| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
B Cascade Is Important in Bcl-xL Expression and for the Anti-Apoptotic Effects of the CD28 Receptor in Primary Human CD4+ Lymphocytes1
*
Division of Clinical Immunology and Allergy, Department of Medicine, University of California, Los Angeles, CA 90095; and
Signal Pharmaceuticals, San Diego, CA 92121
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B pathway in the survival of
primary human CD4+ T lymphocytes during CD28 costimulation.
Transduction of proliferating CD4+ T cells with a
tetracycline-regulated retrovirus encoding for a dominant-interfering,
degradation-resistant I-B
(inhibitor of B factor) mutant
induced apoptosis. Using DNA arrays, we show that Bcl-xL
features as a prominent anti-apoptotic member among a number of
early CD28-inducible genes. A 1.2-kb segment of the proximal
Bcl-xL promoter, linked to a luciferase reporter, responded
to CD3/CD28 stimulation in Jurkat cells. Mutation of an NF-B site
around -840 decreased, while ectopic expression of I-B kinase-ß
(IKKß) enhanced reporter gene activity. Na+-salicylate
and cyclopentenone PGs, direct inhibitors of IKKß, interfered in the
activation of the Bcl-xL promoter and induced apoptosis in
CD28-costimulated CD4+ T cells. Moreover, salicylate
blocked nuclear localization of NF-B factors that bind to the
NF-B binding site in the Bcl-xL promoter, as well as the
expression of Bcl-xL protein. HuT-78, a lymphoblastoid T
cell line with constitutive NF-B activity, contained elevated levels
of Bcl-xL protein and, similar to proliferating
CD4+ T cells, was resistant to apoptotic stimuli such as
anti-Fas and TNF-. In contrast, the same stimuli readily induced
apoptosis in a Jurkat T cell clone with no detectable
Bcl-xL expression. Jurkat BMS2 cells also differed from
HuT-78 in collapse of mitochondrial membrane potential and superoxide
generation in the mitochondrium. Taken together, these data demonstrate
that CD3/CD28-induced activation of IKKß and expression of
Bcl-xL promote the survival of primary human
CD4+ T lymphocytes. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B cascade on genes that regulate
above biological responses in mature T cells. In this regard, we have
recently shown in primary human CD4+ lymphocytes
that the TCR is critically dependent on the CD28 accessory receptor for
the activation of the multisubunit inhibitor of B
(I-B)3 kinase (IKK)
complex, which induces I-B and I-Bß phosphorylation
(4, 5). Phosphorylation of these inhibitory proteins leads
to degradation and the release of Rel/NF-B proteins from the cytosol
to the nucleus (6, 7). IL-2 is among the genes that are
induced by NF-B transcription factors and contains key NF-B
regulatory sites in its promoter (4, 5, 8). While IL-2
plays a role in the effect of the CD28 receptor on T cell growth and
prevention of anergy, CD28 also assists in the expression of genes that
regulate cell viability (9, 10, 11, 12, 13). In this communication, we
will focus on the role of the CD28 receptor and the NF-B cascade in
preventing activation-induced cell death in human T lymphocytes.
The Rel/NF-B family of proteins has diverse and opposing effects on
apoptosis regulation in different tissues (14, 15, 16, 17, 18, 19). While,
for instance, Rel/NF-B factors suppress apoptosis in immature B
lymphocytes and developing thymocytes, c-Rel is highly expressed in
cells undergoing apoptosis in developing embryos (20, 21).
However, analysis of c-Rel-deficient mice does not support an
obligatory role for c-Rel in apoptosis (22). While there
are no definitive studies looking at the role of NF-B pathway in
human peripheral blood T lymphocytes, murine studies indicate that the
NF-B pathway has anti-apoptotic effects in mature T lymphocytes
(1, 23). Included among the anti-apoptotic genes that
may be NF-B regulated and that have demonstrated links to the CD28
receptor is Bcl-xL (24, 25). Both
the human and mouse Bcl-xL promoters contain
NF-B binding sites that have been shown to be activated in B cell
lines by CD40 or in a T cell line by Tax overexpression (26, 27).
We wished to determine which apoptosis-related genes are induced by
CD28 costimulation in primary human CD4+
lymphocytes, and whether activation of the NF-B pathway plays a role
in the expression of anti-apoptotic genes. We show through the use
of cDNA expression arrays that while CD28 costimulation is associated
with early expression of several pro-apoptotic genes, there is
prominent expression of only one anti-apoptotic gene, namely
Bcl-xL. Interference in the NF-B cascade in
CD4+ lymphocytes by a retrovirally expressed,
degradation-resistant I-B mutant or inhibitory drugs induced
apoptosis in parallel with decreased Bcl-xL
expression. Furthermore, NF-B-mediated activation of
Bcl-xL in CD4+ T cells is
influenced by IKKß function. CD3/CD28 stimulation as well as ectopic
expression of IKKß regulated Bcl-xL promoter
activity. In addition, Bcl-xL expression
correlated with resistance of proliferating CD4+
T cells to a variety of apoptotic stimuli. This indicates that
CD3/CD28-mediated activation of IKKß and subsequent expression of
Bcl-xL may coordinate the life span of
CD4+ T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
OKT3 (anti-CD3) was obtained from Ortho Pharmaceuticals
(Raritan, NJ), and the 9.3 mAb (anti-CD28) was generously provided
by Bristol-Meyer Squibb (Princetown, NJ). For the purpose of cellular
activation, these primary stimulating Abs were cross-linked with mAb
187.1 (Bristol-Meyer Squibb). Anti-Bcl-xL Abs and
human TNF- were purchased from PharMingen (San Diego, CA), while
anti-cIAP2 goat Abs were obtained from Santa Cruz Biotechnology
(Santa Cruz, CA). Anti-Flag (M2) Abs were obtained from Sigma (St.
