioscience,
For DISC analysis 1 ? 107/ml T cells were treated with cross-linking
CD95 mAb APO-1 (IgG3; 1 ?g/ml) (31) for 10 min at 37°C. Control cells
were incubated at 37°C in the absence of CD95 Ab. Cells were once
washed with ice-cold PBS and lysed in lysis buffer (1% Triton X-100, 150
mM NaCl, and 50 mM Tris-HCl (pH 8)), followed by high speed centrif-
ugation for 10 min at 4°C. APO-1 (1 ?g/ml) was added to the lysates of the
control cells. To precipitate CD95, 10 ?l of Pan mouse IgG Dynabeads
(Dynal Biotech, Hamburg, Germany) were added and incubated for 4 h at
4°C. Beads were then washed four times with 1 ml of washing buffer (1%
IGEPAL CA-630, 500 mM NaCl, and 50 mM Tris-HCl (pH 8)) and once
with 25 mM Tris-HCl (pH 7.5). Beads were resuspended in 6? SDS-
reducing sample buffer, boiled for 5 min at 95°C, and separated by Dynal
Magnetic Particle Concentrators, and the supernatant was separated by
10–20% gradient SDS page, followed by Western blotting as described
above. To analyze DISC formation after wortmannin treatment, T cells
1173 The Journal of Immunology
were incubated for 3.5 h in the presence of wortmannin (10 ?M; Sigma-
Aldrich) and subsequently analyzed for DISC formation as described
For Western blotting the following Abs were used: caspase-8 (clone
12F5; Alexis, Grunberg, Germany), caspase-10 (clone 4C1; MBL, Gottin-
gen, Germany), perforin (clone KM585(P1–8); Kamiya Biomedical, Seat-
tle, WA), granzyme B (GrzB; clone 2C5/F5; Serotec, Eching, Germany),
?-actin (clone AC-15; Sigma-Aldrich), lamin B (Ab-1; Oncogene, Bad
Soden, Germany), phospho-Akt (Ser473; New England Biolabs, Frankfurt,
Germany), caspase-3 (clone 19), caspase-7 (clone 51), caspase-9 (poly-
clonal rabbit), FADD (clone 1), death receptor-interacting protein (RIP)
(clone G322-2), Bcl-2 (clone Bcl-2/100), Bcl-x (clone 2H12), CD95L
(clone G247-4), TRAIL (B35-1), DFF45 (clone 19), PKB?/Akt (clone 55;
BD Transduction Laboratories, Heidelberg, Germany), c-inhibitor of apo-
ptosis (c-IAP2; H-85), Bak (N-20), poly-(ADP-ribose)-polymerase
(PARP) (F-2), FAS (C20), PTEN (A2B1), goat anti-mouse IgG-HRP, goat
anti-rabbit IgG-HRP, anti-goat IgG-HRP (Santa Cruz Biotechnology, Eu-
rope), X-linked IAP (XIAP) (MAB822), cIAP-1 (AF8189), survivin
(AF886), TNF-? (MAB610), Bid (AF846), Bad (AF819), and Bax (2282-
MC-100; R&D Systems, Wiesbaden, Germany). FLIP-Ab was provided by
P. Krammer (Heidelberg, Germany). HRP-conjugated goat anti-mouse
IgG, goat anti-rabbit IgG, and anti-goat IgG were obtained from Santa Cruz
Caspase-3 and -8 assay
To detect active caspase-3 and -8 in nonlysed cells, 5 ? 105T cells were
incubated in the absence or the presence of the caspase substrate DEVD-
R110 (caspase-3 specific; 5 ?M; Roche, Mannheim, Germany) or IETD-
Afc (caspase-8 specific; BioVision, Palo Alto, CA) in PBS on 96-well
plates and incubated at 37°C. Medium without cells was used as a control.
After 1 h, free R110 or Afc was measured fluorometrically at ? ? 535 nm
or ? ? 505 nm on a microplate fluorescence reader (1420 Victor Multilabel
Counter; Wallac, Rodgau-Jugesheim, Germany). Experiments were set up
in triplicate. X-increase is the mean emission DEVD-R110/mean
T cells were cultured to adherence in collagen I-coated culture slides (Fal-
con, Heidelberg, Germany) overnight and fixed with methanol for 5 min.
Cells were stained with FITC-labeled APO-1 for 1 h at room temperature
and embedded with fluorescent mounting medium HC08 (Calbiochem-No-
vabiochem, Schwalbach, Germany). APO-1 binding was monitored by
confocal laser scanning microscopy (Leica TCS; Leica Microsystems,
Continuous Ag stimulation induced CD95-resitant, long-term
activated T cells of the effector/memory phenotype
Changes in the apoptosis sensitivity of T cells characterize the
course of an immune response. While naive T cells are CD95
resistant, activated T cells become CD95 sensitive after 3–6 days
of stimulation. To mimic the course of an immune response in
vitro, we established HLA-A1-specific T cells, weekly restimu-
lated with the alloantigen HLA-A1-expressing lymphoblastoid cell
line 721. CD95 sensitivity was monitored after each round of stim-
ulation (Fig. 1A). T cells were incubated with agonist CD95 Ab
APO-1 (100 ng/ml) on day 5 after each stimulation, and apoptosis
was measured after 24 h in the CD4?and CD8?T cell population.
CD95 sensitivity developed between the second and fifth rounds of
stimulation. These CD95-sensitive T cells (two to five rounds of
activation) are called short-term activated T cells. After the fifth
stimulation, however, apoptosis sensitivity decreased again, and T
cells became progressively CD95 resistant. CD95-resistant T cells
obtained after the eight stimulation are designated long-term acti-
vated T cells. Staurosporine mediating apoptosis via mitochondrial
death pathways equally induced apoptosis in long- and short-term
activated T cells (data not shown). Apoptosis resistance was not
due to the loss or down-regulation of CD95 expression on the
surface of CD4?and CD8?T cells (Fig. 1B). To verify that T cells
are dying by apoptosis, we compared two different methods to
determine programmed cell death. Apoptosis induction after CD95
treatment was measured either by changes in cell size (FSC/SSC)
after different rounds of stimulation. After 24 h, cells were double stained for CD4 and CD8 expression, and cell death was determined in each subpopulation
by measuring FSC/SSC. Values are the mean of triplicate determinations. B, T cells after the first, fifth, and ninth stimulations were double stained for CD4
or CD8 and CD95 expression (solid line). Dotted lines showed staining of an isotype-matched control Ab. C, T cells after the 5th and 10th stimulations
were treated with APO-1 (100 ng/ml) for 24 h. Specific apoptosis was determined by measuring FSC/SSC or binding of annexin V-FITC to externalized
phosphatidylserine. Values are the mean of triplicate determinations. D, T cells after the first, fifth, and ninth stimulations were used as effector cells in a
standard chromium release assay. The percentage of CD8?T cells increased from 35% (first stimulation) to 60% (fifth stimulation) to 95% (eighth
stimulation). The HLA-A1-expressing C1R. A1 cell line and the untransfected control line C1R were used as51Cr-labeled target cells. Each experiment
was performed at least three times.
