This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
TAT-RasGAP317-326 Requires p53 and PUMA toSensitize Tumor Cells to Genotoxins
David Michod and Christian Widmann
Department of Physiology and Department of Cell Biology and Morphology,Lausanne University, Lausanne, Switzerland
AbstractAlthough chemotherapy has revolutionized cancer
treatment, the associated side effects induced by lack of
specificity to tumor cells remain a challenging problem.
We have previously shown that TAT-RasGAP317-326,
a cell-permeable peptide derived from RasGAP,
specifically sensitizes cancer cells to the action of
genotoxins. The underlying mechanisms of this
sensitization were not defined however. Here, we report
that TAT-RasGAP317-326 requires p53, but not the Ras
effectors Akt and extracellular signal-regulated kinase,
to mediate its tumor sensitization abilities. The
TAT-RasGAP317-326 peptide, although not modulating
the transcriptional activity of p53 or its phosphorylation
and acetylation status, nevertheless requires a
functional p53 cellular status to increase the sensitivity
of tumor cells to genotoxins. Genes regulated by p53
encode proapoptotic proteins, such as PUMA, and cell
cycle control proteins, such as p21. The ability of
TAT-RasGAP317-326 to sensitize cancer cells was found
to require PUMA but not p21. TAT-RasGAP317-326 did
not affect PUMA levels, however, but increased
genotoxin-induced mitochondrial depolarization and
caspase-3 activation. These results indicate that
TAT-RasGAP317-326 sensitizes tumor cells by activating
signals that intersect with the p53 pathway downstream
of, or at the level of, proapoptotic p53 target gene
products to increase the activation of the mitochondrial
death pathway. (Mol Cancer Res 2007;5(5):497–507)
IntroductionSince its introduction as a cancer therapyf50 years ago (1),
chemotherapy has remained a widely used and efficient
antitumor procedure. Most chemotherapeutic agents kill tumor
cells by inducing DNA damage and are therefore called
genotoxins. Ideally, genotoxins should only target cancer cells
to induce their demise with minimal collateral damage to
normal cells. In reality, however, the effectiveness of chemo-
therapy has suffered from a range of confounding factors,
including systemic toxicity due to a lack of specificity, rapid
drug metabolism, and both intrinsic and acquired drug
resistance. The efficiency of chemotherapy would therefore
be ameliorated by increasing the specificity of chemotherapeu-
tic agents toward cancer cells.
We have recently described a peptide derived from the
RasGAP protein that increases the sensitivity of cancer cells
to genotoxins. RasGAP, a regulator of Ras, bears two caspase-
3 cleavage sites (2, 3). At low levels of caspase activity,
RasGAP is cleaved into two fragments (4). The NH2-terminal
fragment (fragment N) seems to be a general blocker of
apoptosis downstream of caspase activation and is in fact
crucially required for the survival of stressed cells (5). At
higher levels of caspase activity, fragment N is further cleaved
into fragments N1 and N2, abrogating its antiapoptotic
activity (4, 6). Fragment N2, in contrast to fragment N,
potently sensitizes cancer cells toward genotoxin-induced
apoptosis (4, 6). A minimal sequence within fragment N2 that
can still sensitize tumor cells to the action of genotoxins has
been identified (7). This sequence has been rendered cell
permeable by linking it to the TAT48-57 peptide. This construct,
called TAT-RasGAP317-326, efficiently sensitizes cancer cells to
genotoxin-induced apoptosis. Importantly, TAT-RasGAP317-326does not sensitize nontumor cells (7). However, the molecular
mechanisms underlying the sensitizing properties of TAT-
RasGAP317–326 are still poorly understood.
Genotoxin-induced DNA damage leads to apoptosis mainly
in a p53-dependent manner (8, 9). At the top of the signaling
networks induced by DNA damage lie three related protein
kinases, DNA-dependent protein kinase, ataxia-telangiectasia
mutated, and ataxia-telangiectasia mutated and Rad3-related,
which orchestrate the damage response, sometimes in concert
and sometimes separately, by activating kinases that phosphor-
ylate p53 (8). The p53 protein transmits the apoptotic signal by a
complex mechanism involving, at least in part, its ability to
transactivate proapoptotic target genes, such as PUMA (10).