Louis, MO). HRP conjugated to protein A was from Amersham (Arlington
Heights, IL). Tosyl-activated magnetic beads and M-450 anti-CD4
beads were purchased from Dynal (Great Neck, NY). PMA (P), ionomycin
(I), pyrrolidine-dithiocarbamate, and sodium salicylate
(Na+) were from Sigma. Anti-Fas mAb (CH11) was obtained
from MBL (Nagoya, Japan). The annexin V/propidium iodide (PI) staining
kit system was purchased from Trevigen (Gaithersburg, MD). Retronectin
was from Takara Biomedicals (Otsu, Shiga, Japan), and 3,3'
dihexyloxacarbocyanine iodide (DiOC6) and
hydroethidine (HE) were obtained from Molecular Probes (Eugene, OR).
Staurosporine and cycloheximide were purchased from ICN (Costa Mesa,
CA). Prostaglandins PGA1,
PGB2, and
15-deoxy-
12–14-PGJ2
(15dPGJ2) were purchased from Cayman Chemical
(Ann Arbor, MI).
Cell culture and preparation of primary CD4+ T cells
Jurkat (clone BMS2) and HuT-78 T cell lines were grown in RPMI medium, supplemented with 10 mM HEPES, 10% FCS, 2 mM glutamine, 100 U penicillin, and 100 µg streptomycin/ml. The 293-T epithelial cells were cultivated in DMEM containing 10% FCS, 2 mM glutamine, 100 U penicillin, and 100 µg streptomycin/ml. For preparation of CD4+ T cells, mononuclear cells were isolated from human peripheral blood by density centrifugation and depletion of adherent cells on plastic culture dishes. CD4+ T cells were positively selected with anti-CD4 Dynabeads according to manufacturer’s instructions (Dynal). This yielded a T cell subset that was >98% positive for the CD4 marker, as determined by dual color CD4/CD8 flow cytometry. Isolated CD4+ T cells were stimulated with anti-CD3 and anti-CD28 mAb coupled to tosyl-activated magnetic beads, as previously described (5, 28). Cells were replenished with 40 U/ml rIL-2 (Chiron, Emeryville, CA) on day 3, and allowed to complete their growth cycle over the course of 12–13 days. At this point, IL-2 and beads were removed, and cells allowed to return to their resting state over a 48-h time period. Resting vs activated state was assessed by cell size analysis in a Coulter counter (Coulter Pharmaceutical, Palo Alto, CA). Cells were subjected to flow cytometry for a second time to confirm that >95% of the cells remained CD4+.
Construction and preparation of a tetracycline (tet)-regulated
DN/I-B-expressing retroviral vector
The mutant I-B construct, with mutational alteration of
Ser32 and Ser36 to Ala
residues, has been previously described (7). The Hermes
HRIgfp tet-regulated retroviral vector was kindly provided by Dr. Helen
Blau (Stanford University, Palo Alto, CA) (29). The
backbone, in addition to long terminal repeats and packaging signal,
contains six copies of a tet operon (Tet O) linked to a minimal CMV
promoter (PhCMV-1) (see Fig. 1
A).
Green fluorescent protein (GFP) is expressed from an internal ribosome
entry site inserted downstream of the multiple cloning site
(29). N-terminal Flag-tagged DN/I-B generously
provided by Dr. G. Cheng (University of California at Los Angeles), was
directionally cloned into the HpaI-BamHI site of
the Hermes vector. In-frame insertion was confirmed by DNA sequencing
(see Fig. 1A). The second retroviral vector (pREV Tet-off)
encoding for the tet-repressible transactivating (tTA) protein was
obtained from Clontech (Palo Alto, CA). Individual retroviruses were
produced in the 293-Ampho (American Type Culture Collection, Manassas,
VA) packaging cell line. Briefly, packaging cell lines were plated in
10-cm dishes at a confluency of 
70% and transfected with 20 µg of
the Hermes or pREV Tet-off vector by calcium-phosphate precipitation
procedures in the presence of 25 µM chloroquine. After overnight
incubation at 37°C, the culture media were replaced with 6 ml of T
cell growth medium and transferred to 32°C. Viral supernatants were
harvested 48 h posttransfection, cleared by centrifugation at 2500
rpm for 5 min, and either stored in aliquots at -80°C or used to
infect cells. The pREV Tet-off viral titer was determined by selecting
neomycin-resistant colonies in NIH 3T3 cells and adjusted to a final
concentration of 5–8 x 106 infectious
particles/ml. Although a quantitatively exact titer could not be
obtained for the DN/I-B viral stocks, we used supernatants from
293-Ampho cultures that were >70% positive for GFP expression.
