| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
Laboratories of
* Receptor Biology and Gene Expression and
Pathology, Advanced Technology Center, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892; and
Department of Microbiology, Howard University College of Medicine, Washington, D.C. 20059
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B pathways. Paradoxically, IL-2
repression by p53 is not reversed by mdm2. Instead,
mdm2 represses the IL-2 promoter by a mechanism that is
synergistic with p53 and resistant to Tax reversal. The p300
structure-function studies show that these effects are linked to
competitive associations among p53, Tax, and mdm2 with
multiple domains of p300. The functional outcome of these antagonistic
associations is revealed further by the observation that Tax and p53
induce apoptosis in activated T cells through separate and mutually
exclusive pathways. Interestingly, both pathways are abrogated by
mdm2. These results provide evidence that a dynamic
interplay, between Tax and specific elements of the p53 network,
mediates growth factor expression and programmed cell death in
activated T cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The function of p53 as a sequence-specific transcription factor is critical to its ability to regulate the expression of the multiple genes that control cellular proliferation and apoptosis. Recent studies have shown that p53 is able to exert this control through its association with a ubiquitous class of large (>2400-aa) transcriptional coregulatory proteins named p300 and CREB-binding protein (CBP)2 (7, 8, 9). Several laboratories have shown that p53 assembles with p300/CBP at the promoters of various p53-dependent genes through multiple protein-protein interactions. In many cases, these assembled complexes lead to direct activation of transcription; in others, these interactions have been proposed to repress (or trans repress) targeted genes through the sequestration of limiting quantities of p300 (10, 11). Whether the end result is activation or repression, an essential feature of p53 action has been its association with p300/CBP-containing complexes that assemble on the regulatory sequence of the target gene. In the past 3 years, at least four different domains along the 2400-aa length of p300/CBP have been found to interact with p53 (7, 9, 12, 13, 14). Moreover, separate N-terminal and core sequences within p53 are capable of independent interactions with p300/CBP (7, 11, 14, 15, 16). Thus, the mechanisms through which p53 is able to act as a master switch to control the expression of several different genes in diverse cellular pathways are likely to be the result of multiple complex interactions with p300/CBP that are selectively interwoven at each affected promoter.
The T cell oncogene, Tax, is a product of the human T cell leukemia virus type I (HTLV-1), and is the etiologic agent for adult T cell leukemia. Although it is not clear exactly how Tax transforms human T cells, this property is linked to the ability of Tax to activate the transcription of several pro-proliferation genes including IL-2, IL-2R, c-fos, GM-CSF, and numerous other genes that promote entry into the cell cycle (17). Like p53, a major mechanism through which Tax is able to control the activity of its target genes is by the formation of multiple complexes with p300/CBP (18, 19). Interestingly, Tax has also been found to antagonize the transcriptional activity of p53, although the precise mechanism of this antagonism remains controversial. Several studies suggest that this property arises from a site-specific competition between p53 and Tax for binding domains within p300/CBP, while others suggest that Tax may also induce inhibitory modifications of p53 that abrogate its ability to activate the transcription of target genes (13, 20, 21, 22). Regardless of the mechanism, these collective observations demonstrate that Tax mediates transformation through its dual ability to: 1) up-regulate the expression of genes involved in promoting growth and survival, and 2) inhibit the activity of those genes responsible for restraining growth, survival, and preventing genome instability in HTLV-1-infected cells.