Continuous Ag stimulation induced CD95-resistant, long-term activated T cells. A, T cells were treated with ?-CD95 Ab APO-1 (100 ng/ml)
1174 CD95 RESISTANCE OF LONG-TERM ACTIVATED T CELLS
or by measurement of externalized phosphatidylserine on the
membrane (annexin V). Fig. 1C indicated that both methods are
comparable, and therefore apoptosis was determined in the following
experiments by analyzing FSC/SSC. All experiments were performed
in the presence of exogenously added rIL-2. As IL-2 withdrawal in-
duced death in short- and long-term activated T cells (data not
shown), we could not investigate the influence of IL-2 on CD95 sen-
sitivity in our system. Short- and long-term activated T cells are func-
tional cytotoxic T cells because they specifically lysed the HLA-A1-
expressing target cell line C1R.A1 and not the non-HLA-A1-
expressing parental cell line C1R. Only T cells at the beginning of the
stimulation process did not exhibit HLA-A1 specificity (Fig. 1D). To
further characterize CD95-resistant, long-term activated T cells, the
expression of surface markers used to dissect naive T cells from ef-
fector and memory T cells was analyzed (Table I). Long-term acti-
vated CD4?and CD8?T cells are devoid of CD28, CD62L, and
CCR7 expression and displayed the phenotype of effector/memory
cells. In addition, CD8?T cells strongly up-regulated CD56 as a
marker of increased cytolytic capacity (32). These results demonstrate
that continuous Ag stimulation induced the development of a CD95-
resistant T cell population of the effector/memory phenotype.
AICD in long-term activated T cells is decreased, and CD95
and caspase independent
AICD mediated via the CD95/CD95L system serves as the major
mechanism to remove activated T cells from the periphery during
the termination of an immune response. AICD can be induced in
T cells by TCR/CD3 triggering via immobilized mAb OKT3. T
cells after the 5th and 10th stimulations were incubated in the
presence or the absence of mAb OKT3. Simultaneously, the broad
caspase inhibitor ZVAD-fmk or APO-1 Fab, preventing the CD95-
CD95L interaction, was added. After 24 h cell death was deter-
mined in the CD8?T cell populations (Fig. 2). Long-term acti-
vated CD8?T cells exhibited decreased AICD compared with
their short-term activated counterparts. Interestingly, long-term ac-
tivated CD8?T cells were barely protected from AICD by ZVAD-
fmk or APO-1 Fab, indicating that AICD in these cells proceeds
through a caspase- and CD95/CD95L-independent mechanism.
The decline in AICD varied for each experiment performed and
ranged between 10 and 50%. As a control for AICD induction and
the effects of inhibitors we used the human T cell line H9, which
was efficiently protected from AICD in the presence of ZVAD-fmk
and APO-1 Fab. These data indicate that CD95 resistance of long-
term activated T cells correlates with a caspase- and CD95-inde-
pendent decrease in AICD.
Expression of pro- and antiapoptotic molecules in short- vs
long-term activated T cells
Differences in the expression of pro- and antiapoptotic molecules
in long-term activated T cells could explain CD95 resistance com-
pared with their short-term activated counterparts. The expression
of IAP and members of the Bcl-2 family was analyzed by Western
blot analysis in T cells after the first, fifth, and ninth rounds of
stimulation. No difference in FLIP, XIAP, and CIAP-1 and -2 ex-
pression was detected. Survivin, which is weakly expressed after
the first stimulation, was up-regulated after the fifth stimulation,
but protein expression did not further increase in long-term acti-
vated T cells (Fig. 3A). Bcl-2 was also up-regulated after the first
stimulation, but expression levels did not differ in short- and long-
term activated T cells. Bcl-xL, however, was strongly up-regulated
after the first round of stimulation in CD95-sensitive cells, with a
slight increase in long-term activated, CD95-resistant T cells. With
progressive Bcl-xLinduction, the proapoptotic molecules Bak and
Bad were down-regulated, while the Bax expression level was un-
changed. Bid, which links the receptor-mediated apoptosis path-
way to the mitochondrial pathway, disappeared in the beginning of
T cell stimulation, but was not differently expressed between short-
and long-term activated T cells (Fig. 3B). T cells can kill target
cells either by death-inducing ligands such as CD95L, TRAIL, and
TNF-? or via the granzyme/perforin system. During the stimula-
tion process, CD95L, GrzB, and perforin were strongly up-regu-
lated, indicating that the cytotoxic capacity increased during re-
peated stimulation (Fig. 3, C and D). An increase in CD95L
were incubated with immobilized OKT3 (anti-CD3; 10 ?g/ml) in the ab-
sence or the presence of ZVAD-fmk (100 ?M) or APO-1 Fab (50 ?g/ml).
After 24 h, cells were stained with CD8-PE, and cell death was determined
in the CD8?fraction by FSC/SSC. The human T cell line H9 served as a
positive control for CD3-mediated killing and functionality of the inhibi-
tors. Values are the mean of triplicate determinations of one experiment of
three independent ones performed.
Decreased AICD in long-term activated T cells. T cells
Table I. Expression of surface markers used to dissect naive T cells from effector and memory T cellsa
% Positive Cells
CD28 CD45ROCD69 CD27 CD62LCD56 CCR7
aT cells after different rounds of stimulation or without any stimulation (0 stimulation) were double stained for CD4 or CD8
expression and the indicated surface markers and analyzed by flow cytometry. Numbers indicate the percentage of positively
1175 The Journal of Immunology
expression was also detected by intracellular staining (data not
shown). These data suggest that long-term activated T cells might
have an enhanced survival capacity by increased expression of
Bcl-xLand diminished expression of Bak and Bad.
Constitutive caspase activation in long-term activated T cells
Caspases are the main mediators of cell death after CD95 trigger-
ing, mitochondrial dysfunction, and granzyme-induced cell death.
During death induction the caspase proform is converted into its
active caspase fragment. The expression of procaspases and their
active cleavage fragments in HLA-A1-specific T cells after the
first, fifth, and ninth stimulations was analyzed by Western blot
analysis (Fig. 4A). Surprisingly, long-term activated T cells did not
express the proforms of caspase-3, -8, and -10. While caspase-8
and -3 were completely converted into their intermediate cleavage
products (p43/41 for caspase-8) and the active fragments (p20/18
for caspase-8 and p12 for caspase-3), active caspase-10 fragments
could not be detected due to Ab specificities. In contrast, short-
term activated T cells expressed procaspase-8 (55/54 kDa), the
intermediate cleavage products p43/41, and only low amounts of
the active fragment p20/18. Caspase-10 was not processed, and
partial processing of caspase-3 was found. Constitutive activation
of caspase-9 was not detected in long-term activated T cells.