|
tTA and DN/I-B encoding viral supernatants were tested in
293-T cells and Jurkat lymphoblastoid cells. The 293-T epithelial cells
were seeded in six-well plates to 70% confluency and sequentially
infected with the pREV and Hermes viral supernatants. Briefly, 1 ml of
each supernatant was diluted 1/1 with complete DMEM containing 4
µg/ml of polybrene and added to the cells. The virus was centrifuged
onto the cells at 2200 rpm for 90 min at 37°C. Cells were rested for
6 h between infections. Infected cells were grown in the absence
or presence of 5 µg/ml tet. The infection efficiency of 293-T cells
was about 60%, as determined by GFP expression. Jurkat cell infection
followed the same guidelines, with slight modification. Briefly, 2
x 105 Jurkat cells were mixed with 1 ml of each
viral supernatant, diluted 1/1 in RPMI containing 4 µg/ml of
polybrene, transferred to 24-well plates, and centrifuged at 2200 rpm
for 90 min. The infection efficiency of Jurkat cells was about 40%, as
assessed by flow cytometry. An emission setting of 488 nm was used to
detect GFP.
Transduction of primary CD4+ T cells
Primary CD4+ T cells were amplified by CD3/CD28 costimulation, as described above. Three days postamplification, magnetic beads were removed and cells were transferred to 24-well plates coated with Retronectin (Takara Biomedicals) and incubated for 2 h. A total of 1 ml of viral supernatant, diluted 1/1 in T cell-conditioned medium, was added to cells and further centrifuged at 2200 rpm for 90 min. Cells were sequentially treated with the viral supernatants, using three rounds of infection over a 72-h time period in the presence of 5 µg/ml tet and 200 IU/ml of IL-2. On the fourth day postinfection, cells were washed three times in T cell culture medium, and incubated in the presence or absence of 5 µg/ml tet. The percentage of infected T cells was analyzed by flow cytometry, as described above.
Gene expression array
Total cytoplasmic RNA was isolated as described (30). The RNA pellet was resuspended in water and quantitated by absorbance at 260 nm. Approximately 200 µg of cytoplasmic RNA was used to isolate poly(A) RNA using Oligotex resin, as per the manufacturer’s instructions (Qiagen, Valencia, CA). Poly(A) RNA was eluted with 2 x 40 µl elution buffer, and the entire solution was quantitated at 260 nm. Typical recovery was 1.5–3 µg. The poly(A) RNA was precipitated with 95% ethanol in the presence of 0.3 M sodium acetate, pH 5.2, and 4 µg of glycogen. The pellet was washed with 75% ethanol, dried, and resuspended in water to a concentration of 0.5 µg/ml poly(A) RNA. The poly(A) RNA (1.5 µg) was used to produce a complex cDNA probe for screening of the Atlas Human cDNA Expression Array (Clontech), as described in the manufacturer’s protocol. The gene filters were sealed in plastic and the image data collected with a PhosphorImager and ImageQuant software (Molecular Dynamics, Sunnyvale, CA). Spot labeling, hit picking, and densitometric analysis were performed using a series of Excel-based macros developed at Signal Pharmaceuticals (San Diego, CA).
Apoptosis assays
Apoptosis in primary CD4+, HuT-78, and
Jurkat cells was assessed by dual color annexin V/PI staining.
Apoptosis was assessed in primary T cells 48 h after removal of
tet from retrovirus-infected cells or after incubation for 16 h in
the presence of sodium salicylate, prostaglandin
PGA1, PGB2, and
15dPGJ2 at the indicated concentrations.
Apoptosis was also assessed in all three cell types 16 h after the
introduction of 100 ng/ml anti-Fas mAb, 50 ng/ml TNF- (in the
presence of 10 µg/ml cycloheximide), or 1 µM staurosporine. Cells
were incubated with 1 µl annexin V plus 10 µl PI in 500 µl
binding buffer for 15 min, as recommended by the manufacturer. Flow
cytometry was performed using a FACScan (Becton Dickinson, Mountain
View, CA), equipped with a single 488-nm Argon laser. Annexin V-FITC
and PI were analyzed at emission settings of 535 and 575 nm,
respectively. Dual color staining for mitochondrial membrane potential
(
m) and superoxide production was performed
by DiOC6(3) and HE staining (31).
Briefly, 1 x 106 cells were incubated with
20 nM DiOC6 and 2 µM HE for 30 min at 37°C in
the dark. Cells were subjected to flow cytometry using emission
settings of 501 nm for DiOC6(3) and 605 nM
for HE.