Although these overlapping and antagonistic properties of Tax and p53 may predict a precise biological interplay in HTLV-1-infected cells, the potential functional outcome of these interrelations at downstream target genes has yet to be described in lymphoid cells. The T cell growth factor, IL-2, is a p300-controlled gene and a T cell-specific transcriptional target of Tax (23). As such, it is poised to be the focus of a true regulatory convergence between the interlinked and antagonistic functions of p53 and Tax. In this study, we provide evidence that the IL-2 gene, in Tax-expressing cells, is regulated by multiple interactions between the p53 pathway and the molecular processes controlled by Tax. This work is the first observation that the IL-2 promoter can be inhibited in trans by the p53 tumor suppressor, and also demonstrates that both the trans activation and trans repression activities of p53 are abrogated by the Tax oncogene. These combined observations illustrate that convergent targeting of p300 is a common denominator that dictates the differential outcome of cellular fate through its integration of cellular oncogene and tumor suppressor action at downstream effectors. Functional extension of this concept to the in vitro and in vivo analysis of IL-2 and Fas ligand (FasL) activation details a clear regulatory pathway through which both growth factor expression and programmed cell death are cross-regulated by Tax and p53.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Jurkat T cells were cultured in RPMI medium containing 10% FCS
and 100 U/ml of penicillin and streptomycin (Life Technologies,
Gaithersburg, MD) at 37°C in 5% CO2, as
previously described (24). Cells were activated with a
combination of either 1 µM ionomycin (Calbiochem, La Jolla, CA) and
50 ng/ml PMA (Sigma-Aldrich, St. Louis, MO), or plate-bound (1/200) mAb
OKT3 (
-CD3) ascites (American Type Culture Collection, Manassas, VA)
and 10 ng/ml PMA. Cell selection was conducted following transfection
using the pHook selection system (Invitrogen, Carlsbad, CA), according
to the manufacturer’s instructions. Transfection of Jurkat T cells
with plasmid reporter and expression vectors was accomplished by
electroporation, as previously described (24). Luciferase
(LUC) and chloramphenicol acetyltransferase (CAT) assays were performed
as described (24). LUC and CAT activity in the lysate from
each transfection was normalized for total protein using the Bio-Rad
(Hercules, CA) Protein Dye Reagent and expressed as fold activity over
that of cells that were either unstimulated, contained the reporter
alone, or both. All transient transfection studies were performed in
duplicates or triplicates. The error bars expressed in each figure
represent SEM. In all cases, each result is representative of at least
three independent experiments.
Plasmids
The pcDNA 3.1+ (Invitrogen, Carlsbad, CA), CD28RE-3'-12-O-tetradecanoylphorbol-13-acetate response element (TRE)-CAT, IL-2-CAT, IL-2-LUC, UAS-CAT E1B (23, 24), UAS-LUC (25), FasL-LUC (26), p21-LUC (27), mdm2-LUC, and PG13 (9, 28, 29) reporter plasmids have been described previously. The p53 wild-type and mutant expression vectors (p53 (WT), p53 (22, 23), and p53 (273H)) have been previously described (30, 31). The Tax expression vector used in this study was a generous gift from K. T. Jeang (National Institute of Allergy and Infectious Diseases, Bethesda, MD). The wild-type and mutant Tax expression vectors (pcTax (WT), pcTax (M22), pcTax (M47)) were provided by W. Greene (Gladstone Institute of Virology and Immunology, University of California, San Francisco, CA) (23). The Gal4-p300 fusion expression vectors used in this study have been previously described (23, 32).
Reagents and Abs
PMA, ionomycin, and mammalian protease inhibitor cocktail were from Sigma-Aldrich. Fas/Fc chimera protein was purchased from R&D Systems (Minneapolis, MN). Abs against p300 were affinity purified, as previously described, from antisera obtained from New Zealand rabbits inoculated with a GST-fusion construct containing aa 1–743 of p300. Anti-mouse and anti-rabbit Abs conjugated with Alexa Fluor 488 or Alexa Fluor 594 were purchased from Molecular Probes (Eugene, OR). Biotinylated oligonucleotides encoding the CD28RE-TRE sequences have been described (23). Streptavidin beads were from Pierce (Rockford, IL).
Ab staining and imaging
Cytological samples of Jurkat or primary human T cells (0.5–1 x 105 cells/ml) were isolated on glass microscope slides by centrifugation at 600 rpm for 5 min in a Cytospin 3 centrifuge (Shandon, Pittsburgh, PA). Cells were air dried for 1 h and fixed in methanol at -20°C for 30 min, rehydrated with PBS, then permeabilized in PBS containing 1% BSA and 0.5% Triton X-100, and blocked with PBS containing 1% BSA and 10% normal goat serum before staining with primary Ab. Secondary staining with Alexa Fluor-conjugated Abs was performed at 1/250 to 1/400 dilution in PBS with 1% BSA and 1% normal goat serum. Samples were stained for chromosomal DNA by incubation with 4',6'-diamidino-2-phenylindole at 0.1 µg/ml for 1 min at room temperature. Fluorescent confocal images were collected on a Zeiss (Oberkochen, Germany) Pascal 510 laser-scanning microscope at magnifications of x40 and x100, and analyzed using the Zeiss LSM Image Examiner Software v2.5.
