Guy’s, King’s and St. Thomas’ School of Medicine, Rayne Institute, London, United Kingdom
Correspondence: Alistair Noble, Department of Immunology, Rayne Institute, 123 Coldharbour Lane, London, SE5 9NU, U.K. E-mail: alistair.noble{at}kcl.ac.uk
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
/TNFR interactions because although type 1
cells secreted more TNF- than type 2 cells, the addition of TNFR:Fc
fusion protein did not inhibit AICD. Instead, a reduction in AICD was
observed in the presence of EGTA or concanamycin A. The inhibition of
apoptosis by a granzyme B inhibitor z-AAD-CMK in Tc1 cells further
indicated an involvement of the granule exocytosis mechanism in
AICD.
Key Words: apoptosis • CD4+ • CD8+ • TNFR • perforin
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(IFN-), and Th2 cells, producing
IL-5, IL-6, IL-10, IL-13, and IL-4 [1
, 2
].
The pattern of cytokines produced reflects the type of immune response
initiated. Type 1 cytokines tend to induce a cellular immune response
with inflammation, the activation of macrophages for anti-microbial
action, and delayed type hypersensitivity (DTH). Type 2 cytokines tend
to induce a humoral response, providing B-cell help for antibody
production and enhancing allergic responses with eosinophilia and
immunoglobulin (Ig)E formation [2
, 3
].
CD8+ T cells of type 1 (Tc1) and type 2 (Tc2)
cytokine-secreting phenotype have been increasingly described in rodent
and human models [4
, 5
]. In the human
system, Tc1 and Tc2 cells have been isolated ex vivo from peripheral
blood and draining lymph nodes of individuals with chronic infectious
diseases or allergies [3
, 6
]. Their
cytokine patterns were found to be similar to their CD4+
counterparts, although Tc2 cells are rare. CD4+ and
CD8+ subsets have been shown to have distinct functions and
are important in many aspects of disease resolution and pathology
[7
].
In addition to cytokine production, the most striking distinction
between type 1 and type 2 T-cell subsets is their susceptibility to
activation-induced cell death (AICD), a specific form of apoptosis
initiated in previously activated T cells following restimulation via
the CD3/T-cell receptor (TCR) complex. This form of apoptosis also
requires the interaction of specific receptors, termed death receptors,
with their ligands. For example, CD95 (Fas)/CD95 ligand (CD95L)
interactions are the major mechanism for the AICD of murine
CD4+ T cells [7
8
9
10
]. Another mechanism
using tumor necrosis factor (TNF-) and the TNF receptor (TNFR)
has been described as important in AICD of murine CD8+
cells [11
, 12
]. Moreover, mechanisms other
than those mediated by death receptors have also been found to be
important in T-cell apoptosis. Recently, a role for perforin in AICD of
murine T cells has been shown [13
14
15
]. In addition to
the critical importance of the perforin/granzyme pathway in the
induction of apoptosis in target cells by cytotoxic CD8+ T
cells [16
, 17
], the pathway may also
contribute to homeostatic control of activated T cells by AICD of
effectors themselves via suicidal or fratricidal mechanisms
[18
19
20
].
A comprehensive appreciation of how AICD is controlled in human,
primary effector cells is lacking. Therefore, our aim in this study is
to assess the differential susceptibility of CD4+ and
CD8+ T-cell subsets to AICD. We identify in each subset the
relative contributions of three major mechanisms of AICD: CD95,
TNF-, and perforin/granzyme B. We demonstrate that Th1 and Tc1 cells
are more susceptible to AICD than their type 2 counterparts; CD95/CD95L
interactions play a partial role in AICD; and TNF- is not involved
in CD4+ or CD8+ T-cell death. We also show a
clear involvement of perforin/granzyme B in AICD of CD8+
cells, demonstrating that the perforin/granzyme cytotoxic pathway
contributes to the high levels of apoptosis observed in Tc1 T cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CD8+ and CD4+ T cells were separated from PBMCs by positive selection using anti-CD8 or anti-CD4 Dynabeads (Dynal, Cheshire, UK), respectively, as previously described in our laboratory [21 ] and by others [22 ]. Positively selected cells were resuspended in 2% fetal calf serum (FCS)/phosphate-buffered saline (PBS) and released from beads by the addition of DETACHaBEADTM antibody (Dynal), as described in the manufacturer’s instructions. CD4+ and CD8+ T cells were diluted to 2 x 106 cells per ml in complete medium: RPMI 1640 with glutamine (Gibco BRL, Paisley, UK; 0.3 g/l), sodium pyruvate (Sigma Chemical Co.; 110 g/l), 2-mercaptoethanol (Gibco BRL; 0.05 M), gentamycin (Sigma Chemical Co.; 0.1 mg/ml), 1% modified Eagle’s medium (MEM) nonessential amino acids (Gibco BRL), and 10% AB human serum (Serotech, Oxford, UK). This method has been shown not to activate the cells [23 ].