Here, we have assessed the role of the p53 pathway in the ability
of TAT-RasGAP317-326 to sensitize cancer cells to genotoxins.
ResultsTAT-RasGAP317-326 Does Not Modulate the Activity of Aktor the Extracellular Signal-Regulated Kinase Mitogen-Activated Protein Kinase PathwayRasGAP, depending on the circumstances, can regulate Ras
either negatively or positively (11, 12). Conceivably, therefore,
the TAT-RasGAP317-326 peptide could have an effect on Ras
activity. However, this peptide does not affect the Ras effector
pathways leading to nuclear factor-nB, c-Jun NH2-terminal
Received 8/15/06; revised 2/22/07; accepted 3/6/07.Grant support: Oncosuisse and Swiss National Science Foundation.The costs of publication of this article were defrayed in part by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.Requests for reprints: Christian Widmann, Department of Physiology, LausanneUniversity, Rue du Bugnon 7/9, 1005 Lausanne, Switzerland. Phone: 41-21-692-5123; Fax: 41-21-692-5255. E-mail: [email protected] D 2007 American Association for Cancer Research.doi:10.1158/1541-7786.MCR-06-0257
kinase, and p38 mitogen-activated protein kinase activation (7).
But as it has been shown that fragment N activates Akt (13) and
because the 317-326 sequence of RasGAP is born by fragment
N, we determined if the TAT-RasGAP317-326 peptide also
modulates the activity of Akt. Figure 1A shows that neither the
control HIV-TAT48-57 peptide nor the TAT-RasGAP317-326–
sensitizing peptide induced the phosphorylation of Akt.
Moreover, these peptides did not affect the ability of cisplatin
FIGURE 1. Akt and ERKactivation is not modulated byTAT-RasGAP317-326. U2OScells were treated for 16 h withthe indicated combinationsof 20 Amol/L HIV-TAT48-57,20 Amol/L TAT-RasGAP317-326,and 30 Amol/L cisplatin. Theextent of Akt (A) and ERK (B)activation was detected byWestern blot using antibodiesrecognizing the phosphorylat-ed (phospho ) forms of Akt andERK. The total levels of eachprotein were assessed usingantibodies recognizing allforms of Akt and ERK. Quan-titations of the bands are pre-sented below the Westernblots. Columns, mean of threeindependent experiments nor-malized to the values obtainedin untreated cells; bars, SD.NS, not significant.
FIGURE 2. TAT-RasGAP317-326 re-quires p53 to sensitize cancer cell linesto cisplatin-induced apoptosis. A. SAOSand U2OS cells were incubated withincreasing concentrations of cisplatin inthe presence or absence of 20 Amol/L ofHIV-TAT48-57 or TAT-RasGAP317-326. Theextent of apoptosis was scored 22 h later.B. HCT116 cells expressing or not p53were treated as in A. C and D. H1299cells encoding a p53 construct whoseexpression is negatively regulated bytetracycline were cultured with (+ tet ) orwithout (� tet) 1 Amol/L tetracycline. C.Cells were then treated as described in A.D. Expression of p53 was determined byWestern blot. E. The p53-repressibleH1299 cells cultured in the presence orabsence of tetracycline were transfectedwith an empty vector (pcDNA3) or a vectorexpressing the N2 fragment of RasGAP.The cells were then treated with increas-ing concentrations of cisplatin. The per-centage of cells undergoing apoptosiswas scored 22 h later. Points, mean ofthree independent determinations; bars,SD. Asterisks, significant differences be-tween the genotoxin-treated cells incubat-ed with TAT-RasGAP317-326 and those leftuntreated or incubated with HIV-TAT48-57.
to decrease the phosphorylation of Akt. These results indicate
that Akt is not a target of TAT-RasGAP317-326 and, conse-
quently, that Akt is probably not involved in the ability of this
peptide to sensitize cancer cells to genotoxins.