DNA affinity precipitation assays
The p53-programmed nuclear extracts were generated by transfection of 60 x 106 Jurkat cells with 15 µg of p53 expression vector. Following stimulation with phorbol ester and ionomycin for 6 h, the viable cells were isolated by sedimentation on Ficoll gradients. Ficoll-purified cells were washed five times in 4°C PBS, and nuclear extracts were prepared, as described previously (33). DNA affinity precipitations were performed as described with 28-bp biotinylated duplex oligonucleotides encoding a single copy of the IL-2 CD28RE-TRE regulatory sequence (23).
Measurement of endogenous IL-2
Cells cotransfected with the indicated expression vectors and
the pHook single chain Ab expression plasmid were selected on Ag-coated
magnetic beads, according to the manufacturer’s specifications
(Invitrogen, Carlsbad, CA), 6 h following transfection. Cells were
resuspended in 5 ml of medium and were either not treated or stimulated
with plate-bound -CD3 Abs and 10 ng/ml PMA for 16–24 h
postselection. Following the incubation, the cells were removed by
centrifugation, and 4 ml of each supernatant was concentrated to 100
µl using a Microcon-30 Concentrator (Millipore, Bedford, MA). Final
IL-2 concentration was determined using the Quantikine IL-2 ELISA kit
(R&D Systems) in accordance with the manufacturer’s
specifications.
Apoptosis assay
Transfected, Ficoll-separated, and PMA/ionomycin
(P/I)-stimulated cells were incubated for 7–9 h at 37°C in medium.
The cells were isolated on glass slides (6 x
104 cells/slide) by Cytospin centrifugation at
500 rpm for 5 min. Following paraformaldehyde fixation and Triton
permeabilization, apoptotic cells were identified by TUNEL assay using
the In Situ Cell Death Detection kit (Roche, Basel, Switzerland),
according to the manufacturer’s instructions. TUNEL-positive (FITC
staining) and TUNEL-negative cells were visualized and counted by
fluorescent microscopy. Three representative fields (
200
cells/field) were counted for the number of TUNEL-positive cells per
total cells and reported as mean percent ± SD of the
TUNEL-positive (apoptotic) cells.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human Jurkat T cell line expresses IL-2 and is well known for
its p53-negative phenotype (34, 35, 36). This reduced
functional background provides a suitable experimental system to assess
the influence of wild-type and mutant forms of p53 and the Tax on IL-2
expression. As shown in Fig. 1
A, enforced expression of Tax
and p53 has opposing effects on endogenous IL-2 expression in Jurkat T
cells. Although expression of Tax alone causes the expected increase in
IL-2 production, expression of p53 induces a near 90% repression. This
repression is completely reversed by the coexpression of Tax.
Comparable effects of p53 are seen at a transfected reporter construct
containing the IL-2-proximal promoter, in which the abrogation of
activation of the IL-2 in the presence ofp53 is relieved by the
coexpression of Tax (Fig. 1B). Moreover, the IL-2 CD28RE-TRE
element, the major tax-responsive regulatory sequence of the IL-2
promoter, behaves in a manner similar to the endogenous gene. As shown
in Fig. 1C, repression of the IL-2 CD28RE-TRE by p53 is
completely reversed by coexpression of Tax.
|
B trans
activation potential of Tax
The inhibition of the IL-2 CD28RE-TRE by p53 and its reversal by
Tax are both dose-dependent phenomena (see Fig. 2, A and B). As
has been shown previously (23), the B and CREB
activation potential of Tax are required for full induction of the IL-2
promoter and the CD28RE-TRE. The Tax mutant M22 is defective in NF-B
signaling, while the Tax mutant M47 is defective in CREB signaling.