Culture plates were coated with anti-CD3 (BD Pharmingen, Oxford, UK; 5 µg/ml) and anti-CD28 (CLB Amsterdam, Netherlands; 2 µg/ml) in PBS for 1 h at 37°. The wells were washed with complete medium prior to the addition of the purified T cells. Cells were added to the culture wells with equal volumes of culture medium containing recombinant (r)IL-12 (R&D Systems, Abingdon, UK; 500 U/ml) for the generation of type 1 cells or rIL-4 (R&D Systems; 200 U/ml) and neutralizing anti-IL-12 (R&D Systems; 10 µg/ml) for the generation of type 2 cells. Cytokine-containing medium was replaced on day 2 of culture, then with medium alone as required until day 6 in order to provide short-term primary effector cells. Cells were cultured at 37°C in 5% CO2.
Induction of apoptosis
AICD of short-term primary effector T-cell subsets was induced
by stimulation of cells with immobilized anti-CD3 (BD Pharmingen; 5
µg/ml) in the presence of 100 U/ml recombinant human (rh)IL-2 (R&D
Systems) unless otherwise stated. Controls included cells cultured in
the presence of rIL-2 alone with no anti-CD3 antibody.
Alternatively, apoptosis was induced by direct ligation of CD95 with an anti-CD95 antibody (CH11; Upstate Biotechnology, Lake Placid, NY). As a control, cells were incubated with an IgM control antibody (Caltag Laboratories, South San Francisco, CA).
Analysis of cell death
Apoptosis was measured by flow cytometry using Annexin V-FLUOS
(Roche Molecular Biochemicals, Mannheim, Germany) and propidium iodide
(PI; Sigma Chemical Co.) staining [24
]. Cells were
washed and incubated for 15 min in 100 µl Annexin V labeling solution
prepared in incubation buffer (10 mM Hepes/NaOH, pH 7.4, 140 mM NaCl, 5
mM CaCl2) with 10 µl PI (50 µg/ml). The test volume was
made up to 500 µl with incubation buffer, and the samples were
analyzed within 1 h of preparation using a FACScalibur flow
cytometer and Cellquest software (BD Pharmingen). Annexin V
single-positive cells were considered apoptotic and Annexin V/PI,
double-positive necrotic. For anti-CD3-induced AICD, apoptosis was
expressed as % specific apoptosis [25
]:
 |
Induced apoptosis is defined as the apoptosis induced with anti-CD3 in the presence of rIL-2. Control apoptosis is the apoptosis measured in the presence of rIL-2 alone with no cross-linking antibody present.
Blockade of AICD
AICD was blocked with the anti-CD95L antibody NOK1 (BD
Pharmingen; 10 µg/ml) [26
], TNFR:Fc fusion protein
(kindly provided by Immunex Corp., Seattle, WA) [27
],
the granzyme B inhibitor z-AAD-CMK (Calbiochem, Nottingham, UK)
[28
29
30
], or the caspase 8 inhibitor boc-AEVD-CHO
(Bachem, Saffron Waldon, UK) [31
]. In addition, EGTA, a
calcium chelator classically used to inhibit the exocytosis of lytic
granules from cytotoxic cells, was used to inhibit the exocytosis of
lytic granules from the T-cell subsets [32
].
Concanamycin A (CMA, Sigma Chemical Co.), an inhibitor of the proton
pump that maintains acidity of endosomes and lytic granules
[33
], was used to prevent the processing of the inactive
(70 kDa) precursor form of perforin to its active (60 kDa) form
[34
, 35
]. T cells were induced to undergo
AICD and cultured in the presence or absence of the inhibitors. Where
appropriate, the background levels of apoptosis were measured in the
presence of isotype-matched control antibody or Me2SO.
Additional controls included culture in the presence of rIL-2 alone and
medium alone.
Mixed lymphocyte reaction (MLR)
CD4+ T cells (105) were co-cultured with
5 x 104 allogeneic antigen-presenting cells (APC) in
a final volume of 200 µL in 96-well plates, as described by Becher
and colleagues [36
]. APC were prepared by negative
selection, and CD4+ and CD8+ cells were removed
using Dynabeads (Dynal). The remaining cells were irradiated at 5000
rads prior to co-culture with CD4+ T cells. T cells and
allogeneic APC were co-cultured for 5 days, and the culture supernatant
was harvested for detection of IFN- by specific cytokine
enzyme-linked immunosorbent assay (ELISA).
Flow cytometric analysis
For the analysis of cell-surface expression of CD95L and
TNF-, cells were restimulated in the presence of matrix
metalloprotease inhibitor KB8301 (BD Pharmingen) at 20 µM for 2 h. Cells (5x105) were then washed and stained with
anti-CD95L (Alexis, Nottingham, UK) or anti-TNF- mAb11 (BD
Pharmingen) for 20 min at 4°C in 100 µl wash buffer (PBS, 2%
FCS). Cells were analyzed using a FACScalibur flow cytometer and
Cellquest software (BD Pharmingen).