Caspase-7 was processed in all T cells independently of the stim-
ulation status. Expression levels of FADD, the adapter molecule
between CD95 and caspase-8 and -10 at the DISC, were un-
changed at all stages of stimulation. Down-regulation of RIP,
which was recently reported to mediate CD95-induced T cell death
(33), could not consistently be detected. Although long-term acti-
vated T cells expressed active caspases, no differences in the cleav-
age of caspase substrates such as PARP, DFF45, and lamin B,
were detected between the different stages of T cell stimulation,
indicating that cellular substrates were not fully cleaved by the
active caspase fragments. To rule out that caspase activation in
long-term activated T cells was mediated via GrzB released from
cytolytic granule during protein preparation, we measured
caspase-3 and -8 activities fluorometrically in non-lysed T cells
(Fig. 4B). Medium and isolated nonstimulated CD4?and CD8?T
cells from healthy donors served as a negative control. While un-
stimulated T did not process the caspase-3 substrate DEVD-R110,
T cells after the fourth stimulation showed enhanced capase-3 ac-
tivity, which was further increased in T cells after the 12th stim-
ulation. Caspase-8 activity in long-term activated T cells was less
pronounced, possibly reflecting a decreased diffusion of the pep-
tide into the cells. Considering the CD95 resistance of long-term
activated T cells, these data indicate that constitutive caspase ac-
tivation might be associated with apoptosis resistance.
Short-term activated CD4?and CD8?T cells do not differ in
their protein expression pattern
Our in vitro system of repeated Ag stimulation led to a predomi-
nant expansion of CD8?T cells. After five rounds of stimulation,
B and NK cells disappeared, and the ratio of CD4?/CD8?T cells
varied from 1/1 to 1/5. Further stimulation increased the percent-
age of CD8?T cells up to 85–98%. We did not observe any re-
lationship between the composition of lymphoid cells in the start-
ing culture and the acquisition of CD95 resistance in long-term
activated T cells (data not shown). To rule out that some of the
observed differences in apoptosis sensitivity were attributed to the
significant amount of CD4?T cells present in the initial culture,
we isolated both subpopulations at the beginning of the stimulation
and restimulated them separately. Fig. 5A shows that proforms and
active fragments of caspase-3, -8, -9, and -10 were comparably
expressed in both T cell subsets. No differences were found for
PARP, RIP, and FADD expression. Bcl-xL, Bcl-2, survivin,
regulate the proapoptotic molecules Bak and Bad. After the first, fifth, and
ninth stimulations, 20 ?g of protein lysates of T cells were subjected to
Western blot analysis. Expression of inhibitors of apoptosis (A), pro- and
antiapoptotic molecules of the Bcl-2 family (B), death-inducing ligands
(C), and GrzB and perforin (D) was determined. Expression of ?-actin
served as the loading control. The percentage of CD8?T cells increased
from 20% (first stimulation) and 40% (fifth stimulation) to 90% (ninth
stimulation). The experiment shown is representative for T cells from one
donor of two analyzed.
Long-term activated T cells up-regulate Bcl-xLand down-
cells. A, Cell lysate (40 ?g) of T cells after different rounds of stimulation
was used for Western blot analysis to detect the expression of procaspases
and their cleavage products. Arrows indicate the m.w. of procaspases and
cleavage products. To detect the expression of FADD, RIP, PARP, lamin
B, and DFF45, 20 ?g of cell lysate was loaded. ?-Actin served as the
loading control. Blots shown are representative of two different experi-
ments. B, Medium, nonstimulated, isolated CD4?and CD8?T cells from
healthy donors and T cells after the 4th, 9th, and 12th stimulations were
incubated with caspase-3 substrate DEVD-R110 or the caspase-8 substrate
IETD-Afc, and caspase-3 and -8 activity was determined fluorometrically.
Values are the mean of triplicate determinations, and the experiment was
Constitutive caspase activation in long-term activated T
1176 CD95 RESISTANCE OF LONG-TERM ACTIVATED T CELLS
CD95L, and GrzB were up-regulated after the first stimulation in
both T cell subsets, and Bid was down-regulated. Perforin, how-
ever, strongly expressed in CD8?T cells after the fifth stimulation
was nearly absent in the CD4?T cell population (Fig. 5B), con-
sistent with the observation that short-term activated CD4?T cells
do not lyse HLA-A1-expressing target cells (data not shown).
These data indicate that short-term activated CD4?and CD8?T
cells do not differ in the expression of caspases and apoptosis reg-
ulators. For additional experiments short-term activated T cells
were considered a unique cell population.
Analysis of the CD95 signaling complex
As long-term activated T cells exhibited an impaired CD95 sig-
naling pathway, we analyzed CD95 receptor expression by confo-
cal microscopy on isolated short-term activated CD8?T cells and
their long-term activated counterparts (Fig. 6A). While short-term
activated T cells displayed a regular distribution of the CD95 mol-
ecules on the cell surface with minimal CD95 clustering, extensive
CD95 receptor clustering was found in long-term activated T cells.
CD95 clusters are completely absent during the early phase of the
activation process (first and second stimulations; data not shown).
Although Fig. 1B showed that CD95 expression on short and long
term activated T cells is similar, Fig. 6A might indicate that long-
term activated T cells express lower levels of CD95 than their
short-term activated counterparts. To exclude that CD95 is down-
regulated on apoptosis-resistant cells, we incubated T cells of the
third and ninth stimulations with soluble CD95L for 24 h and
measured CD95 expression on the resistant CD8?population. No
difference in CD95 expression was detected (Fig. 6B), indicating
that clustering of the CD95 receptor might cause quenching of the
fluorescence intensity by confocal microscopic analysis, leading to
relatively less intense staining of clustered CD95 receptors.
Since long-term activated T cells expressed processed caspase-8
and -10 and clustered CD95, we next analyzed whether formation
of the CD95-DISC was intact (Fig. 7A). T cells were incubated
with the cross-linking agonistic CD95 Ab anti-APO-1 or left un-
treated. After cell lysis, CD95 was precipitated, and subsequently
Western blot analysis was performed to determine compounds
bound to the receptor. In the absence of CD95 cross-linking, no
association between caspase-10 and the CD95 receptor was de-
tected in short-term activated T cells, and only small amounts of
caspase-8 were bound to CD95. Long-term activated T cells, how-
ever, displayed a strong constitutive association between CD95
and caspase-8. Also, small amounts of caspase-10 were bound to
the receptor in the absence of cross-linking. After CD95 triggering,
short-term activated T cells recruited caspase-8, caspase-10, and
FADD to the DISC. Recruitment of caspase-8 and -10 and FADD,
however, was impaired in long-term activated T cells. We did not
observe differences in FLIPshortor processed FLIPlong(p43) re-
cruitment in short- vs long-term activated T cells. Unprocessed
FLIPLwas not found in the DISC. RIP, however, was not present
in the DISC of long-term activated CD8?T cells independently of
anti-APO-1 treatment. In addition, full-length procaspase-8 (55/54
kDa) was hardly detectable in any of the DISC assays performed,
showing that even in short-term activated T cells caspase-8 is rap-
idly converted into the intermediate cleavage product p43/41 after
receptor triggering. As short-term activated T cells after the fifth
stimulation presented a mixture of CD4?and CD8?T cells, DISC
formation of both subsets, isolated via biomagnetic cell separation
before the experiment, were analyzed (Fig. 7B). Differences in
isolated CD4?and CD8?T cells. CD4?and CD8?T cells were isolated
by biomagnetic sorting from PBL of an HLA-A1-negative donor. Each
subpopulation was separately stimulated up to five times. Proteins were
isolated after the first and fifth stimulations and were subjected to Western
blot analysis. A, Expression of caspases, PARP, RIP, and FADD. B, Ex-
pression of pro- and antiapoptotic molecules, PTEN, CD95L, perforin,
GrzB, and the loading control ?-actin.