Both of these mutants are significantly impaired in their ability to
reverse the trans repression of the CD28RE-TRE by p53,
although the M22 (NF-B) mutant shows a slightly greater deficit
(Fig. 2C).
|
In addition to Tax, mdm2 is also a well-known
regulator of p53 through its ability to form transcriptionally inactive
complexes with p53 that mediate its nuclear export and degradation
(37). As shown on an artificial promoter containing
multiple tandem copies of a consensus p53 response element (PG13), the
addition of either Tax or mdm2 leads to significant
down-regulation of transcription (Fig. 3A). Although Tax is well able
to repress the mdm2 promoter when coexpressed with p53 (Fig. 3B), its combination with p53 produces synergistic
activation of the p21 promoter (Fig. 3C). Thus,
trans repression of p53 by Tax is highly promoter specific.
The ability of p53 to trans repress target genes does not
require its transcriptional activity. Variants of p53 with a mutation
in either its DNA binding (p53 273H) or trans activation
domain (p53 M22, 23) show minimum activation of the p53 reporter (Fig. 3A). However, both mutants retain significant ability to
trans repress the IL-2 CD28RE-TRE. The M22, 23 mutant
retains 50% of wild-type activity, while the 273H mutant retains
nearly 90% of wild-type inhibitory activity (Fig. 3D).
|
Several laboratories have shown that p53 influences the
transcription of regulated genes through a direct interaction with p300
that either recruits it to active promoters, or conversely, impedes the
activation of p300-targeted genes through sequestration and/or steric
inhibition (38, 39). The p300/CBP has two well-described
trans activation domains, in its C terminus and N terminus,
whose modular nature confers robust transcriptional activity to hybrid
fusions containing the DNA binding domain of Gal4. When such Gal4
fusions are expressed with a reporter driven by the sequence-specific
recognition site for Gal4 (UAS), they readily activate transcription.
An analysis of the influence of combinations of Tax and p53 on the
intrinsic transcriptional potential of full-length and truncation
mutants of p300 is presented in Fig. 4.
As shown in Fig. 4A, the intrinsic transcriptional activity
of full-length p300 is significantly repressed in the presence of p53,
and this repression is partially reverse by coexpression of Tax.
Expression of Tax alone has minimal effect. Moreover, both p300 N- and
C-terminal fusions with Gal4 are repressed by p53 in a manner that is
readily reversed by coexpression of Tax (Fig. 4, A,
B, and D).
|
C). These findings
suggest that part of the mechanism underlying the transcriptional
repression of p300 by p53 involves domain-specific interactions with
p300. This correlation is further supported by the reduced ability of
the p53 M22, 23 mutant (by comparison with the 273H mutant) to
trans repress the IL-2 CD28RE-TRE (see Fig. 3D).
Prior studies have shown that the M22, 23 mutant has a markedly reduced
affinity for the CBP/p300 class of transcriptional coactivators
(7). Thus, reversal of the repression is likely to result
from competitive displacement of p53 from a common or overlapping
domain by Tax. The p53 requires p300 for its assembly into protein-DNA complexes at the IL-2 CD28RE-TRE element
A finding that was critical to the understanding of mechanism of
p300/CBP involvement in p53 signaling was the demonstration that p53
could assemble onto target DNA in complexes with p300 (9, 39). Similarly, p53 forms complexes on the CD28RE-TRE element in
vitro in the absence of a DNA-binding p53 consensus sequence. Jurkat
nuclear extracts, programmed to express wild-type p53 by transfection,
were used to assess the ability of p53 to form complexes with
biotinylated duplex DNA, encoding the CD28RE-TRE element, via a DNA
affinity precipitation assay (Fig. 5).