EMSA
Gel-shift analysis of nuclear extracts from primary
CD4+ T cells and T cell lines was performed as
previously described (5). A total of 6 x
107 resting primary CD4+ T
cells or Jurkat was left untreated or incubated in the presence of 20
mM sodium salicylate for 2 h before stimulation for 30 min with
anti-CD3 mAb or a combination of anti-CD3 plus anti-CD28
mAbs, secondarily cross-linked with 187.1 mAb. Cells were immediately
washed and processed for extraction of nuclear extracts. A
double-stranded oligonucleotide containing the putative NF-B binding
site at positions -840 to -848 of the murine
Bcl-xL promoter (26)
(5'-cgatAAAGGGACTTCCAAGat-3' and 3'-taTTTCCCTGAAGGTTCtagc-5') was
end labeled with the Klenow polymerase in the presence of
[32P]GTP and [32P]dCTP.
DNA binding was performed at room temperature for 20 min in a final
volume of 20 µl, which contained 5 µg of nuclear extract, 5 x
104 cpm labeled oligonucleotide, and binding
buffer containing 100 µg/ml of poly(dI-dC) as nonspecific competitor.
Shift complexes were resolved on nondenaturing 5% polyacrylamide gel
in 0.5x TBE. Specific competition was performed with a 50-fold excess
of the unlabeled probe as well as Abs to p50 and p65, as previously
described (26).
Immune complex kinase assays
A total of 4 x 107 resting
CD4+ T cells was treated with 20 mM sodium
salicylate for 2 h before stimulation for 30 min with 2 µg/ml
anti-CD3 or a combination of 2 µg/ml anti-CD3 + 2 µg/ml
anti-CD28 mAb, secondarily cross-linked with 10 µg/ml mAb 187.1.
Stimulation with PMA (100 nM) plus ionomycin (1 µg/ml) was used as a
positive control. After cell lysates were precleared with protein
A-Sepharose beads, 200 µg lysate was treated with 2 µg
anti-IKKß absorbed onto protein A-Sepharose beads for 2 h.
Immune complexes were washed and equilibrated in kinase buffer, as
described (32). Kinase reactions were initiated by the
addition of 10 µCi [
-32P]ATP and 3 µg
GST-I-B (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54) substrate. The reaction was
conducted for 30 min at 30°C. Products were analyzed on SDS-PAGE and
detected by autoradiography.
Luciferase assays
A 1.2-kb Bcl-xL promoter element, cloned
upstream of a luciferase gene (pGL2 vector), was kindly provided by Dr.
G. Nunez (University of Michigan, Ann Arbor, MI) (33),
while wild-type (pGL848) and NF-B mutant
(pGL2BM) versions of a 848-bp luciferase
reporter were a generous gift from Dr. Mari Kannagai (Tokyo Medical and
Dental University, Japan) (26). A total of 12 µg of the
indicated reporter gene constructs was transiently transfected into
107 Jurkat cells in the presence or absence of
indicated amounts of IKKß (34). The cells were rested
for 24 h and then stimulated for 6 h with a combination of 2
µg/ml anti-CD3 + 2 µg/ml anti-CD28 mAb, secondarily
cross-linked with 187.1 mAb. Where indicated, transfectants were
incubated in the presence of 20 mM sodium salicylate or 20 µM
15dPGJ2 prior to stimulation. Control stimulation
was with 100 nM PMA + 1 µg/ml ionomycin. The cells were lysed in
luciferase buffer (Analytical Luminescence, Ann Arbor, MI), and
luciferase activity was measured in 50 µg of lysate in a Monolight
2010 Luminometer (Analytical Luminescence). Transfection efficiency was
monitored by cotransfection of a ß-galactosidase plasmid
(CMV-ß-gal); ß-galactosidase activity was used for adjusting
luciferase values between different transfected cell populations.
Western blot analysis
A total of 50 µg of cellular lysate was separated by SDS-PAGE and transferred to nitrocellulose membranes, as previously described (5). Membranes were sequentially overlaid with specific primary and secondary Abs conjugated to HRP, as indicated in the figure legends. Products were detected by enhanced chemiluminescence (ECL).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B mutant
leads to apoptosis in primary human CD4+ T cells
Coculture of primary human CD4+ lymphocytes
with anti-CD3 plus anti-CD28 mAb coimmobilized on beads induces
a prolonged (12–16 days) proliferative cycle during which the majority
of CD4+ T cells are protected from
activation-induced cell death by a signal imparted via the CD28
receptor (25, 28). This signal may involve the NF-B
cascade, which is activated in a synergistic fashion by TCR and CD28
costimulation (5). We asked whether the introduction of a
degradation-resistant I-B mutant will affect the survival of
CD3/CD28-stimulated human CD4+ lymphocytes. That
goal was accomplished by using DN/I-B expression by a
tet-regulated system that uses a combination of the pREV Tet-off and
Hermes HRIgfp retroviruses. The former virus encodes for a tTA protein,
while the latter expresses a copy of I-B(32A/36A) downstream a
heptamerized tet operator (Fig. 1
A) (29). The
Hermes virus also coexpresses GFP, which allows the tracking of
infected cells by fluorescence microscopy or flow cytometry (Fig. 1A). The efficiency of these retroviruses was demonstrated
in 293 and Jurkat cells, in which tet removal from the culture medium
could be seen to induce Flag-tagged DN/I-B (Fig. 1B)
and GFP (not shown) expression. These infection conditions were used to
develop optimal conditions for infection of primary lymphocytes.