Comparison of protein expression in short-term activated,
activated T cells. A, CD8?short- and long-term activated T cells were
grown on collagen-coated culture slides overnight. Methanol-fixed slides
were stained with APO-1-FITC, and the fluorescence images were ana-
lyzed by confocal laser scanning microscopy. Magnification: a, ?630; b,
?2400. B, T cells after the third and ninth stimulations were incubated with
soluble CD95L (200 ng/ml). After 24 h, cells were double stained for
CD95 and CD8, and apoptosis-resistant cells were gated via FSC/SSC. The
expression of CD95 or isotype control (dotted lines) was analyzed on the
Expression of CD95 clusters on the surface of long-term
1177 The Journal of Immunology
PKB?/Akt phosphorylation. DISC assays were always performed on day 6 after the last stimulation. DISC analysis was performed after cross-linking of
the CD95 receptor by APO-1 Ab (?APO-1) or in the absence of the Ab (?APO-1). After CD95 precipitation, proteins were subjected to Western blot
analysis, and association of caspase-8 and -10, FADD, FLIP, and RIP to the CD95 receptor was determined. CD95 expression served as the loading control.
A, DISC formation of T cells after the 5th and 10th stimulations. Ninety-five percent of T cells were CD8?after the 10th stimulation, and 70% were CD8?
after the fifth stimulation. B, DISC formation of isolated CD4?and CD8?T cells after the third stimulation. C, Proteins isolated from T cells of two donors
after different rounds of stimulation were subjected to Western blot analysis, and the expression of PTEN, PKB?/Akt, and phospho-Akt (Ser473; pAkt) was
analyzed. The percentage of CD8?long-term activated T cells was ?95%, while 50–60% of CD8?T cells were present in short-term activated T cells.
D, Short- and long-term activated T cells after the fourth and ninth stimulations were incubated with APO-1 (100 ng/ml) or medium in the absence or the
presence of wortmannin (10 ?M). After 24 h cells were stained for CD8 expression, and the percent specific apoptosis was calculated in the CD8?T cell
population using FSC/SSC analysis. Data are representative of three different experiments performed with T cells from different donors. E, DISC formation
of T cells after the fourth and ninth stimulations in the presence or the absence of wortmannin (10 ?M). The blots shown are representative of at least three
independent experiments with T cells from different donors.
Impaired DISC formation in long-term activated T cells is associated with the loss of PTEN expression and increased PI3-K activity, and
1178CD95 RESISTANCE OF LONG-TERM ACTIVATED T CELLS
DISC formation after CD95 triggering were not observed in both
subsets. In all assays CD95 served as a loading control. Taken
together, CD95 resistance of long-term activated T cells might be
due to a defective CD95 signaling pathway, including constitutive
caspase activation, CD95 cluster formation, and impaired DISC
Down-regulation of PTEN expression, elevated PI3-K activity,
and increased phosphorylation of PKB?/Akt in long-term
activated T cells
Activation of PKB?/Akt has been shown to mediate the inhibition
of CD95-DISC formation and apoptosis in mouse T cells. Hyper-
phosphorylation and activation of PKB?/Akt may be conse-
quences of an increase in the intracellular phospholipid phospha-
tidylinositol-3,4,5-triphosphate (PIP3) (14, 34). PIP3is generated
by PI3-K (35) and is negatively regulated by the lipid phosphatase
PTEN, which cleaves the D3 phosphate from PIP3(36). Since
long-term activated T cells exhibited a CD95-resistant phenotype
with defective DISC assembly, we analyzed the expression of
PTEN and phosphorylated PKB?/Akt in short- vs long-term acti-
vated T cells from two donors (Fig. 7C). PTEN was strongly
down-regulated, and elevated levels of phosphorylated PKB?/Akt
were detected in long-term activated T cells from both donors.
PTEN down-regulation in long-term activated T cells was not due
to the significant amount of CD4?T cells still present after early
stimulations, as PTEN expression did not vary in the isolated
CD4?and CD8?short-term activated subpopulations (Fig. 5B).
The amount of unphosphorylated inactive PKB?/Akt was unal-
tered between the different stages of T cell stimulation. Activation
of PI3-K also leads to increased phosphorylation of PKB?/Akt.
Short- and long-term activated T were incubated in the presence of
the PI3-K inhibitor wortmannin, and CD95 sensitivity and DISC
formation were analyzed. After wortmannin treatment, long-term
activated T cells exhibited CD95 sensitivity comparable to that of
untreated short-term activated T cells. A modest increase in CD95
sensitivity was also observed in short-term activated T cells after
PI3-K inhibition (Fig. 7D). To further clarify whether inhibition of
PI3-K restored caspase-8 and-10 recruitment to the DISC, we com-
pared DISC formation of short- and long-term activated T cells in
the presence or the absence of wortmannin. Caspase-10 recruit-
ment to the DISC was increased in short- and long-term activated
T cells, while caspase-8 association to the DISC was only mar-
ginally increased in long-term activated T cells (Fig. 7E). These
results suggest that impaired DISC formation of long-term acti-
vated T cells might be due to the loss of PTEN expression and the
increased activation of PI3-K inducing constitutive activation of
PKB?/AKT and preventing the formation of a functional DISC.
Thymic selection (central tolerance) prevents the development of
autoreactive lymphocytes leading to chronic tissue damage. This
elimination process, however, is often incomplete, and peripheral
regulatory mechanisms, such as anergy, ignorance, suppression, or
apoptosis (37–39) are needed to control mature self-reactive cells.
Activated T cells are deleted from the periphery by either CD95-
mediated apoptosis or AICD. Both mechanisms are affected by
alterations in CD95 sensitivity during the course of an immune
response. According to current concepts, naive T cells are CD95
resistant, whereas activated T cells acquire CD95 sensitivity after
3–6 days of activation (3). AICD, mainly mediated via the CD95/
CD95L system, terminates the immune response after repeated
TCR triggering (4). Molecular mechanisms of apoptosis defects,
however, leading to uncontrolled expansion and prolonged sur-
vival of self-reactive T cells, are poorly understood.
To mimic the in vivo situation of an immune response after
continuous Ag stimulation we established an in vitro system of
HLA-A1-specific T cells that were weekly restimulated with HLA-
A1-expressing cells. Short-term activated T cells (stimulated for
2–5 wk) were CD95 sensitive and susceptible to AICD. Surpris-
ingly, long-term activated T cells (?8 wk of stimulation) acquired
constitutive CD95 resistance and exhibited decreased AICD.