Bound protein-DNA complexes were separated from free by sedimentation
with avidin-coated beads. As shown in Fig. 5, p53 was readily affinity
precipitated with avidin beads only in the presence of the CD28RE-TRE
element. This precipitation required the presence of p300 because
immunodepletion of the extracts with p300-specific Ab resulted in
complete loss of p53 from the CD28RE-TRE element. Depletion of the
programmed nuclear extract with nonspecific Ab did not interfere with
the DNA affinity precipitation of p53. Thus, p53 is likely to exert
most of its effect on repression of the IL-2 promoter through the
direct formation of an inhibitory DNA-protein complex, containing p300
and p53, at the IL-2 CD28RE-TRE element.
|
The mdm2 is a well-known regulator of p53 through its
ability to induce nuclear export and degradation of p53 by the
ubiquitin pathway (4). Therefore, it was anticipated that
mdm2 would readily reverse the p53 repression of the IL-2
promoter. Unexpectedly, mdm2 did not reverse the repression
of IL-2, but instead, acted in synergy with p53 (Fig. 6A). As shown in Fig. 6A, either p53 or mdm2 can inhibit activation of
the endogenous IL-2 gene, and their coexpression leads to
superrepression. Similarly, transcriptional activation of both the IL-2
reporter and the isolated CD28RE-TRE element is synergistically
repressed by p53 and mdm2 (Fig. 6, B and
C).
|
The mechanism through which mdm2 is able to mediate
repression of p300-controlled promoters is currently unknown. Several
laboratories have reported that mdm2 is able to form stable
complexes with p300 through interactions with its N-terminal domains
(14, 16, 44). In some of these reports, this interaction
has been shown to participate in competitive displacement of
trans-acting factors from the p300 N terminus
(16). To investigate the role of mdm2-p300
interaction in the repression of the IL-2 promoter, mdm2 was
expressed in reporter assays using the Gal4-p300 fusion constructs
described above. As shown in Fig. 7, coexpression of mdm2 is able to cause transcriptional
repression of each Gal4 fusion with p300, including the full-length,
the C-terminal truncation, the N-terminal truncation, and the
N-terminal truncation with a point mutation in the KIX domain. This
repression is increased by coexpression of p53 and is completely
resistant to reversal by Tax (Fig. 7).
|
As shown above, the linked interplay between Tax and p53, when
compared with that between mdm2 and p53, is functionally
distinct. Although mdm2 antagonizes p53-dependent
trans activation in a manner similar to Tax (see Fig. 3A), it conversely synergizes with p53-dependent
trans repression of the IL-2 promoter and p300. These
functional differences are reflected morphologically in experiments
that examine the subcellular localization of p53 when coexpressed with
Tax in lymphoid cells. Immunofluorescent detection of p53 and p300 in
cells transfected with p53 shows a high level of nuclear localization
of p53 in a manner similar to p300 (Fig. 8, top). Although coexpression
of Tax and p53 does not alter the nuclear accumulation of p53 (Fig. 8, middle), the coexpression of p53 and mdm2 leads
to significant exclusion of p53 from the nucleus (Fig. 8, bottom).
|
Both p53 and Tax previously have been shown to induce apoptosis in
lymphoid cells, although the precise mechanism of their effects remains
to be determined (45, 46, 47). As shown in Fig. 9A, enforced expression of p53
or Tax in activated Jurkat cells is associated with a marked increase
in apoptosis (Fig. 9A). As expected, this p53-dependent
apoptosis is completely abolished by coexpression of mdm2;
however, coexpression of Tax and p53 together results in a significant
decrease in apoptosis below the level induced by either factor alone.