In experiments in which these retroviruses were used to infect
CD3/CD28-stimulated CD4+ blasts, we observed GFP
expression in >40% of the tet(-) cells (Fig. 2A). In contrast, only 7% of
tet(+) cells expressed GFP (Fig. 2A).
When the same populations were used to assess cell viability by flow
cytometry, we could demonstrate an increase from 9 to 63% in the
percentage of annexin
V+/PI+ cells in the
tet(-) as compared with
tet(+) population (Fig. 2B).
DN/I-B expression in tet(-) cells was
confirmed by anti-Flag immunoblotting (Fig. 2C,
top panel). Examination of similar lysates with an antiserum
that recognizes I-B protein revealed that DN/I-B was
overexpressed compared with endogenous I-B levels (Fig. 2C, bottom panel). Taken together, these data
indicate that interference in the NF-B pathway by a
degradation-resistant I-B mutant negates the anti-apoptotic
effect of the CD28 receptor in primary human CD4+
lymphocytes.
|
B activity is linked to Bcl-xL expression and
cellular survival in primary CD4+ T lymphocytes
To determine which genes function as early response elements
during CD3/CD28 costimulation, we used the human Atlas (Clontech) cDNA
expression array, which includes an apoptosis-related panel.
Reverse-transcribed RNA from CD4+ lymphocytes,
treated for 1 h with the anti-CD3 + anti-CD28 mAb
combination or anti-CD3 mAb alone, was used for this analysis.
While there was no early message expression for anti-apoptotic
proteins such as cIAP1, cIAP2, cIAPx, Bfl-1, and Bcl-2, CD3 + CD28
stimulation induced Bcl-xL
expression (Table I). In contrast, CD3
ligation alone had a negligible effect on Bcl-xL
expression (Table I). All of the aforementioned genes have been linked
to NF-B activation (26, 27, 35, 36, 37, 38). Interestingly, CD3
+ CD28 costimulation also up-regulated expression of a number of
pro-apoptotic genes, including TNF- and Fas ligand (Table I). CD28
costimulation also induced nonapoptosis-related genes, belonging to the
transcription factor, enzyme, cytokine/chemokine, receptor, and stress
response categories (Table II). The
expression of some of these genes may involve activation of their
promoter by AP-1 and NF-B response elements, which are CD28
inducible (Table II).
|
|
B response elements in the human
and murine Bcl-xL promoters, which may play a
role in expression of the protein in T lymphocytes (26, 27). We confirmed at protein level that
Bcl-xL expression in primary human
CD4+ lymphocytes is linked to CD28 costimulation
(Fig. 3A). While CD3 + CD28
costimulation induced Bcl-xL expression in
CD4+ T cells, anti-CD3 mAb alone had little
effect (Fig. 3A). In contrast, cIAP2 was constitutively
expressed in these cells, and its relative abundance was not further
increased during cellular stimulation (Fig. 3B). The same
was true for cIAP2 expression in the Jurkat T cell line (Fig. 3B).
|
B activation, we made use of the inhibitory effects of
Na+-salicylate on IKKß activation
(41). This drug effectively suppressed IKK activation in
CD4+ lymphocytes during CD3 + CD28 costimulation
(Fig. 3C), and also had an inhibitory effect on
Bcl-xL expression during immunoblotting (Fig. 3A, lane 6). Although it has been suggested that
salicylate can induce apoptosis through the activation of the
p38MAPK cascade (42), we could not
demonstrate that a specific inhibitor of that cascade (SB203580)
exerted any effect on Bcl-xL expression or that
it could negate the inhibitory effects of
Na+-salicylate (Fig. 3D). A second
chemical inhibitor of the NF-B cascade in T lymphocytes,
pyrrolidine-dithiocarbamate (43), also interfered in
Bcl-xL expression in primary human
CD4+ T cells (not shown).
In the next of experiments, we asked whether
Na+-salicylate affected
CD4+ cell viability, and demonstrated a
dose-dependent increase in apoptosis in CD3 + CD28-treated cells (Fig. 4A). To confirm the
specificity of IKKß in this cellular effect, we also examined the
effect of the cyclopentenone PGs, 15dPGJ2 and
PGA1, which have recently been shown to directly
inhibit the IKKß subunit of the IKK complex in Jurkat cells
(44). As shown in Fig. 4B, treatment of
proliferating CD4+ T cells with these
cyclopentenone PGs induced apoptosis in a dose-dependent fashion. In
contrast, PGB2, which does not inhibit IKKß,
had no effect on CD4+ T cell survival (Fig. 4B). Collectively, these data indicate that IKKß activity
is essential for the survival of primary human
CD4+ T cells.