The extracellular signal-regulated kinase (ERK) mitogen-
activated protein kinase pathway is one of the main Ras effector
pathways (14-16) implicated in the increased proliferative rate
of tumor cells (17). As shown in Fig. 1B, neither the control
HIV-TAT48-57 peptide nor the TAT-RasGAP317-326–sensitizing
peptide activated the phosphorylation of ERKs. They were also
not able to modulate the effect of cisplatin on ERK
phosphorylation. These and earlier results (7) indicate that
TAT-RasGAP317-326–mediated enhancement of genotoxin-
induced apoptosis does not depend on Ras.
TAT-RasGAP317-326 Enhances Cisplatin-Induced Apopto-sis in a p53-Dependent MannerThe p53 transcription factor is activated in response to
genotoxic stress. It can induce cell cycle arrest, DNA repair, and
apoptosis (10, 18). Several approaches were used to determine
whether p53 was required for the TAT-RasGAP317-326 peptide
to favor genotoxin-induced apoptosis. First, the U2OS and
SAOS osteosarcoma cell lines were analyzed. The former
expresses p53, whereas the latter is deficient in this protein.1
Figure 2A shows that the p53-negative SAOS cells were not
sensitized by the TAT-RasGAP317-326 peptide to undergo
cisplatin-induced apoptosis, in contrast to the U2OS p53-
positive cells. Despite their common origin, these osteosarcoma
cell lines could differ in other aspects, and therefore, a second
approach was undertaken that used the wild-type p53-contain-
ing HCT116 colorectal carcinoma cell line and its p53-negative
variant obtained by somatic homologous recombination (19).
Figure 2B shows that the parental HCT116 cells were
efficiently sensitized by the TAT-RasGAP317-326 peptide to
cisplatin-induced death but that the variant lacking p53
remained unaffected by the peptide. To confirm these results,
we used the H1299 lung carcinoma cell line in which p53 can
be repressed by tetracycline (20). As shown in Fig. 2C, this
repressible cell line expresses p53 in the absence of tetracycline
but is totally devoid of p53 in the presence of the drug. Figure
2D shows that only in the condition allowing p53 expression
was this cell line sensitized by the TAT-RasGAP317-326 peptide
to apoptosis induced by cisplatin. These results indicate that
p53 is required for the sensitization property of the TAT-
RasGAP317-326 peptide. Consistent with this notion is the
observation that, in several tumor cell types, the TAT-
RasGAP317-326 peptide does not sensitize cell death induced
by staurosporine (Fig. 3), a stimulus known to kill cells
independently of p53 (21).
The TAT-RasGAP317-326 peptide is derived from the N2
fragment of RasGAP that is physiologically produced in cells
undergoing apoptosis. To determine whether this N2 fragment
also required p53 to sensitize tumor cells, the repressible
H1299 cells were transfected with a plasmid encoding frag-
ment N2 and subjected to increasing concentrations of
cisplatin. Figure 2E shows that fragment N2 requires p53 to
sensitize this cancer cell line to genotoxin-induced apoptosis.
This indicates that the TAT-RasGAP317-326 peptide and
fragment N2 from which it is derived regulate genotoxin-
induced death in a similar manner.
TAT-RasGAP317-326 Does Not Modulate p53 Transcrip-tional Activity and StabilizationIn wild-type p53-expressing cells, genotoxins induce stabi-
lization of p53 through inhibition of interaction between p53 and
Mdm2, an E3 ligase targeting p53 to proteasomal degradation.