One of the pathways through which Tax can induce apoptosis is by
activation of FasL expression (47). To investigate the
influence of p53 on this pathway, FasL reporter assays were conducted
in the presence and absence of Tax and p53. Tax expression in Jurkat T
cells leads to a dramatic up-regulation of the FasL promoter. Both this
Tax-induced activation and the endogenous activation of the FasL
reporter are significantly inhibited by coexpression of wild-type p53
(Fig. 9B). Interestingly, this repression is not reversed by
the addition of mdm2. In fact, expression of mdm2
caused significant repression of both endogenous and Tax-induced
activation of the FasL promoter (Fig. 9B). Finally,
Tax-expressing Jurkat cells are rescued from apoptosis via receptor
blockade by the addition of Fas/Fc (a soluble Fas-IgG fusion that acts
as a competitive inhibitor of FasL) (48) (Fig. 9C). Jurkat cells overexpressing p53 are not rescued, and
the cross-abrogation of Tax-induced apoptosis by the expression p53 is
uninfluenced by this treatment (Fig. 9C). Consistent with
the FasL reporter observations (Fig. 9B), mdm2
continues to show a superdominant effect: repressing both Tax-dependent
and p53-induced apoptosis. Thus, Tax and p53 induce apoptosis through
mutually exclusive pathways that intersect to create a functional
antagonism at the level of FasL expression and action.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The ability of p53 to trans repress target genes has been a
well-recognized aspect of p53 function that was initially observed over
a decade ago (49, 50, 51). The finding that p300/CBP
coregulator complexes could form stable interactions with p53 led to
the general conceptual consensus that the inhibitory effect of p53
could be explained by competitive interactions between p53 and various
transcription factors for finite binding sites on limiting amounts of
nuclear CBP/p300. The relative validity of this concept was established
in numerous studies that showed that overexpression of p300/CBP could
ultimately relieve or rescue cells from their repressed states
(reviewed in Ref. 38). What has remained to be established
is what form this level of competition ultimately takes within the
microenvironment of the nucleus. Two distinct models for this
competition have emerged. In one model, promoter-targeted transcription
factors compete for limiting amounts of available p300/CBP in the
nucleus, and therefore, depending on cellular context, actively
sequester p300 away from specific gene clusters in favor of function at
selected gene cassettes. In a second general model, competition exists
at the level of steric interference. p300/CBP engages in specific gene
regulation through the differential formation of active multicomponent
complexes at the promoters. By this mode of regulation, the binding of
p53 to sites on p300 disrupts appropriate assembly of the
promoter-specific complex through steric hindrance. Although both
models are dependent on a direct interaction between p300/CBP and p53,
the former model also requires sequence-specific interactions with
promoters that are envisioned to be competing for available p300/CBP.
Thus, the first model would require that the sequence-specific
DNA-binding activity of p53 remain intact. As shown in Fig. 3B, a tumor-derived DNA-binding mutation in p53 (273H)
retained near wild-type activity in repression of IL-2 regulatory
sequences, while the p53 mutant of p53 that is defective in p300/CBP
interactions (M22, 23) shows significantly reduced
repression.
The p53 interacts with p300 along four different stretches of its
primary sequence (see schematic summary in Fig. 10B). It binds to the very C
terminus of p300 via its N-terminal sequences (7, 12). A
second binding site, with apparent lower affinity, is located within
the CH3 domain (7, 9). A third
binding site that interacts with the core domain of p53 is located in
the CH1 domain of p300, and the KIX domain of
p300/CBP is the fourth region of p300 demonstrated to interact with p53
(13, 14). At present, it is not clear whether these
multiple binding sites for p53 act independently or are assembled into
a single binding pocket(s) through tertiary folding of the p300
structure (see Fig. 10B, alternative structure).
Nonetheless, the Gal4-p300 fusion assay shows that p53 targeting of at
least three of four of these domains results in significant repression
of the intrinsic transcriptional activity of p300 (summarized in Fig. 10A). This suggests that the p53-binding activity retained
by the isolated p300 segments is sufficient to be functionally relevant
(see Figs. 4 and 7). Although both of these models provide a unifying
framework upon which the results from multiple laboratories can be
explained, they do not distinguish between the mechanisms of
sequestration vs that of steric hindrance that have been proposed as
explanations for the trans repression of p300-dependent
genes, as described above. Notwithstanding, the binding pocket proposed
in the alternative model (Fig. 10B) implies that multiple
allosteric interactions could exist along the length of p300/CBP that
could be targeted for regulation by a variety of protein:protein
interactions and covalent modifications.
|
Tax expression has pleiotropic direct and indirect effects on nuclear
and cytoplasmic processes within the cell. Two of these effects are
related to the ability of separate domains within Tax to up-regulate
CREB and rel/B transcriptional activity. Both of these
factors have binding sites in or near the KIX domain of p300. Thus, in
addition to binding directly to p300, Tax induces the nuclear activity
of factors that may also limit the accessibility of p53 to domains
within p300. This possibility is supported by the finding that Tax
mutants deficient in the ability to induce either CREB or
rel/B activity (M47 and M22 mutants, respectively) are
also deficient in reversing p53 repression, although the M22 mutant
appears more defective (Fig. 2C). Finally, Tax induces a
differential phosphorylation of p53, and this activity has been linked
to its ability to inhibit p53-dependent activation of target genes
(22). The work described in this study does not address
the competence of different phosphorylation states of p53 to
participate in trans repression of target genes.