|
B
response element in T lymphocytes
To show more directly activation of the
Bcl-xL gene during CD28 costimulation, we used a
1.2-kb fragment of the murine promoter linked to a luciferase reporter
(Bcl-xL-Luc) for transfection into the Jurkat T
cell line (33). Although Jurkat cells failed to show
endogenous Bcl-xL expression during CD28
costimulation (see Fig. 5A),
CD3 + CD28 ligation had a stimulatory effect on reporter gene activity
(Fig. 5A). This response was enhanced by the addition of PMA
or cotransfection of IKKß cDNA (Fig. 5A). Please notice
that exogenously expressed IKKß (Fig. 5B) is
constitutively active in Jurkat cells (4, 5), which may
explain the 2.1-fold increase in basal luciferase activity in these
transfected cells (Fig. 5A). Moreover, the costimulatory
effects of CD3 + CD28 on the Bcl-xL promoter were
inhibited by treating the cells with the pharmacological inhibitors,
Na+-salicylate and PG
15dPGJ2, confirming that IKKß activity plays a
role in Bcl-xL promoter activity (Fig. 5C).
|
B-binding motif (GGGACTTCC), similar to the NF-B element in
the immunoglobin promoter, has been identified in the murine
Bcl-xL promoter (26). Mutation of
this site, located at bp -848 to -840 from the start site in an
848-bp promoter, significantly decreased the promoter response to
NF-B activation in a murine CTLL cell line (26). When
used in Jurkat cells, the mutant promoter could not be induced by CD3 +
CD28 costimulation, and did not respond to ectopic expression of IKKß
either (Fig. 5D). Moreover, we demonstrated that nuclear
extracts from Jurkat cells contain CD3 + CD28-inducible factor(s) that
binds to a labeled oligonucleotide probe corresponding to that NF-B
response element (Fig. 6A). Complex formation was
competed by a 50-fold excess of the cold probe (Fig. 6A, lane 4), and the complex was partially
supershifted with Abs to p50 and p65, as previously described
(26). It is interesting that P+I treatment, which leads to
IKK activation, induced the same shift complex (Fig. 6B,
lane 5). Similar effects were seen in
CD4+ T cells during treatment with anti-CD3 +
anti-CD28 or P+I (Fig. 6B, lanes 3 and
4). Na+-salicylate interfered in the
generation of these shift complexes in Jurkat as well as
CD4+ T cells (Fig. 6, A and
B). Similarly, the cyclopentenone PGA1
inhibited PMA-induced NF-B shift complexes and IKK activation in
Jurkat cells (44).
|
While the anti-apoptotic effects of the NF-B cascade cover
a wide range of stimuli, the role of Bcl-xL to
diverse stimuli has not been studied in any depth in primary human T
cells. To perform those studies, we used
Bcl-xL-expressing CD4+
blasts to look at the effect of Fas, TNF-, and staurosporine. For
comparison, we included a Jurkat cell clone, which does not express
detectable levels of endogenous Bcl-xL (Fig. 7A), as well as HuT-78, which
shows constitutive Bcl-xL expression (Fig. 7A). Interestingly, this constitutive
Bcl-xL expression in HuT-78 cells occurs in
parallel with constitutive NF-B activation (45). That
fact is demonstrated by the ability of nuclear extracts from
unstimulated HuT-78 cells to generate a shift complex with a probe
representative of the NF-B site in the
Bcl-xL promoter (Fig. 7B). Cellular stimulation with CD3 + CD28 mAb or P+I did not
increase the relative abundance of this shift complex (not shown).
|
, and staurosporine for 16 h, and subjected to two-color
annexin V/PI staining. While these stimuli readily induced apoptosis in
Jurkat cells, there was a comparatively minor effect in
CD4+ blasts and HuT-78 cells (Fig. 8A). This is compatible with
the wide range of anti-apoptotic effects of
Bcl-xL in T cells. Included among those effects
is regulation of the mitochondrial permeability transition pore
complex, which has an effect on m as well
as release of apoptogenic factors from the mitochondrium (46, 47). The drop in m during
mitochondrial perturbation can be assessed in a flow cytometer with the
fluorochrome, DiOC6(3). We therefore looked at
the decrease in DiOC6(3) fluorescence in HuT-78
vs Jurkat cells, and combined that with measurement of superoxide
(O2-) generation, as determined by HE staining
(31). O2- generation is a reflection of
damage to the mitochondrial inner membrane, which disrupts electron
transfer between complexes I and III (31). This experiment
showed that while staurosporine, anti-Fas, and TNF- had a
considerable effect on m as well as
O
2 generation in Jurkat cells, the same stimuli had
only a minor effect on HuT-78 cells (Fig. 8B). This suggests
that the constitutively active NF-B cascade and accompanying high
levels of Bcl-xL expression in HuT-78 cells exert
important anti-apoptotic effects at mitochondrial level.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B response site in
the promoter of that gene. Retroviral-mediated expression of a
DN/I-B mutant as well as treatment with direct inhibitors of
IKKß activity, namely Na+-salicylate and
cyclopentenone PGs, induced apoptosis in primary
CD4+ T cells. Since this suggests that the
CD3/CD28-induced NF-B cascade exerts anti-apoptotic effects in
primary T cells via Bcl-xL expression, we showed
that the activation of the promoter of that gene required NF-B
activation and was induced by IKKß expression in Jurkat T cells.
Moreover, HuT-78, a T cell line with constitutive activation of the
NF-B cascade, exhibited high basal levels of
Bcl-xL expression. HuT-78 cells, like
proliferating primary CD4+ T cells, have a
relatively high resistance to Fas-, TNF--, and staurosporine-induced
apoptosis compared with a Jurkat cell clone in which
Bcl-xL was not detectable. Therefore, among a
range of apoptosis-related proteins induced by CD28 costimulation,
Bcl-xL plays a major role in apoptosis
prevention.