For analysis of cell-surface expression of CD95, cells were restimulated and stained with anti-CD95 antibody DX2 (BD Pharmingen) or the appropriate isotype-control antibody. For the analysis of cell-surface expression of TNFR1 and TNFR2, cells were restimulated, and samples were taken for staining at intervals until 24 h post-stimulation. Cells (5x105) were stained with anti-TNFR1 (mouse IgG1; anti-human TNFR p60, Genzyme, Cambridge, MA) or anti-TNFR2 (rat IgG2b; anti-human TNFR p80, Genzyme) for 20 min at 4°C in 100 µl wash buffer. Cells were washed and incubated with the appropriate fluorescein isothiocyanate (FITC)-conjugated secondary antibody using anti-mouse Ig or anti-rat Ig (BD Pharmingen).
For intracellular cytokine analysis, T-cell cultures were restimulated
with immobilized anti-CD3 and anti-CD28 for 6 h in the presence of
Brefeldin A (Sigma Chemical Co.; 5 µg/ml). Cells were harvested,
washed, and fixed for 20 min in 4% formaldehyde. After an additional
wash (200 g, 10 min) and resuspension in permeablization buffer [PBS,
1% bovine serum albumin (BSA), 0.5% saponin; Sigma Chemical
Co.], cells were incubated at room temperature for 20 min. After
washing, all subsets were incubated with anti-IL-4 PE (BD Pharmingen)
or IgG2a (isotype control) and anti-IFN- FITC (BD Pharmingen) or
IgG2b (isotype control) for 30 min at 4°C. Cells were washed twice
before analysis.
Cytokine ELISA
TNF- and IFN- were measured using commercially available
antibody pairs, a coating antibody anti-TNF-, mAb1 (5 µg/ml), and
the biotinylated anti-TNF-, mAb11 (1/100 dilution), both from BD
Pharmingen. For the IFN- ELISA, the coating antibody anti-IFN-,
NIB42 (2 µg/ml), and the biotinylated anti-IFN-, 4S.B3 (1
µg/ml), both from BD Pharmingen, were used. The ELISA was performed
according to the manufacturer’s instructions using the substrate
p-nitrophenyl phosphate (Sigma Chemical Co.). The ELISA was
read at 405 nm using an Emax microplate reader, and data were analyzed
using Softmax software (Molecular Dynamics, Sunnyvale CA).
Western blotting
Cell lysates were prepared by resuspending 2 x
106 cells in 40 µl hot lysis buffer [2% sodium dodecyl
sulfate (SDS), 62.5 mM Tris, pH 6.8, 10% glycerol, 770 mM
2-mercaptoethanol]. The lysate was boiled for 2 min and then passed
through a fine gauge needle to disrupt DNA. The cell lysates were
resolved on SDS-polyacrylamide gel electrophoresis (PAGE) gels (10%)
and blotted onto Hybond nitrocellulose membranes (Amersham Pharmacia
Biotech, Uppsala, Sweden) using a Trans-blot SD, semi-dry transfer cell
(Biorad, Hertfordshire, UK). Proteins were detected with anti-CD95L
(Transduction Laboratories, Lexington, KY) followed by goat anti-mouse
Ig coupled to horseradish peroxidase and detected using the ECL Western
blot detection system (Amersham Pharmacia Biotech).
Statistical analysis
Data were analyzed using Student’s t-test with
P values <0.05 considered significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2.7% CD8+CD56+,
0.9% CD19+, and 0.5% CD14+. Positive
selection of T cells with Dynabeads did not induce expression of the
early activation marker CD69 on CD4+ cells (0.5%
CD69+) or CD8+ cells (2% CD69+).
On day 6 of culture, Th1 and Th2 cells remained >99%
CD3+CD4+, and Tc1 and Tc2 cells were >97%
CD3+CD8+ with contaminants consisting of 2%
CD4+ cells, 1% CD56+ cells, and <1%
CD19+ or CD14+ cells.
In vitro generation of CD4+ and CD8+ T
cells secreting type 1 or type 2 cytokines
Th1 and Th2 or Tc1 and Tc2 primary effector T cells were generated
from human peripheral blood by stimulating CD4+ and
CD8+ T cells, respectively. T cells were stimulated with
immobilized anti-CD3 and anti-CD28 in the presence of recombinant IL-12
or neutralizing anti-IL-12 plus recombinant IL-4 to generate type 1 or
type 2 phenotypes, respectively. After 6 days of culture, polarization
of the populations to type 1 or type 2 phenotype was assessed by
intra-cytoplasmic staining for IFN- and IL-4 (Fig. 1
). Small amounts of IFN- were produced by CD4+ and
CD8+ cells cultured in type 2 conditions (to a maximum of
2%); this has been shown previously in CD8+ cells
[37
, 38
]. In addition, type 1 cells
produced very little IL-4. There were very few cells producing IL-4 and
IFN-, which suggests only minor contamination from Th0 or Tc0 cells.