Nonetheless, it is noteworthy that mutations within the KIX domain of
p300 block the ability of Tax to reverse the p53-induced repression of
the activity of the p300 N-terminal trans activation domain
(Fig. 4C). This implies that a direct association between
Tax and p300 is a required step in reversing the repression. However,
the reported inability of the M22 Tax mutant to induce the
tax-dependent phosphorylation of p53 suggests that both competitive and
noncompetitive influences of Tax on p53 may contribute to the
cross-regulation (22).
The role of mdm2 as a negative regulator of p53 through
increased p53 degradation and nuclear export is well known. That is why
the inability of mdm2 to reverse the p53-induced repression
of the IL-2 promoter and p300 trans activation domains was
surprising (see Figs. 6 and 7). However, this observation is consistent
with a prior report that mdm2 possesses an intrinsic
transcriptional repression domain located adjacent to its p300
interaction domain (54). Although not completely
characterized, the action of this mdm2 domain appears to be
specific for TATA-containing promoters and is linked to its ability to
interfere with basal transcriptional machinery. It may be that the
mdm2 interaction with p300 and p53 may aid in its targeting
of genes for repression. The presence of p53 enhances the stability and
specificity of the repressive p300/p53/mdm2 complexes until
p53, and possibly p300 are ubiquitinated. Thus, p53-dependent
repression of p300-controlled genes must occur in two sequential
stages. The first stage is due to the action of p53 alone, is
independent of its transcriptional activity (see Fig. 3), and can be
reversed by the expression of Tax. The second stage requires the
transcriptional activity of p53. This consequently increases
mdm2 levels, which results in a more profound and
Tax-resistant repression (see Figs. 6, 7, and 8). The targeting of p300
by mdm2 is likely to be the primary mechanism through which
mdm2 is able to antagonize the apoptotic pathways induced by
Tax and p53 (see Fig. 9) and may contribute to the reported ability of
mdm2 overexpression to contribute to tumorigenesis
independent of p53 (55).
The ability of Tax to induce apoptosis in T cells through the
up-regulation of FasL expression and its antagonism by coexpression of
p53 may provide insights into the different mechanisms of tumor
induction by the oncogenic viruses. The transforming proteins of
several oncogenic viruses interact with the p53 tumor suppressor
protein to alter its function. The SV40 large T Ag and the adenovirus
E1B protein form complexes with p53 that both stabilize its levels
while disabling its function (56, 57, 58). The E6 protein of
the human papilloma virus forms a complex with p53 that ultimately
leads to it degradation through the ubiquitin-directed pathway
(59). What is common to these tumor virus proteins and the
endogenous regulation of p53 by mdm2 is that they all
disable p53 function through a mechanism that has significant effects
on the compartmentalization and the steady state levels of p53. These
properties are clearly distinct from the activity of Tax. Although Tax
does influence many of the activities of p53, it does not lead to
dramatic changes in either its levels or its subcellular localization
(compare effects of mdm2 and Tax on p53 in Fig. 8). These
findings in addition to the observation that few, if any, tumors of T
cell origin are associated with loss of heterozygosity at the
p53 locus (60) suggest that some aspect of p53
action must be retained during the initial transformation process. The
finding that p53 can antagonize the induction of apoptosis by Tax
expression in activated T cells suggests that this may be a point of
symbiosis between Tax and p53. Work by Rivera-Walsh et al.