Stimulus-induced degradation of inhibitory I-B and I-Bß
proteins activates NF-B transcription factors (6, 7).
We have been using Ab-coated beads as surrogate APCs to amplify primary
CD4+ T cells in studying the role of NF-B
pathway in proliferation, cell death, and anergy (5, 28).
Using a retrovirus-mediated gene transfer approach, we found that
interruption of NF-B pathway in CD4+ T cells
with superinhibitory DN/I-B inhibited cell proliferation and
induced apoptosis (Fig. 2B). The prosurvival effect of the
NF-B pathway has been demonstrated in many systems, including T
cells (18, 26, 48). For example, Kolenko et al.
(48) have demonstrated that pharmacological inhibition of
NF-B pathway in PBLs induces caspase-dependent apoptosis. Moreover,
CTLA-4 ligation, which has been shown to inhibit the proliferation and
survival of T cells, suppresses CD28-mediated activation of NF-B by
blocking I-B phosphorylation (49, 50). Therefore,
CD28-mediated NF-B activation is a prerequisite for the
proliferation and survival of mature T cells during ligation of
the TCR.
The anti-apoptotic effects of NF-B have been attributed to
induction of pro-survival gene products, including Bcl-2,
Bcl-xL, cIAPs, Al/Bfl-1, IEX-1L, TNFR-associated
factor 1 and 2, and superoxide dismutase (26, 27, 35, 36, 37, 38).
Examination of CD4+ T cells revealed that among
the anti-apoptotic proteins regulated by NF-B, CD3/CD28
stimulation resulted in the induction of Bcl-xL
(Table I). Consistent with previous reports, CD3 + CD28 costimulation
also induced expression of the pro-apoptotic Fas ligand (Table I). It
is interesting, therefore, that Bcl-xL has been
associated with inhibition of Fas-induced apoptosis in murine T cells
and exogenously transfected Jurkat cells (39, 51).
Bcl-xL also has the ability to rescue B cells
from Fas-induced apoptosis (27, 39). While interference in
Fas-induced cell death may explain the prosurvival effect of
Bcl-xL in CD28-costimulated primary
CD4+ T cells (Fig. 8A), it is equally
possible that this may also represent an interference in
TNF--induced apoptosis. In this regard, our cDNA arrays showed that
CD3/CD28 costimulation also induced TNF- message expression (Table I). While our studies were not designed to address
Bcl-xL interference in a particular cell death
pathway, it is noteworthy that the NF-B pathway has a known negative
regulatory effect on TNF--induced apoptosis (1).
Whatever the primary stimulus for cell death in
CD4+ lymphocytes during inhibition of the NF-B
pathway, CD28-costimulated cells were resistant to TNF-- and
Fas-induced apoptosis (Fig. 8A).
Several reports, including ours, have suggested that IKKß activity is
responsible for I-B phosphorylation in response to the majority
of NF-B-inducing signals (5, 52). Seminal studies in
mice in which the IKKß locus has been deleted support that notion
(53, 54, 55). Embryonic cells from IKKß knockout mice do not
activate the NF-B pathway in response to TNF- or IL-1 stimulation
(53), and die at midgestation due to severe liver
apoptosis (55). In this study, we show that treatment of
primary CD4+ T cells with
Na+-salicylate, an inhibitor of IKKß
(41), blocks Bcl-xL expression (Fig. 3D) as well as inducing apoptosis (Fig. 3E). This
notion was further substantiated by the pro-apoptotic effects of
recently discovered cyclopentenone IKKß inhibitors, namely
prostaglandins PGA1 and
15dPGJ2 (Fig. 4) (44). The induction
of apoptosis by Na+-salicylate in B cells and
myeloid leukemia cells has been associated with activation of caspases
(56, 57). It is also interesting that
Bcl-xL has been shown to interact with caspases
and interferes with the execution of the death phase
(58, 59, 60). Therefore, it is plausible that inhibition of
IKKß activity by cyclopentenone PGs and
Na+-salicylate is responsible for induction of
apoptosis in CD28-costimulated T cells.
Proliferating primary CD4+ and HuT-78
lymphoblastoid T cells resisted death in response to various apoptotic
stimuli, including TNF-, anti-Fas mAb, and staurosporine (Figs. 7 and 8). This resistance to apoptosis was also reflected in the
resistance of mitochondrial membrane perturbation in HuT-78 cells (Fig. 8B). This is compatible with the regulatory effects of
Bcl-xL on the mitochondrial permeability
transition pore (47), but does not rule out other
anti-apoptotic effects of this regulatory protein. For instance,
Bcl-xL has recently been demonstrated to interact
with Apaf-1 and several caspases (58, 59, 60). Thus,
heterodimerization with pro-apoptotic proteins and competition with
Apaf-1 for binding to caspases may represent additional
Bcl-xL activities in primary
CD4+ T cells.