AICD induction could not be inhibited by caspase inhibitors or
CD95 Fab that prevent the interaction between CD95 and its li-
gand. Inhibition of perforin/GrzB (40, 41) and TRAIL (42) activ-
ity, described as further mediators of AICD in human T cells,
could also not prevent T cell death (data not shown). Phenotypi-
cally, long-term activated T cells were identified as effector/mem-
ory T cells, characterized by the loss of CD62L and CCR7 ex-
pression and their immediate effector function (24). IL-5 and
IFN-? production decreased in long-term activated T cells, while
the amount of the immunomodulatory cytokine IL-10, which pro-
motes proliferation and differentiation into effective cytotoxic T
cells (43, 44), was significantly increased, underlining their effi-
cient cytotoxic capacity (data not shown). In agreement with other
studies (12, 13), only a subpopulation of activated T cells in the
early stimulation process (second to fifth stimulations) was CD95
sensitive. CD95 sensitivity varied in T cells from different donors
from 30–70%, indicating that the total T cell population always
contained CD95-resistant T cells. Changes in CD95 sensitivity
might reflect the expansion of CD95-resistant cells or indicate in-
dividual changes from CD95-sensitive cells into CD95-resistant
cells. The final result of continuous Ag stimulation, however, is the
development of CD95-resistant T cells.
Apoptosis resistance of T cells was described in different auto-
immune diseases, such as rheumatoid arthritis (18), multiple scle-
rosis (45), and autoimmune hematologic cytopenias (46). CD95-
resistant T cell clones from patients with multiple sclerosis
exhibited up-regulation of apoptosis inhibitors survivin or FLIP
(15, 16). Patients with rheumatoid arthritis have an expanded pool
of long-lived, functional CD4?CD28?T cells that are insensitive
to apoptosis signals such as CD95 ligation, AICD, and growth
factor withdrawal and express elevated levels of the apoptosis in-
hibitors Bcl-2 and FLIP (17–19). Also, apoptosis resistance of
memory T cells specific for viral peptides and alloantigens was
linked to the transcriptional up-regulation of FLIP RNA (47). Up-
regulation of FLIP or IAP, however, was not detected in our long-
term activated T cells. Also, the differential expression of Bcl-2
family members did not show a clear association between the sen-
sitive and resistant phenotypes. Although the proapoptotic mole-
cules Bak and Bad were decreased during the development of ef-
fector/memory cells in vitro, Bax and Bcl-2 levels remained stable.
Also, Bcl-xLwas up-regulated in apoptosis-sensitive short-term
activated cells, with a further increase in apoptosis-resistant, long-
term activated T cells. Staurosporine, a mediator of mitochondrial
cell death (48), induced apoptosis in short- and long-term activated
T cells (data not shown), indicating that differences in the expres-
sion of Bcl-2 family members do not predominantly account for
apoptosis resistance, but may contribute to an enhanced survival
caspase-3, -8, and -10 was strongly decreased in long-term acti-
vated T cells of the effector/memory phenotype, and only active
caspase fragments were detected. Caspase activation in the ab-
sence of apoptotic stimuli has also been observed in TCR-triggered
human T cells (49–53). Caspase inhibitors or Fas-Fc, preventing
the interaction between CD95 receptor and endogenously pro-
duced and secreted CD95L, inhibited T cell proliferation in the
early phase of T cell stimulation, suggesting a model in which T
procaspase expression of
1179 The Journal of Immunology
cell activation via TCR up-regulates CD95L, which binds to sur-
face CD95 receptor and induces caspase activation. In our long-
term activated T cells, increased CD95L expression might also be
induced by TCR triggering through the stimulator cells, which
subsequently leads to caspase activation. Activation of T cells and
caspases, however, is always synchronized with the up-regulation
of molecules of the intrinsic death machinery, such as perforin,
GrzA, GrzB, and CD95L. To rule out that caspase activation in
long-term activated T cells is induced by the release of granzymes
after protein isolation (54) we measured internal caspase-3 and -8
activity in living cells. While caspase-3 activity was not observed
in unstimulated T cells, a 3-fold increase in activity was detected
during the development from short- into long-term activated T
cells, which was, to a lesser extent, also detected for caspase-8
activity. Although effector caspases were activated, cleavage of
nuclear targets such as lamin B or DFF45 was not found, indicat-
ing that caspase activation in long-term activated T cells does not
lead to nuclear apoptosis. Partial PARP-cleavage was already
present after the first stimulation and did not increase after further
stimulations, corresponding to published data (50, 51). Caspases,
however, may still be cleaved by granzymes that are exported from
cytolytic granules into the cytoplasm of T cells (55). In addition, it
is not clear how T cells are protected from apoptosis in the pres-
ence of active caspase fragments. Safety catch, e.g., is a regulatory
peptide, that controls the autocatalytic cleavage of procaspase-3
(56), but to our knowledge no molecules regulating the apoptosis-
inducing capacity of active caspase fragments have been described.
Since caspase-8 and -10 are initiator caspases involved in DISC
formation, we further analyzed whether functional DISC formation
occurs in the presence of constitutive activated caspases. While
short-term activated T cells formed a functional DISC after CD95
cross-linking, long-term activated T cells showed an impaired re-
cruitment of caspase-8 and -10 and FADD to the DISC. Up-reg-
ulation of c-FLIPshortand thereby reduced procaspase-8 cleavage
at the DISC have been described as being responsible for apoptosis
resistance of TCR/CD3-restimulated, short-term activated T cells
(12). However, we did not detect differences in FLIPsor processed
FLIPLrecruitment to the DISC in uninduced or CD95-triggered T
cells. This may be due to differences in the T cell populations used,
i.e., short-term activated T cells vs long-term activated effector/
memory T cells. Murine T cells, transgenic for active PKB?/Akt
or haploinsufficient for PTEN expression, a phosphatase that neg-
atively regulates PKB?/Akt phosphorylation, exhibit a CD95-re-
sistant phenotype induced by decreased recruitment of caspase-8
to the DISC (14). In our system impaired recruitment of initiator
caspases-8 and -10 in long-term activated T cells was linked to the
loss of PTEN expression and increased levels of PKB?/Akt phos-
phorylation. PKB?/Akt is phosphorylated in response to PIP3, a
second messenger generated by PI3-K. Incomplete procaspase-8
processing at the DISC was also observed in the presence of active
PI3-K (13), as shown for TCR/CD3-triggered apoptosis-resistant
Th2 cells. Inhibition of PI3-K activity restored CD95 sensitivity in
long-term activated T cells and increased the recruitment of
caspases to the DISC, indicating that active PI3-K might mediate
protection from CD95-induced cell death. How PKB?/Akt phos-
phorylation inhibits recruitment of effector caspases to the DISC,
however, is currently unknown. Phosphorylation of Bad and
caspase-9 by PKB?/Akt inhibits mitochondrial-mediated apopto-
sis (57, 58), but nonfunctional Bad or caspase-9 would not influ-
ence recruitment of FADD and initiator caspases to the CD95 re-
transcriptional factors, such as NF-?B (59, 60) or c-Myb (61), on
the up-regulation of molecules preventing the association of
caspases and FADD to the CD95 receptor. Loss of PTEN expres-
acts via regulationof
sion in mouse T cells renders them insensitive to different apopto-
sis stimuli, such as anti-CD95 treatment, IL-2, IL-2 plus serum
withdrawal, and AICD (62, 63), reflecting CD95 resistance and
decreased AICD induction in our human effector/memory T cells.