This leads to an up-regulation of p53 cellular levels (compare
also lanes 1 and 2 in Fig. 4A; refs. 22-25). The increase in p53
levels and activity is accompanied by an increased transcription
of the p21 gene and a concomitant augmentation of p21 protein
levels (compare also lanes 1 and 2 in Fig. 4A; ref. 26). We
therefore investigated if TAT-RasGAP317-326 could modulate the
expression levels of p53 or alter its transcriptional activity. As
shown in Fig. 4A-C, cisplatin induced an increased expression
of p53 and its target gene product p21 independently of the
presence of the TAT-RasGAP317-326 peptide. This indicates that
the TAT-RasGAP317-326 peptide does not affect the stability
or the transcriptional activity of p53. This latter point was
FIGURE 3. TAT-RasGAP317-326 does not modulate staurosporine-induced tumor cell death. The cancer cell lines used in Fig. 2 were treatedwith increasing concentrations of staurosporine in the presence of20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326 during 22 h. The extentof apoptosis was then scored. Points, mean of three independentdeterminations; bars, SD.
cisplatin-induced cell death in the presence of TAT-Ras-
GAP317-326. These results indicate that the ability of TAT-
RasGAP317-326 to sensitize tumor cells is drastically affected
when the functionality of p53 to mediate genotoxin-induced
apoptosis is affected (even to limited extents).
To further characterize the resistance of the cells expressing
the p53 forms bearing the point mutations to the action of the
TAT-RasGAP317-326 peptide, their ability to modulate p53 and
p21 cellular levels in response to cisplatin was assessed in the
presence or absence of the peptide. In wild-type p53-expressing
cells, cisplatin induced an f20-fold increase in p53 levels. Incontrast, this induction was decreased to 3- to 5-fold in cells
expressing the p53 mutants (Fig. 6). Moreover, although the
p21 levels were increased by cisplatin in wild-type p53-
expressing cells, they remained unchanged in mutants 15,
33-37, and 9-15-33-37 or minimally increased in mutant 20
(Fig. 6). In none of the mutants were the levels of p53 and p21
modulated by the TAT-RasGAP317-326 peptide (Fig. 6).
Taken together, the results presented in Figs. 5 and 6 indicate
that the p53 mutants used here, despite being still more
F IGURE 4 . TAT -Ra s -GAP317-326 does not modulatethe expression and the tran-scriptional activity of p53 inU2OS cells. A. Western blotanalysis of p53 and p21 expres-sion in U2OS cell treated with15 Amol/L cisplatin in the pres-ence or absence of 20 Amol/Lof HIV-TAT48-57 or TAT-Ras-GAP317-326 for 16 h. Thecorresponding quantitations arepresented in B and C. Theywere done on three independentWestern blots and normalizedagainst the positive controls(cells without treatment). D.U2OS cells were transfectedwith 0.1 Ag of a firefly lucifer-ase reporter plasmid bearingp53-binding elements and with0.5 Ag of a plasmid encodingthe Renilla luciferase. Thecells were treated 2 d later with15 Amol/L cisplatin in the pres-ence or absence of 20 Amol/Lof the HIV-TAT48-57 or TAT-RasGAP317-326 peptides during16 h. Firefly luciferase activitynormalized to the Renilla lucif-erase activity and expressed asfold increase of the basal p53activity obtained in control un-treated cells. Columns, meanof three independent determi-nations; bars, SD; NS, not sig-nificant.
sensitive to cisplatin-induced apoptosis compared with p53�/�
cells, are strongly affected in their ability to regulate the trans-
cription of at least one of its target genes (p21) in response to
DNA damage. This could explain why TAT-RasGAP317-326 was
not able to sensitize the mutant p53-expressing cells to cisplatin.
TAT-RasGAP317-326 Does Not Affect p53 Phosphorylationand Acetylation
The results described above indicate that TAT-RasGAP317-326does not modulate p53 transcriptional activity. The peptide
might, however, affect p53 posttranslational modifications that
could modulate p53 activity in ways that could not have been
detected in the experimental conditions used above. The p53
protein can be phosphorylated on about 15 serine and
threonine residues and acetylated on 4 lysine residues (29).
We therefore assessed, using commercially available anti-
bodies, the capacity of TAT-RasGAP317-326 to modulate the
phosphorylation of p53 at some of the serine residues, as well
as its acetylation on Lys382. Figure 7 shows that the increase
in p53 phosphorylation at Ser15, Ser20, Ser37, and Ser46
induced by cisplatin was not affected by TAT-RasGAP317-326.