(45) clearly demonstrates that Tax expression can also
induce apoptosis through T cell activation-independent (and therefore
FasL-independent) pathways by the up-regulation of the TNF-related
apoptosis-inducing ligand gene in response to chronic
overexpression of Tax. By striking a balance between tax-regulated
pathways of T cell proliferation and activation-induced cell death, it
is therefore likely that both the activation state of the T cell and
the steady state levels of p53 may be critical determining factors in
whether HTLV-1 infection results in cellular transformation. The
demonstration of the upstream linkage of Tax expression and
FasL-induced apoptosis (see Fig. 9) adds credence to this
hypothesis.
IL-2 also plays an active role in promoting apoptosis in activated T cells through its ability to increase FasL expression and suppress the FLIP inhibitor of Fas-induced cell death (61). This paradoxical function of IL-2 suggests that a complicated and kinetically determined hierarchy of interactions and targeting by Tax and p53 may exist that must proceed through thresholds during the course of T cell activation. Thus, the early symbiosis between Tax and p53 must give way to other secondary events during later stages of T cell activation as cells acquire high affinity IL-2R and become more susceptible to apoptosis. The nature of these secondary events is not clear, but they are likely to result from the increasing genetic instability of the proliferating T cells due to the down-regulation of p53 by Tax. These secondary events are also likely to contribute to the mechanisms through which HTLV-1-infected cells eventually become IL-2 and Tax independent as they evolve into tumors (62).
Although prior Tax transgenic studies show accelerated Tax-dependent tumor growth in heterozygous p53 null mice, the early demise of the homozygous p53 mice from Tax-independent tumors makes it hard to determine whether there is a true dose-dependent role for p53 in Tax-induced cellular transformation (63, 64). Future studies designed to compare the ability of infectious clones of HTLV-1 to immortalize wild-type vs p53 heterozygous or homozygous null lymphocytes would be better able to test this hypothesis.
The finding that elevated levels of p53 and mdm2 can antagonize IL-2 expression in activated T cells suggests the possibility that the p53 network may have a role in negative feedback inhibition of proliferation and cytokine gene expression. Although p53 levels are elevated in activated T cells (65), it is not clear whether during normal T cell activation, the presence of p53 and later mdm2 serves to limit IL-2 expression. Studies with p53 knockout mice show increased proliferation of the deficient cells (6), yet detailed studies that examine differences in cytokine production are not available. Several studies have reported that environmental exposures to DNA damage agents, including chemotherapeutic drugs, ionizing radiation, naturally occurring heavy metals, and pesticides, will elevate the levels of p53 in peripheral blood lymphocytes (66). Our data predict that these exposures can lead to a chronic immunosuppression of naive and memory T cells characterized by down-regulation of IL-2 expression through activation of the p53 network.
Finally, the interaction and functional outcome of the association of
Tax and p53 with multiple domains within p300/CBP have been described
by numerous labs; however, the precise manner in which these
interactions target and ultimately influence the expression of cellular
genes has received little attention. Fig. 10A presents a
rudimentary conceptual summary of how such interactions influence
various p300-targeted promoters in vitro. Whether the implications of
these observations will have application to our understanding of how
the variety of p300-controlled promoters are regulated in vivo will
require approaches that place these regulatory systems in a more
physiological context. The ability to obtain kinetic and spatial
information on protein-protein and protein-DNA interaction through
chromatin precipitation assays promises to be a very powerful approach
in the future to identify those in vitro determined biochemical
interactions that have biological relevance in vivo.
|
Footnotes |
|---|
2 Abbreviations used in this paper: CBP, CREB-binding protein; CAT, chloramphenicol acetyltransferase; FasL, Fas ligand; HTLV, human T cell leukemia virus; LUC, luciferase; P/I, PMA and ionomycin.
Received for publication June 19, 2002. Accepted for publication October 21, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B and p53. Mol. Cell. Biol. 19:3485.B and p53 signaling. J. Biol. Chem. 274:1879.B. Cytokine Growth Factor Rev. 12:207.[Medline]
B signaling pathway governs TRAIL gene expression and human T-cell leukemia virus-I Tax-induced T-cell death. J. Biol. Chem. 276:40385.
is mediated through competition for the C/H1 domain of CBP. J. Biol. Chem. 276:15720.