 View larger version (43K): [in a new window] |
Figure 1. CD4+ and CD8+ T cells secreting type 1 or
type 2 cytokines can be generated in vitro. CD4+ and
CD8+ T cells were purified from human peripheral blood and
activated with immobilized anti-CD3 and anti-CD28 in the presence of
recombinant IL-12 or neutralizing anti-IL-12 plus recombinant IL-4 to
generate type 1 or type 2 cells, respectively. After 6 days of culture,
the cytokine-secreting phenotype of the cells was determined by flow
cytometric analysis. Cells were stained for IFN- (x-axis)
and IL-4 (y-axis) or the appropriate isotype controls to set
quadrant markers. The dot plot shows representative data from a single
individual. Similar data were obtained in six further experiments.
|
 View larger version (15K): [in a new window] |
Figure 2. Type 2 effector cells are resistant to AICD. T-cell subsets were
induced to undergo AICD by stimulation with immobilized anti-CD3 in the
presence of rIL-2 (100 U/ml) or IL-2 alone. (A) AICD of Th1 ( ), Th2
( ), Tc1 ( ), and Tc2 (x) subsets was measured by Annexin V/PI
staining at timed intervals following restimulation and expressed as
percentage specific apoptosis. The graph is representative of four
independent experiments. (B) AICD of T-cell subsets at 6 h
post-stimulation showing pooled data from 16 donors. Bars indicate
mean ± SE. Data were analyzed using Student’s
t-test, with ** representing P < 0.006 and
* representing P < 0.03.
|
 View larger version (39K): [in a new window] |
Figure 3. T-cell subsets express equal levels of CD95 and CD95L and are
susceptible to CD95-induced apoptosis. (A) CD95 expression was measured
on all subsets by flow cytometry using an anti-CD95 antibody (DX2,
solid histogram) compared with an isotype control (open histogram). (B)
CD95 ligand (solid histogram) is up-regulated on all subsets following
activation with anti-CD3 and IL-2 in the presence of a matrix
metalloprotease inhibitor (KB8301) compared with isotype control (open
histogram). The percentages refer to the number of CD95- or
CD95L-positive cells in the solid histograms. (C) Cellular extracts of
un-activated (un) and anti-CD3 re-activated (re) T cells were resolved
on a 10% SDS-PAGE gel, and CD95L protein was detected by Western blot
analysis using an anti-CD95L antibody. The lower molecular weight band
is nonspecific and is shown to verify that the total protein
concentration per well was similar. (D) A time course of apoptosis
induced by direct ligation of CD95 with an anti-CD95 antibody CH11 (1
µg/ml). The results show representative data from one of three
experiments. Apoptosis was measured in Th1 (), Th2 (), Tc1 (),
and Tc2 cells (x) by Annexin V staining. Differences between subsets
were not statistically significant.
|
To further assess the involvement of the CD95/CD95L pathway in apoptosis, AICD was initiated in each of the four subsets from five different donors, in the presence or absence of the anti-CD95L antibody NOK1. This antibody is shown to block CD95/CD95L interaction [26 ]. AICD was partially but significantly inhibited in Th1 and Tc1 subsets (P<0.02 and P<0.05, respectively) but not in Th2 or Tc2 subsets (Fig. 4 ). Similar results were obtained using a Fas:Fc fusion protein (BD Pharmingen) at 50 µg/ml (unpublished results).
 View larger version (16K): [in a new window] |
Figure 4. Blockade of CD95/CD95L-dependent AICD of Th1, Th2, Tc1, and Tc2 cells
with an anti-CD95L antibody NOK1. AICD was induced with immobilized
anti-CD3 in the presence of rIL-2 (100 U/ml). The T cells were
incubated in the presence (0.1–10 µg/ml) or absence of anti-CD95L
(NOK1). Apoptosis was measured by Annexin V staining 6 h
after reactivation. AICD was expressed as percentage specific
apoptosis. The data show the percentage of specific apoptosis of cells
from five individuals; each donor is represented by the same symbol in
each graph. Data were analyzed using a paired Student’s
t-test between the lowest (0 µg/ml) and the highest (10
µg/ml) concentration of NOK 1 used for each subset. P
values <0.05 were considered significant.
|
-mediated apoptosis in AICD was detected in culture supernatant from cells
restimulated with anti-CD3 in the presence of rIL-2. It was observed
that Th1 and Tc1 cells secreted more TNF- than their type 2
counterparts (Fig. 5A
). This is consistent with studies of differential TNF-
secretion in type 1 and 2 murine T cells [2
,
40
, 41
]. When restimulated cells were
cultured in the presence of matrix metalloprotease inhibitors, TNF-
was found to be expressed on the surface of all T-cell subsets, but a
greater number of cells in the type 1 populations expressed the
membrane-bound form of the cytokine (unpublished results).