Similarly, the acetylation of p53 at position 382 induced by
the genotoxin was not changed in the presence of the peptide
(Fig. 7, bottom). This suggests that TAT-RasGAP317-326 does
not modulate the posttranslational modifications occurring on
p53, although we cannot exclude that phosphorylation and
FIGURE 5. Cells expressing p53 mutants are not sensitized by TAT-RasGAP317-326. A. HCT116 p53+/+ or HCT116 p53�/� cells stably infected withlentiviruses encoding wild-type (WT) or the indicated mutant forms of p53 were lysed, and the expression of p53 and p21 was assessed by Western blotanalysis. B. Alternatively, these cells were treated with a low (20 Amol/L) or a high (100 Amol/L) concentration of cisplatin in the presence or absenceof 20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326 for 22 h. The percentage of cells undergoing apoptosis was then scored. C. HCT116 p53+/+ or HCT116p53�/� cells stably infected with lentiviruses encoding wild-type or mutant forms of p53 bearing alanine substitutions at the indicated phosphorylation sites weresubmitted to increasing concentrations of cisplatin for 22 h, and the percentage of apoptosis was then scored. Asterisks, statistically significant differencebetween the p53 9-15-33-37 mutant-expressing cells and the cells expressing the other p53 mutants. The difference between the p53 mutant-expressing cellsand either wild-type p53-expressing cells or p53�/� cells was statistically significant at cisplatin concentrations of z30 Amol/L (not indicated on the figure).D. Alternatively, these cells were treated with low (20 Amol/L) and intermediate (50 Amol/L) cisplatin concentrations in the presence or absence of theHIV-TAT48-57 or TAT-RasGAP317-326 peptides as indicated in Fig. 4B. Columns, mean of three independent determinations; bars, SD. Asterisks, significantdifferences between the genotoxin-treated cells incubated with TAT-RasGAP317-326 and those left untreated or incubated with the HIV-TAT48-57 peptide. NS,not significant.
acetylation sites other than those tested here are regulated by
the RasGAP peptide.
TAT-RasGAP317-326 – Induced Sensitization of TumorCells Requires PUMA but not p21The results presented above suggest that some p53 target
genes may be required for the sensitization of tumor cells
by TAT-RasGAP317-326. We therefore assessed the implica-
tion of two important p53 target genes, the cell cycle inhi-
bitor p21cip1 and the Bcl2 family member PUMA , in this
response. Figure 8A shows that HCT116 cells lacking p21
were still sensitized to cisplatin-induced apoptosis by the
TAT-RasGAP317-326 peptide. As p21 is a critical mediator of
p53-induced cell cycle arrest (30), this suggests that the TAT-
RasGAP317-326 peptide does not mediate its effect through
modulation of the cell cycle. In contrast, HCT116 cells devoid
of PUMA were completely resistant to the sensitization
mediated by TAT-RasGAP317-326 (Fig. 8B). We next assessed
if TAT-RasGAP317-326 was able to modulate the expression of
PUMA. As shown in Fig. 9A and B, PUMA expression was
up-regulated by cisplatin. However, the TAT-RasGAP317-326peptide did not further modulate PUMA protein levels in cells
whether treated with the genotoxin or not. These results show
that TAT-RasGAP317-326 requires the expression of PUMA to
FIGURE 6. p53 and p21 expression in HCT116 cells expressing p53 mutants bearing alanine substitutions at NH2-terminal p53 phosphorylation sites. A.Western blot analysis of p53 and p21 expression of the cells described in Fig. 5 after treatment with 50 Amol/L cisplatin (+) in the presence or absence of20 Amol/L of the HIV-TAT48-57 or TAT-RasGAP317-326 peptides during 16 h. B. Corresponding quantitative analyses done on three independent Westernblots and normalized against the values obtained in untreated control cells. Columns, mean; bars, SD.