Transfection of PTEN in long-term activated T cells will show
whether CD95 sensitivity and functional DISC formation can be
Defects in the CD95 signaling pathway of long-term activated T
cells of the effector/memory phenotype could be further explained
by preassociation of caspase-8 cleavage products (p43/41) to the
CD95 receptor in the absence of receptor cross-linking and cluster
formation of the receptor on the cell surface. It is not clear, how-
ever, whether cluster formation induces continuous caspase acti-
vation and impaired DISC formation or whether preassociation of
caspase-receptor complexes leads to CD95 resistance. Possibly,
CD95 receptors on long-term activated T cells are recruited into
clusters by extracellular oligomerization domains (64, 65), induc-
ing conformational changes or continuous caspase activation and
preventing CD95 signaling. Furthermore, constitutive complex
formation of CD95 receptor with its ligand CD95L, which is
strongly up-regulated in long-term activated T cells, might explain
CD95 resistance. DISC assays, immunoprecipitations, or immu-
nofluorescence analysis, however, did not reveal any complex for-
mation between CD95L and its receptor (data not shown).
In vitro-derived, long-term activated T cells resemble self-reac-
tive T cells in patients with chronic graft-vs-host disease after bone
marrow transplantation from unrelated donors. Clinically these T
cells can only be controlled by aggressive cytotoxic therapy, but
many patients develop resistance to drugs such as corticosteroids,
cyclosporin A, or anti-thymocyte globulin, reflecting their apop-
tosis-resistant status (66, 67). Also, in vitro-generated long-term
activated T cells were less susceptible to cyclophosphamide- and
methotrexate-induced apoptosis than short-term activated T cells
(data not shown). Immunosuppressive therapy that prevents T cell
activation and induces T cell apoptosis (25) makes it difficult to
identify long-term activated T cells in patients with chronic graft-
vs-host disease in vivo. A small proportion of CD8?T cells from
three haplo-identical transplanted patients displayed the phenotype
of long-term activated T cells and showed decreased CD95 sensi-
tivity and AICD induction (data not shown). Also, in the synovial
fluid of one patient with rheumatoid arthritis, 50% of the CD8?T
cells did not express CD28, CD62L, CD27, and CCR7 and showed
up-regulation of CD56, resembling effector/memory T cells in
vivo. Further analysis of these T cells will show whether they
exhibit CD95 resistance by similar mechanisms as in vitro derived
long-term activated T cells.
In conclusion, the present data provide strong evidence that ap-
optosis resistance of long-term activated T cells is due to an im-
paired CD95 signaling pathway. Constitutive caspase activation,
continuous clustering of CD95 receptors, and impaired DISC for-
mation are associated with the CD95-resistant phenotype. Under-
standing the molecular mechanisms of internal caspase activation
and CD95 clustering in the absence of cell death can provide novel
therapeutic strategies to sensitize CD95-resistant T cells in vivo.
We thank Peter H. Krammer for kindly providing the mAb ?-FLIP (NF6).
1. Smith, C. A., G. T. Williams, R. Kingston, E. J. Jenkinson, and J. J. Owen. 1989.
Antibodies to CD3/T-cell receptor complex induce death by apoptosis in imma-
ture T cells in thymic cultures. Nature 337:181.
2. Kawabe, Y. and A. Ochi. 1991. Programmed cell death and extrathymic reduc-
tion of V?8?CD4?T cells in mice tolerant to Staphylococcus aureus entero-
toxin B. Nature 349:245.
1180 CD95 RESISTANCE OF LONG-TERM ACTIVATED T CELLS
3. Klas, C., K. M. Debatin, R. R. Jonker, and P. H. Krammer. 1993. Activation
interferes with the APO-1 pathway in mature human T cells. Int. Immunol. 5:625.
4. Dhein, J., H. Walczak, C. Baumler, K. M. Debatin, and P. H. Krammer. 1995.
Autocrine T-cell suicide mediated by APO-1/(Fas/CD95). Nature 373:438.
5. Stassi, G., and R. De Maria. 2002. Autoimmune thyroid disease: new models of
cell death in autoimmunity. Nat. Rev. Immunol. 2:195.
6. Siegel, R. M., F. K. Chan, H. J. Chun, and M. J. Lenardo. 2000. The multifaceted
role of Fas signaling in immune cell homeostasis and autoimmunity. Nat. Immu-
7. Wang, J., L. Zheng, A. Lobito, F. K. Chan, J. Dale, M. Sneller, X. Yao,
J. M. Puck, S. E. Straus, and M. J. Lenardo. 1999. Inherited human caspase 10
mutations underlie defective lymphocyte and dendritic cell apoptosis in autoim-
mune lymphoproliferative syndrome type II. Cell 98:47.
8. Bleesing, J. J., M. R. Brown, S. E. Straus, J. K. Dale, R. M. Siegel, M. Johnson,
M. J. Lenardo, J. M. Puck, and T. A. Fleisher. 2001. Immunophenotypic profiles
in families with autoimmune lymphoproliferative syndrome. Blood 98:2466.
9. Kischkel, F. C., D. A. Lawrence, A. Tinel, H. LeBlanc, A. Virmani, P. Schow,
A. Gazdar, J. Blenis, D. Arnott, and A. Ashkenazi. 2001. Death receptor recruit-
ment of endogenous caspase-10 and apoptosis initiation in the absence of
caspase-8. J. Biol. Chem. 276:46639.
10. Algeciras-Schimnich, A., L. Shen, B. C. Barnhart, A. E. Murmann,
J. K. Burkhardt, and M. E. Peter. 2002. Molecular ordering of the initial signaling
events of CD95. Mol. Cell. Biol. 22:207.
11. Yang, X., H. Y. Chang, and D. Baltimore. 1998. Autoproteolytic activation of
pro-caspases by oligomerization. Mol. Cell. 1:319.
12. Kirchhoff, S., W. W. Muller, A. Krueger, I. Schmitz, and P. H. Krammer. 2000.
TCR-mediated up-regulation of c-FLIPshort correlates with resistance toward
CD95-mediated apoptosis by blocking death-inducing signaling complex activ-
ity. J. Immunol. 165:6293.
13. Varadhachary, A. S., M. E. Peter, S. N. Perdow, P. H. Krammer, and P. Salgame.
1999. Selective up-regulation of phosphatidylinositol 3?-kinase activity in Th2
cells inhibits caspase-8 cleavage at the death-inducing complex: a mechanism for
Th2 resistance from Fas-mediated apoptosis. J. Immunol. 163:4772.
14. Jones, R. G., A. R. Elford, M. J. Parsons, L. Wu, C. M. Krawczyk, W. C. Yeh,
R. Hakem, R. Rottapel, J. R. Woodgett, and P. S. Ohashi. 2002. CD28-dependent
activation of protein kinase B/Akt blocks Fas-mediated apoptosis by preventing
death-inducing signaling complex assembly. J. Exp. Med. 196:335.