 View larger version (15K): [in a new window] |
Figure 5. Evaluation of TNF- secretion and TNFR expression on T-cell subsets.
(A) TNF- production by the T-cell subsets Th1 (), Th2 (), Tc1
(), and Tc2 (x) was evaluated by specific cytokine ELISA of culture
supernatants collected at timed intervals following restimulation with
anti-CD3 and rIL-2. Data points show the mean of triplicate
samples ± SD values. TNFR1 (B) and TNFR2 (C)
expression was evaluated by flow cytometry. (B and C) Cells were
sampled at timed intervals, and receptor expression was determined by
staining cells with primary antibody, anti-TNFR1 (p60), anti-TNFR2
(p80), or the appropriate isotype control. The cells were then
incubated with the appropriate FITC-conjugated secondary antibody. Note
that Th2 cells have the lowest expression of TNFR1 and the lowest
secretion of TNF-. Results are representative of two independent
experiments.
|
. TNFR2 was not up-regulated on any of the subsets
but remained at its preactivation expression levels (Fig. 5C)
.
The effect of TNFR:Fc on AICD was studied to test whether AICD was
mediated by TNF/TNFR death pathways (Fig. 6A
). TNFR:Fc binds to soluble and membrane-bound TNF- and
prevents its interaction with TNFR1 or TNFR2 [27
].
T-cell subsets were restimulated with immobilized anti-CD3 in the
presence of rIL-2, and levels of AICD were determined at 8 h (Fig. 6A)
, 24 and 48 h (unpublished results) postinduction. The addition
of the TNFR:Fc did not inhibit AICD in any of the T-cell subsets tested
at any of the time points investigated. Addition of recombinant TNF-
in soluble or immobilized form did not induce apoptosis in any of the
T-cell subsets (unpublished results).
 View larger version (15K): [in a new window] |
Figure 6. TNF:Fc fusion protein fails to block AICD of human T-cell subsets. (A)
AICD of Th1 (), Th2 (), Tc1 (), and Tc2 (x) cells was induced
by stimulation with immobilized anti-CD3 (BD Pharmingen) in the
presence of rIL-2. Apoptosis was measured by Annexin V staining after
8-h restimulation in the presence of TNFR:Fc fusion protein. As a
control, apoptosis in the presence of IL-2 was also determined. The
values presented are the percentage of apoptotic cells, and the error
bars represent ± SE. Results show representative data
from one of three independent experiments. (B) Functionality of
TNFR:Fc. TNFR:Fc fusion protein inhibits TNF--mediated IFN-
production in a five-day MLR in which CD4+ T cells were
mixed with allogeneic APC and TNFR:Fc or an irrelevant human IgG.
Supernatant from the MLR was analyzed by ELISA for IFN-. Data
presented here are mean values ± SE.
|
-dependent production of IFN- in an allogeneic MLR
[36
]. Briefly, CD4+ T cells were co-cultured
at a 2:1 ratio with allogeneic APC. TNFR:Fc or the irrelevant IgG
control antibody was added to the reaction, and the cells were cultured
for 5 days. IFN- in the supernatant was measured by ELISA, and it
was shown that IFN- production was inhibited in a dose-dependent
manner by the TNFR:Fc (Fig. 6B)
. It was observed that IFN-
production was inhibited by 44.2%, confirming functionality of the
TNFR:Fc fusion protein.
Chelation of extracellular calcium inhibits AICD
EGTA inhibition was used initially to investigate whether lytic
granule-induced cell death played a role in AICD of T-cell subsets. A
reduction in apoptosis was observed in CD4+ and
CD8+ subsets in the presence of EGTA (5 mM). Higher levels
of inhibition were observed in CD8+ T cells, but this
effect was not statistically significant because of variability between
donors (Fig. 7A
).
 View larger version (25K): [in a new window] |
Figure 7. AICD of T-cell subsets is inhibited by EGTA and Con A. (A) T-cell
subsets were induced to undergo AICD by stimulation with anti-CD3 in
the presence of IL-2 (100 U/ml), in the presence or absence of 5 mM
EGTA. After 6 h, apoptosis was measured by Annexin V staining.