have its effect but does not modulate its expression. We also
tested whether Mcl-1, an antiapoptotic Bcl2 family member
that can be degraded after DNA damage (31, 32), could be
negatively affected by TAT-RasGAP317-326, which could
explain the sensitization induced by the peptide on geno-
toxin-treated cells. In our experimental conditions, however,
cisplatin did not induce reduction in Mcl-1 levels (Fig. 9C,
compare lanes 1 and 2). Surprisingly, TAT-RasGAP317-326induced an f2-fold induction of Mcl-1 in HCT116 cells
(Fig. 9C). The reason for this increase is not known at the
present time. However, the combination of cisplatin and TAT-
RasGAP317-326 led to a down-regulation of Mcl-1 protein levels
(Fig. 9C). This could be the result of increased caspase activity
(see Fig. 10B) as Mcl-1 is a caspase substrate (33-36).
Nevertheless, the fact that TAT-RasGAP317-326 by itself does
not induce a decrease of Mcl-1 suggests that Mcl-1 is not
playing a direct role in the ability of the peptide to sensitize
cancer cells to genotoxin-induced death.
As PUMA is a BH3-only Bcl-2 family member that favors
mitochondrial-dependent apoptotic responses (37), the results
described above indicate that TAT-RasGAP317-326 increases
the sensitivity of tumor cells to genotoxin by enhancing
the intrinsic, mitochondrial-dependent, apoptotic pathway.
Consistent with this interpretation is the observation that the
FIGURE 7. p53 phosphorylation (at Ser15,Ser20, Ser37, and Ser46) and acetylation are notmodulated by the TAT-RasGAP317-326 peptide.HCT116 cells treated with 20 Amol/L cisplatin inthe presence or absence of 20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326 for 16 h. The cellswere then lysed, and the extent of phosphorylationat the indicated serine sites, as well as the extent ofacetylation, was assessed by Western blot analy-sis. The quantitative analyses were done on threeindependent Western blots and normalized againstthe values obtained in untreated cells. Columns,mean; bars, SD; NS, not significant.
TAT-RasGAP317-326 peptide allowed cisplatin to induce a
stronger drop in mitochondrial membrane potential (Fig. 10A)
and an increased activation of caspase-3 (Fig. 10B).
DiscussionRasGAP has first been identified as a Ras regulator (38), but
this protein has also been shown to bear two caspase-3 cleavage
sites and cleaved during apoptosis (2, 3, 5). The NH2-terminal
fragment resulting from the first cleavage event on RasGAP
(called fragment N; ref. 4) displays strong antiapoptotic
properties by activating the Ras/phosphatidylinositol 3-kinase
Akt pathway (13). The ability of fragment N to stimulate Akt
does not result in nuclear factor-nB activation (13) and neitherdoes the TAT-RasGAP317-326 peptide require nuclear factor-nBto mediate its sensitization effects (7). Here, we have extended
these observations by showing that TAT-RasGAP317-326 does
not modulate Akt or ERK activity, which indicates that Ras
effector pathways are not involved in the tumor sensitization
properties of the TAT-RasGAP317-326 peptide.
On the other hand, our results clearly show that TAT-
RasGAP317-326 requires p53 to increase the genotoxin sensitivity
of tumor cells but does not so by modulating its transcriptional
activity or its phosphorylation and acetylation status. The TAT-
RasGAP317-326 peptide does, however, require that p53 is fully
functional in its ability to sense DNA damage via phosphory-
lation of its NH2 terminus by kinases, such as ataxia-
telangiectasia mutated, ataxia-telangiectasia mutated and
Rad3-related, or DNA-dependent protein kinase (34). Indeed,
all the p53 mutants we have used in the present study that bore
mutations at these phosphorylation sites were incapable of
allowing TAT-RasGAP317-326 to mediate its sensitization effect.