The 3.2-kb stretch of DNA upstream of the Bcl-xL
initiation site contains a variety of different transcription factor
binding sites (33). While the contribution of most of
these elements in regulating the expression of
Bcl-xL remains unclear, some progress has been
made. For instance, signaling through IL-6 pathway, which leads to
activation of Stat3 and Janus kinases, has been associated with
induction of Bcl-xL expression (61).
Indeed, constitutive activation of Stat3 has been associated with the
pathogenesis of multiple myeloma through inhibition of apoptosis
(61). Two reports have also identified independent NF-B
binding sites in the Bcl-xL promoter (26, 27). In one of those studies on the human promoter, two tandem
NF-B binding sites were identified in the human promoter
approximately 200 bp upstream of the initiation site (27).
In the second study, a NF-B binding site was identified at 840 bp
upstream of the Bcl-xL initiation site in the
murine promoter (26). The activation of the latter binding
element by Tax protein is putatively involved in the ability of CTLL-2
cells to survive in the absence of IL-2 (26). DN/I-B
as well as mutation of this putative NF-B site abolished
Tax-mediated Bcl-xL promoter activity and
expression (26). While a 600-bp promoter element did not
respond to CD3/CD28-mediated costimulation or IKKß overexpression in
our studies (not shown), 1.2-kb and 848-bp Bcl-xL
promoter elements were responsive to these stimuli (Fig. 5A). The CD3/CD28-induced activation of the 1.2-kb
Bcl-xL-Luc promoter element was blocked by
pharmacological inhibtion of IKKß (Fig. 5C). In addition,
mutation of the NF-B site at position -840 to -848 in the 848-bp
reporter rendered it nonresponsive to CD3 + CD28 ligation (Fig. 5D). Moreover, CD3 + CD28 stimulation resulted in nuclear
localization of NF-B transcription factors that bound to an
oligonucleotide probe representative of that NF-B binding site (Fig. 6) (26). Nuclear localization and binding of these factors
were inhibited in primary CD4+ and Jurkat T cells
treated with Na+-salicylate (Fig. 6).
Furthermore, constitutive activation of the NF-B pathway in
untreated HuT-78 T cells was associated with constitutive
Bcl-xL expression and generation of an identical
NF-B shift complex (Fig. 7). These studies indicate that CD3 +
CD28-induced activation of NF-B pathway influences
Bcl-xL expression through this putative binding
site. Similar to CD3/CD28 stimulation, Tax-mediated NF-B activation,
which also leads to Bcl-xL expression, proceeds
through IKKß activation (26, 62).
The data presented indicate that regulation of IKKß activity in CD4+ T cells and expression of the anti-apoptotic protein Bcl-xL may coordinate the life span and expansion rate of specific T cell clones when they encounter particular Ags. These findings have important implications in the regulation of the mature T cell compartment and disease induction that may result from dysregulated apoptosis in these cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Andre Nel, Division of Clinical Immunology and Allergy, Department of Medicine, University of California, 10833 Le Conte Avenue 52-175 CHS, Los Angeles, CA 90095.
3 Abbreviations used in this paper: I-B, inhibitor of B factor; cIAP, cellular inhibitor of apoptosis protein; 15dPGJ2, 15-deoxy-12–14-PGJ2; DiOC6, dihexyloxacarbocyanine iodide; DN, dominant-negative; GFP, green fluorescent protein; HE, hydroethidine; IKK, I-B kinase; P+I, PMA plus ionomycin; PI, propidium iodide; tet, tetracycline; tTA, tet-repressible transactivating; 
m, mitochondrial membrane potential.
Received for publication February 4, 2000. Accepted for publication May 23, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B kinase cascade. J. Immunol. 162:3176.B kinase complexes: role in activation of the IL-2 promoter. J. Immunol. 163:5444.B: a mechanism for NF-B activation. Mol. Cell. Biol. 13:3301.B- on serines 32 and 36 controls IB- proteolysis and NF-B activation in response to diverse stimuli. EMBO J. 14:2876.[Medline]
B functions as both a proapoptotic and antiapoptotic regulatory factor within a single cell type. Cell Death Differ. 6:570.[Medline]
B activation to suppress p53-independent apoptosis induced by oncogenic Ras. Science 278:1812.B antiapoptosis: induction of TRAF1 and TRAF2 and cIAP1 and cIAP2 to suppress caspase-8 activation. Science 281:1680.B is activated and promotes cell death in focal cerebral ischemia. Nat. Med. 5:554.[Medline]
B to the rescue: RELs, apoptosis and cellular transformation. Trends Genet. 15:229.[Medline]
B. Trends Cell Biol. 8:107.[Medline]
B transcription factors during lymphocyte activation. Curr. Opin. Immunol. 10:353.[Medline]
B regulates Fas/APO-1/CD95- and TCR-mediated apoptosis of T lymphocytes. Eur. J. Immunol. 29:878.[Medline]
B in apoptosis-resistant T-cell transfectants with Tax. J. Virol. 73:7981.B-mediated up-regulation of Bcl-X and Bfl-1/A1 is required for CD40 survival signaling in B lymphocytes. Proc. Natl. Acad. Sci. USA 96:9136.B kinases essential for NF-B activation. Science 278:860.