15. Semra, Y. K., O. A. Seidi, and M. K. Sharief. 2001. Overexpression of the
apoptosis inhibitor FLIP in T cells correlates with disease activity in multiple
sclerosis. J. Neuroimmunol. 113:268.
16. Sharief, M. K., and Y. K. Semra. 2001. Heightened expression of survivin in
activated T lymphocytes from patients with multiple sclerosis. J. Neuroimmunol.
17. Waase, I., C. Kayser, P. J. Carlson, J. J. Goronzy, and C. M. Weyand. 1996.
Oligoclonal T cell proliferation in patients with rheumatoid arthritis and their
unaffected siblings. Arthritis Rheum. 39:904.
18. Schirmer, M., A. N. Vallejo, C. M. Weyand, and J. J. Goronzy. 1998. Resistance
to apoptosis and elevated expression of Bcl-2 in clonally expanded CD4?CD28?
T cells from rheumatoid arthritis patients. J. Immunol. 161:1018.
19. Vallejo, A. N., M. Schirmer, C. M. Weyand, and J. J. Goronzy. 2000. Clonality
and longevity of CD4?CD28nullT cells are associated with defects in apoptotic
pathways. J. Immunol. 165:6301.
20. Namekawa, T., M. R. Snyder, J. H. Yen, B. E. Goehring, P. J. Leibson,
C. M. Weyand, and J. J. Goronzy. 2000. Killer cell activating receptors function
as costimulatory molecules on CD4?CD28nullT cells clonally expanded in rheu-
matoid arthritis. J. Immunol. 165:1138.
21. Hamann, D., P. A. Baars, M. H. Rep, B. Hooibrink, S. R. Kerkhof-Garde,
M. R. Klein, and R. A. van Lier. 1997. Phenotypic and functional separation of
memory and effector human CD8? T cells. J. Exp. Med. 186:1407.
22. Tomiyama, H., T. Matsuda, and M. Takiguchi. 2002. Differentiation of human
CD8?T cells from a memory to memory/effector phenotype. J. Immunol. 168:
23. Wills, M. R., G. Okecha, M. P. Weekes, M. K. Gandhi, P. J. Sissons, and
A. J. Carmichael. 2002. Identification of naive or antigen-experienced human
CD8?T cells by expression of costimulation and chemokine receptors: analysis
of the human cytomegalovirus-specific CD8?T cell response. J. Immunol. 168:
24. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two
subsets of memory T lymphocytes with distinct homing potentials and effector
functions. Nature 401:708.
25. Strauss, G., W. Osen, and K. M. Debatin. 2002. Induction of apoptosis and
modulation of activation and effector function in T cells by immunosuppressive
drugs. Clin. Exp. Immunol. 128:255.
26. Hochman, J. H., Y. Shimizu, R. DeMars, and M. Edidin. 1988. Specific associ-
ations of fluorescent ?-2-microglobulin with cell surfaces: the affinity of different
H-2 and HLA antigens for ?-2-microglobulin. J. Immunol. 140:2322.
27. Zemmour, J., A. M. Little, D. J. Schendel, and P. Parham. 1992. The HLA-A,B
“negative” mutant cell line C1R expresses a novel HLA-B35 allele, which also
has a point mutation in the translation initiation codon. J. Immunol. 148:1941.
28. Pezzutto, A., B. Dorken, P. S. Rabinovitch, J. A. Ledbetter, G. Moldenhauer, and
E. A. Clark. 1987. CD19 monoclonal antibody HD37 inhibits anti-
immunoglobulin-induced B cell activation and proliferation. J. Immunol. 138:
29. Carrera, A. C., F. Sanchez-Madrid, M. Lopez-Botet, C. Bernabeu, and
M. O. De Landazuri. 1987. Involvement of the CD4 molecule in a post-activation
event on T cell proliferation. Eur. J. Immunol. 17:179.
30. Hoffman, R. A., P. C. Kung, W. P. Hansen, and G. Goldstein. 1980. Simple and
rapid measurement of human T lymphocytes and their subclasses in peripheral
blood. Proc. Natl. Acad. Sci. USA 77:4914.
31. Trauth, B. C., C. Klas, A. M. Peters, S. Matzku, P. Moller, W. Falk,
K. M. Debatin, and P. H. Krammer. 1989. Monoclonal antibody-mediated tumor
regression by induction of apoptosis. Science 245:301.
32. Pittet, M. J., D. E. Speiser, D. Valmori, J. C. Cerottini, and P. Romero. 2000.
Cutting edge: cytolytic effector function in human circulating CD8?T cells
closely correlates with CD56 surface expression. J. Immunol. 164:1148.
33. Holler, N., R. Zaru, O. Micheau, M. Thome, A. Attinger, S. Valitutti,
J. L. Bodmer, P. Schneider, B. Seed, and J. Tschopp. 2000. Fas triggers an
alternative, caspase-8-independent cell death pathway using the kinase RIP as
effector molecule. Nat. Immunol. 1:489.
34. Stambolic, V., A. Suzuki, J. L. de la Pompa, G. M. Brothers, C. Mirtsos,
T. Sasaki, J. Ruland, J. M. Penninger, D. P. Siderovski, and T. W. Mak. 1998:
Negative regulation of PKB/Akt-dependent cell survival by the tumor suppressor
PTEN. Cell 95:29.
35. Chan, T. O., S. E. Rittenhouse, and P. N. Tsichlis. 1999. AKT/PKB and other D3
phosphoinositide-regulated kinases: kinase activation by phosphoinositide-de-
pendent phosphorylation. Annu. Rev. Biochem. 68:965.
36. Maehama, T., and J. E. Dixon. 1998. The tumor suppressor, PTEN/MMAC1,
dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-
trisphosphate. J. Biol. Chem. 273:13375.
37. Ring, G. H., and F. G. Lakkis. 1999. Breakdown of self-tolerance and the patho-
genesis of autoimmunity. Semin. Nephrol. 19:25.
38. Chatenoud, L., B. Salomon, and J. A. Bluestone. 2001. Suppressor T cells they’re
back and critical for regulation of autoimmunity! Immunol. Rev. 182:149.
39. Rathmell, J. C., and C. B. Thompson. 2002. Pathways of apoptosis in lymphocyte
development, homeostasis, and disease. Cell 109(Suppl.):S97.
40. Spaner, D., K. Raju, B. Rabinovich, and R. G. Miller. 1999. A role for perforin
in activation-induced T cell death in vivo: increased expansion of allogeneic
perforin-deficient T cells in SCID mice. J. Immunol. 162:1192.
41. Gorak-Stolinska, P., J. P. Truman, D. M. Kemeny, and A. Noble. 2001. Activa-
tion-induced cell death of human T-cell subsets is mediated by Fas and granzyme
B but is independent of TNF-?. J. Leukocyte Biol. 70:756.