Values are from independent repeats using five different donors with
each individual represented by the same symbol in each graph. Data were
analyzed using a paired Student’s t-test comparing AICD
without EGTA and AICD with EGTA for each subset. P values
<0.05 were considered to be significant. (B) T-cell subsets were
pretreated with Con A (10 nM) for 16 h prior to the induction of
AICD. T-cell subset apoptosis was measured by Annexin V staining 6 h after induction of AICD in untreated (solid bars) and pretreated
(hatched bars) cells. The graph shows mean values from triplicate
cultures ± SD. Results are representative of two
independent experiments.
|
Effect of the granzyme B inhibitor z-AAD-CMK on AICD
When primary effector cells were induced to undergo AICD in the
presence of the specific granzyme B inhibitor z-AAD-CMK, AICD of Th1,
Th2, and Tc2 cells was unaffected. However, statistically significant
inhibition (P<0.05) of AICD was observed in the Tc1 subset
(45.4%±12.6%) at 10 µM (Fig. 8
) This indicated that granzyme B, a component of the lytic-granule
mechanism, is involved in AICD of CD8+ T cells.
Furthermore, the granzyme B inhibitor 2-AAD-CMK does not inhibit
CD95-medrated apoptosis whereas boc-AEVD-CHO, a caspase 8 inhibitor,
does (Fig. 8C)
.
 View larger version (13K): [in a new window] |
Figure 8. Granzyme B contributes to AICD of CD8+ but not
CD4+ T cells. (A) Inhibition of AICD by the granzyme
B-specific inhibitor z-AAD-CMK. AICD was induced in Th1 (), Th2
(), Tc1 (), and Tc2 (x) cells in the absence or presence of
z-AAD-CMK (0.001–10 µM). Apoptosis was assessed by Annexin V
staining 6 h after induction of AICD. Results are representative
of three independent experiments. (B) AICD was induced in all subsets
in the presence (hatched bars) or absence (solid bars) of the granzyme
B inhibitor, z-AAD-CMK (10 µM). Control and test samples contained
equal concentrations of Me2SO. Data from three different
donors were analyzed by Student’s t-test with
P < 0.05 considered significant. Bars represent
mean ± SE. Results showed significant inhibition of
granzyme B-mediated apoptosis in Tc1 cells (*, P<0.05). (C)
Direct ligation of CD95 by the anti-CD95 antibody CH11 induces
apoptosis in Th1 cells (solid bars). Background levels of apoptosis, in
the presence of rIL-2 alone (200 µ/ml; hatched bars), remain
unchanged by addition of the inhibitors. CD95-mediated apoptosis of Th1
cells is inhibited by the caspase-8 inhibitor boc-AEVD-CHO but is not
inhibited by the granzyme B inhibitor z-AAD-CMK. Values presented are
means ± SE from triplicate cultures and are
representative of two experiments.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our study, polarized subsets were generated after only a single
polyclonal stimulus in the presence of polarizing cytokines, not by
multiple stimulations as is the case for T-cell clones. Evidence
suggests that the number of cytokine-producing cells depends on
proliferation and cell-cycle progression. A strong relationship has
been found between cell division and cytokine gene expression.
Frequency of IFN- production was found to increase with each
subsequent cell division, and IL-4 expression required a threshold of
three cell divisions [46
]. These authors suggest that
during the cell cycle, there is the opportunity for the "derepression
of epigenetically silenced cytokine genes." In other words, cytokine
genes are reactivated, and their expression is stabilized during each
subsequent passage through the cell cycle. Therefore, our system would
be expected to have fewer cytokine-producing cells than a clonal model
that has had multiple rounds of stimulation.
The polarization of T-cell subsets was promoted by IL-12 and IL-4, which acted as growth factors for type 1 and type 2 cells, respectively. Whether the polarizing cytokines IL-12 and IL-4 selected cells with predetermined phenotypes and promoted their growth, enriching the population with subsets already present, remains a contentious issue and has been recently reviewed by Reiner [47 ]. This type of polarization, termed "selection," whereby cells have already been programmed by the actions of transcription factors such as T-bet in Th1 cells and Gata-3 in Th2 cells into type 1 or type 2 cells. It is therefore possible that our effector populations could have been derived from memory or precursor cells. Alternatively, cytokines may instruct cells to take on a specific phenotype, termed "instruction." A large body of evidence supports a selective [48 49 50 51 ] rather than an instructive mechanism [46 ] for T-cell differentiation. However, it is still possible that cytokines can have both effects on T cells [47 , 52 , 53 ].
In our study, human primary effector CD8+ cells behaved in an analogous way to CD4+ cells. Type 1 CD8+ T cells were found to be more susceptible to AICD than their type 2 counterparts. Although it is well established that CD95/CD95L is involved in AICD of CD4+ and CD8+ cells, we have demonstrated a role for CD95/CD95L interaction in AICD of Th1 and Tc1 cells. However, susceptibility to death was independent of the level of CD95L expression, suggesting other regulatory mechanisms are involved in CD95/CD95L-dependent AICD. Equally, differences in AICD of subsets were not a result of expression of CD95, because this was expressed at equivalent levels. The signaling function of CD95 may play a role in differences in susceptibility, because after direct ligation of CD95 with anti-CD95, slight differences in apoptosis of T-cell subsets were observed. It is possible that differences in signaling pathways of polarized T-cell subsets affect apoptosis. TCR-induced signals are initiated on CD3 ligation and have been implicated in the regulation of apoptosis. For example, MEK 1 mitogen-activated protein (MAP) kinase and extracellular-regulated kinase (ERK) activity have been shown to be important in regulation of CD95-dependent AICD [54 , 55 ]. The p38 kinase is also required for protection against TNF-mediated apoptosis [56 ]. It has been demonstrated that calcium signaling of type 1 cells is stronger than type 2 cells for CD4+ clones [57 , 58 ] and CD8+ clonal populations [59 ]. Calcium signals are important in T-cell activation and induction of apoptosis and may further contribute to differential susceptibility to AICD.
Differential secretion of TNF by type 1 T-cell subsets suggests a
potential role of TNF- in type 1 cell AICD. However, because TNFR:Fc
fusion protein or anti-TNF-receptor antibodies failed to block AICD,
and recombinant TNF- did not induce apoptosis, we concluded that
TNF- did not play a significant role in AICD of primary effector
cells generated in vitro. This may have been caused by a failure of
primary effector cells to up-regulate TNFR2 on reactivation. Human
primary effector cells show expression of TNFR1 on Th1, Tc1, and Tc2
cells and are similar to Jurkat cells in that they express TNFR1 but
not TNFR2 [60
] and are insensitive to TNF-induced
apoptosis [61
]. Our results confirmed the observation
that TNFRI is the TNFR predominantly expressed on T cells
[62
]. A mechanism for TNFR1 and TNFR2 cooperation has
been shown to be necessary to transduce a competent signal in PC60
cells [63
]. TNFR2 has been shown to facilitate
TNFR1-mediated apoptosis in T cells, thymocytes, HeLa transfectants,
and T-cell hybridomas [64
65
66
] by stimulating production
of membrane-bound TNF-. Also, "ligand passing" has been
described, where TNFR2 cooperates with TNFR1 by passing bound TNF-
to TNFR1 [67
]. Therefore, absence of TNFR2 may result in
the lack of sensitivity to death via TNFR1 [68
]. This is
in contrast to studies with murine cells where TNF- was shown to
mediate AICD of CD8+ T cells via the p75 TNFR (TNFR2)
[11
]. These authors showed that CD95 alone accounted for
almost all the AICD of CD4+ murine T cells and suggested
that the CD95L mechanism is primarily used by CD8+ cells to
kill target cells (including CD4+ cells), and the TNF-
pathway is used in auto-regulatory apoptosis (AICD) of the same
CD8+ populations.
Because TNF- was not shown to be involved in AICD of
CD4+ or CD8+ T cells, an alternative mechanism
was required to account for the considerable AICD not attributable to
CD95/CD95L interactions. Perforin has been shown previously to have a
role in AICD of murine T cells [13
14
15
] and along with
granzyme B, is a component of the lytic granules produced by cytolytic
T lymphocytes. Like CD95L and TNF-, granzyme B can also be expressed
following T-cell stimulation via a diverse array of stimuli, e.g.,
anti-CD3, Con A, phorbol 12-myristate 13-acetate, and allogeneic
stimulation [69
]. In the same way that CD95L is
preformed, granzyme B and perforin are stored in preformed lytic
granules ready for release on activation of the effector cell. By using
EGTA to chelate extracellular calcium and prevent lytic granule
formation [32
], we were able to show a partial but
variable inhibition of AICD in CD4+ and CD8+ T
cells. We then showed that components of lytic granules were involved
in AICD with granzyme B, important in the AICD of Tc1 cells. It has
been shown that a small, sustained Ca2+ influx
is sufficient and required for CD95L up-regulation as well as
CD95/CD95L killing [70
]. There have been contrasting
studies indicating that Ca2+ is also required
for CD95L up-regulation but that CD95/CD95L killing is
calcium-independent [71
72
73
]. Because perforin/granzyme
are predominantly produced by CD8+ T cells, inhibition of
AICD in CD4+ T cells by EGTA suggests inhibition of other
calcium-dependent mechanisms of AICD. Therefore, it is possible that
the effects of EGTA on CD4+-cell AICD that we observed were
a result of interference in CD95/CD95L-mediated apoptosis. This may
also be true for calcium-dependent expression of other death ligands.
Other mediators of apoptosis, e.g., caspases and cytochrome c release,
are also calcium-dependent [74
].