42. Martinez-Lorenzo, M. J., M. A. Alava, S. Gamen, K. J. Kim, A. Chuntharapai,
A. Pineiro, J. Naval, and A. Anel. 1998. Involvement of APO2ligand/TRAIL in
activation-induced death of Jurkat and human peripheral blood T cells. Eur.
J. Immunol. 28:2714.
43. Chen, W. F., and A. Zlotnik. 1991. IL-10: a novel cytotoxic T cell differentiation
factor. J. Immunol. 147:528.
44. Santin, A. D., P. L. Hermonat, A. Ravaggi, S. Bellone, S. Pecorelli, J. J. Roman,
G. P. Parham, and M. J. Cannon. 2000. Interleukin-10 increases Th1 cytokine
production and cytotoxic potential in human papillomavirus-specific CD8?cy-
totoxic T lymphocytes. J. Virol. 74:4729.
45. Comi, C., M. Leone, S. Bonissoni, S. DeFranco, F. Bottarel, C. Mezzatesta,
A. Chiocchetti, F. Perla, F. Monaco, and U. Dianzani. 2000. Defective T cell fas
function in patients with multiple sclerosis. Neurology 55:921.
46. Shenoy, S., T. Mohanakumar, T. Chatila, J. Tersak, B. Duffy, R. Wang,
A. R. Thilenius, and J. H. Russell. 2001. Defective apoptosis in lymphocytes and
the role of IL-2 in autoimmune hematologic cytopenias. Clin. Immunol. 99:266.
47. Di Somma, M. M., F. Somma, M. S. Montani, R. Mangiacasale, E. Cundari, and
E. Piccolella. 1999. TCR engagement regulates differential responsiveness of
human memory T cells to Fas (CD95)-mediated apoptosis. J. Immunol. 162:3851.
48. Li, X., M. Marani, R. Mannucci, B. Kinsey, F. Andriani, I. Nicoletti, L. Denner,
and M. Marcelli. 2001. Overexpression of BCL-XLunderlies the molecular basis
for resistance to staurosporine-induced apoptosis in PC-3 cells. Cancer Res. 61:
49. Miossec, C., V. Dutilleul, F. Fassy, and A. Diu-Hercend. 1997. Evidence for
CPP32 activation in the absence of apoptosis during T lymphocyte stimulation.
J. Biol. Chem. 272:13459.
50. Wilhelm, S., H. Wagner, and G. Hacker. 1998. Activation of caspase-3-like en-
zymes in non-apoptotic T cells. Eur. J. Immunol. 28:891.
51. Alam, A., L. Y. Cohen, S. Aouad, and R. P. Sekaly. 1999. Early activation of
caspases during T lymphocyte stimulation results in selective substrate cleavage
in nonapoptotic cells. J. Exp. Med. 190:1879.
52. Kennedy, N. J., T. Kataoka, J. Tschopp, and R. C. Budd. 1999. Caspase activation
is required for T cell proliferation. J. Exp. Med. 190:1891.
53. Budd, R. C. 2002. Death receptors couple to both cell proliferation and apoptosis.
J. Clin. Invest. 109:437.
54. Zapata, J. M., R. Takahashi, G. S. Salvesen, and J. C. Reed. 1998. Granzyme
release and caspase activation in activated human T-lymphocytes. J. Biol. Chem.
55. Bidere, N., M. Briet, A. Durrbach, C. Dumont, J. Feldmann, B. Charpentier,
G. de Saint-Basile, and A. Senik. 2002. Selective inhibition of dipeptidyl pepti-
dase I, not caspases, prevents the partial processing of procaspase-3 in CD3-
activated human CD8?T lymphocytes. J. Biol. Chem. 277:32339.
56. Roy, S., C. I. Bayly, Y. Gareau, V. M. Houtzager, S. Kargman, S. L. Keen,
K. Rowland, I. M. Seiden, N. A. Thornberry, and D. W. Nicholson. 2001. Main-
tenance of caspase-3 proenzyme dormancy by an intrinsic “safety catch” regu-
latory tripeptide. Proc. Natl. Acad. Sci. USA 98:6132.
57. Datta, S. R., H. Dudek, X. Tao, S. Masters, H. Fu, Y. Gotoh, and M. E. Green-
berg. 1997. Akt phosphorylation of BAD couples survival signals to the cell-
intrinsic death machinery. Cell 91:231.
58. Cardone, M. H., N. Roy, H. R. Stennicke, G. S. Salvesen, T. F. Franke,
E. Stanbridge, S. Frisch, and J. C. Reed. 1998. Regulation of cell death protease
caspase-9 by phosphorylation. Science 282:1318.
1181 The Journal of Immunology
59. Kane, L. P., V. S. Shapiro, D. Stokoe, and A. Weiss. 1999. Induction of NF-?B
by the Akt/PKB kinase. Curr. Biol. 9:601.
60. Madrid, L. V., C. Y. Wang, D. C. Guttridge, A. J. Schottelius, A. S. Baldwin, Jr.,
and M. W. Mayo. 2000. Akt suppresses apoptosis by stimulating the transacti-
vation potential of the RelA/p65 subunit of NF-?B. Mol. Cell. Biol. 20:1626.
61. Lauder, A., A. Castellanos, and K. Weston. 2001. c-Myb transcription is activated
by protein kinase B (PKB) following interleukin 2 stimulation of Tcells and is
required for PKB-mediated protection from apoptosis. Mol. Cell. Biol. 21:5797.
62. Di Cristofano, A., P. Kotsi, Y. F. Peng, C. Cordon-Cardo, K. B. Elkon, and
P. P. Pandolfi. 1999. Impaired Fas response and autoimmunity in Pten?/?mice.
63. Suzuki, A., M. T. Yamaguchi, T. Ohteki, T. Sasaki, T. Kaisho, Y. Kimura,
R. Yoshida, A. Wakeham, T. Higuchi, M. Fukumoto, et al. 2001. T cell-specific
loss of Pten leads to defects in central and peripheral tolerance. Immunity 14:523.
64. Papoff, G., P. Hausler, A. Eramo, M. G. Pagano, G. Di Leve, A. Signore, and
G. Ruberti. 1999. Identification and characterization of a ligand-independent oli-
gomerization domain in the extracellular region of the CD95 death receptor.
J. Biol. Chem. 274:38241.
65. Siegel, R. M., J. K. Frederiksen, D. A. Zacharias, F. K. Chan, M. Johnson,
D. Lynch, R. Y. Tsien, and M. J. Lenardo. 2000. Fas preassociation required for
apoptosis signaling and dominant inhibition by pathogenic mutations. Science
66. Sullivan, K. M., R. P. Witherspoon, R. Storb, H. J. Deeg, S. Dahlberg,
J. E. Sanders, F. R. Appelbaum, K. C. Doney, P. Weiden, C. Anasetti, et al. 1988.
Alternating-day cyclosporine and prednisone for treatment of high-risk chronic
graft-v-host disease. Blood 72:555.
67. Schiller, G., and R. P. Gale. 1993. Is there an effective therapy for chronic graft-
versus-host disease? Bone Marrow Transplant. 11:189.
1182CD95 RESISTANCE OF LONG-TERM ACTIVATED T CELLS