Another inhibitor of perforin is CMA, which is an inhibitor of the vascular type H+ ATPase, which acts as a proton pump to maintain acidification of endosomes and lytic granules [33 ]. This inhibitor was used to inhibit perforin-dependent AICD. Perforin is synthesized as an inactive precursor form that is cleaved to the active form only on entry into an acidic compartment. It is only the cleaved form that is stored in granules ready for secretion. Therefore, CMA is known to block selectively and completely the perforin pathway without affecting the CD95 pathway [34 ]. In the current study, we show that CMA inhibits AICD in CD8+ subsets, confirming a role for the granule exocytosis mechanism in AICD of CD8+ T cells.
Inhibition of cell-to-cell contact with an anti-CD18 antibody suggested that cell contact was not required for AICD. However, this does not preclude a role for granzyme in AICD, because although granzyme B has been shown to require perforin for entry into cells, a putative granzyme B receptor has also been described [75 76 77 ]. In addition, a role for granzyme in a cell-autonomous mechanism of apoptosis is possible, whereby death may be a result of internal trafficking of perforin after T-cell activation, which would initiate granzyme activity [14 , 78 ].
It is evident from the data presented here that CD4+ and CD8+ T cells share common mechanisms of AICD but that some events remain distinct between these subsets. The susceptibility of subsets to AICD may be linked to their function. Tc2 cells that are resistant to AICD may regulate their proliferation and clonal expansion by mechanisms other than apoptosis. They may kill their stimulating APC, thereby limiting their own stimulation, or reduce T-cell-growth factor secretion, thereby down-regulating the immune response [16 ]. T cells might exhibit reduced AICD because of differential expression of regulatory molecules such as Bcl-2 and Bcl-xL, but these have not been shown to be important in AICD of murine cells [79 , 80 ]. Some studies have indicated that the CD95-associated protease FAP-1 can block CD95-mediated cell signaling. This protects some cells from AICD, particularly Th2 cells where FAP-1 up-regulation corresponds to apoptosis resistance [45 , 81 , 82 ].
Our study demonstrates that human CD8+ T cells are highly
susceptible to AICD and that their cytotoxic function contributes to
auto-regulatory apoptosis, an important mechanism of self-tolerance.
Tc2 cells, like their Th2 counterparts, are relatively resistant to
AICD despite being susceptible to CD95-mediated death. In contrast to
murine T cells, we show that human CD8+ cells do not use
TNF- as an inducer of AICD. The lack of such regulatory pathways in
Tc2 cells may therefore contribute to the survival of cells producing
type 2 cytokines. This may occur after chronic exposure to antigen in
the lungs of asthmatics [83
] or in immune disorders, for
example in AIDS, hyper-IgE/hypereosiniphilia, and idiopathic
hypereosinophilic syndrome [6
], where type 2 cells
predominate. Therefore, so that appropriate treatments for the control
of immunopathology are developed, it is important that mechanisms
involved in AICD of different T-cell subsets are elucidated.
Received December 12, 2000; revised June 22, 2001; accepted June 22, 2001.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  Y. Zhang, G. Xu, L. Zhang, A. I. Roberts, and Y. Shi Th17 Cells Undergo Fas-Mediated Activation-Induced Cell Death Independent of IFN-{gamma} J. Immunol., July 1, 2008; 181(1): 190 - 196. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Mateo, M. Menager, G. de Saint-Basile, M.-C. Stolzenberg, B. Roquelaure, N. Andre, B. Florkin, F. le Deist, C. Picard, A. Fischer, et al. Perforin-dependent apoptosis functionally compensates Fas deficiency in activation-induced cell death of human T lymphocytes Blood, December 15, 2007; 110(13): 4285 - 4292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R. K. Loeb, J. L. Harris, and C. S. Craik Granzyme B Proteolyzes Receptors Important to Proliferation and Survival, Tipping the Balance toward Apoptosis J. Biol. Chem., September 22, 2006; 281(38): 28326 - 28335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Taylor, C. Gercel-Taylor, K. S. Lyons, J. Stanson, and T. L. Whiteside T-Cell Apoptosis and Suppression of T-Cell Receptor/CD3-{zeta} by Fas Ligand-Containing Membrane Vesicles Shed from Ovarian Tumors Clin. Cancer Res., November 1, 2003; 9(14): 5113 - 5119. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Elflein, M. Rodriguez-Palmero, T. Kerkau, and T. Hunig Rapid recovery from T lymphopenia by CD28 superagonist therapy Blood, September 1, 2003; 102(5): 1764 - 1770. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Strauss, I. Knape, I. Melzner, and K.-M. Debatin Constitutive Caspase Activation and Impaired Death-Inducing Signaling Complex Formation in CD95-Resistant, Long-Term Activated, Antigen-Specific T Cells J. Immunol., August 1, 2003; 171(3): 1172 - 1182. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Wieckowski, G.-Q. Wang, B. R. Gastman, L. A. Goldstein, and H. Rabinowich Granzyme B-mediated Degradation of T-Cell Receptor {zeta} Chain Cancer Res., September 1, 2002; 62(17): 4884 - 4889. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