Presumably, the signals initiated by the TAT-RasGAP317-326peptide integrate with the p53 pathway downstream of one or
several of the p53 target genes. Consistent with this notion is the
observation that the p53 target PUMA is crucially required for
the TAT-RasGAP317-326 peptide to induce tumor cell sensitiza-
tion (Fig. 8B). PUMA levels were not modulated by the peptide,
suggesting that neither the stability of PUMA nor the
transcription rate of its gene or the translation rate of its mRNA
was affected by the peptide. Although we did not detect any
change in the migration pattern of PUMA on polyacrylamide
gels that sometimes happen as a result of posttranslational modi-
fications, it cannot be ruled out that the TAT-RasGAP317-326peptide induced such modifications on PUMA. To our knowl-
edge, however, posttranslational modifications of PUMA have
yet to be described. If the peptide does not directly affect
PUMA levels, it might regulate its ability to stimulate
apoptosis. In this context, it is interesting to note that PUMA
can displace p53 from Bcl-XL, allowing p53 to induce
mitochondrial permeabilization and apoptosis (39). It could
therefore be anticipated that TAT-RasGAP317-326 modulates this
concerted action of p53 and PUMA at the mitochondria level
to increase the sensitivity of tumor cells to the action of
genotoxins. This is consistent with our observation that TAT-
RasGAP317-326 augmented mitochondrial depolarization and
the ensuing caspase activation induced by cisplatin (Fig. 10).
Our results also show that the TAT-RasGAP317-326 peptide
does not facilitate tumor cell death by modulating p21 levels, as
cells bearing deletions of the p21 gene are as sensitive to
cisplatin in the presence of TAT-RasGAP317-326 as p21-positive
cells (Fig. 8A). Recently, the rapamycin derivative mammalian
target of rapamycin inhibitor everolimus has been shown to
sensitize tumor cells to genotoxins in a manner that required
inhibition of p53-induced p21 expression (40). The mechanism
by which TAT-RasGAP317-326 sensitizes tumor cells is therefore
different from those resulting from the inhibition of the
mammalian target of rapamycin pathway. This study and ours
point out to the central role of p53 in sensitizing tumor cells to
the action of genotoxins. They also indicate that there are more
than one p53 target that can be modulated to enhance the p53-
dependent cell death of cancer cells, which may eventually lead
to the development of parallel strategies to increase the efficacy
of genotoxic drugs to specifically target tumors.
Materials and MethodsCells and TransfectionThe U2OS and SAOS osteosarcoma cell lines (LGC Pro-
mochem; American Type Culture Collection) were maintained in
FIGURE 8. TAT-RasGAP317-326 requires PUMA, but not p21, tosensitize HCT116 cells to cisplatin-induced apoptosis. A. HCT116 p21+/+
or HCT116 p21�/� cells were treated with 20 Amol/L cisplatin in thepresence or absence of 20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326for 22 h, and the extent of apoptosis was scored. Columns, mean of threeindependent determinations; bars, SD. B. HCT116 PUMA+/+ or HCT116PUMA�/� cells were treated and analyzed as described in A. Asterisks,significant differences between the genotoxin-treated cells incubated withTAT-RasGAP317-326 and those left untreated or incubated with the HIV-TAT48-57 peptide.
DMEM (Sigma) containing 10% FCS (Sigma) at 37jC and 5%CO2. HCT116 p53
+/+, HCT116 p53�/�, HCT116 PUMA�/�,and HCT116 p21�/� have been described earlier (41, 42) andwere cultured as described above. TheH1299 cell line encoding a
p53 construct under the control of a tetracycline-repressible
promoter (Tet-off system) was described earlier (20). Experi-
ments using cells were done in 2 mL in six-well plates con-
taining 1.5 � 105 cells per well for U2OS, SAOS, and H1299
cells or 4� 105 cells per well for HCT116 cells. H1299 cells weretransfected using LipofectAMINE 2000 (Invitrogen) with 0.5 Aggreen fluorescent protein–expressing plasmid and with 1 Agcontrol plasmid or 1 Ag HA-N2–expressing plasmid. Briefly,1.5 � 105 cells, plated the previous day, were incubated for
5 h with a DNA (1.5 Ag)-LipofectAMINE 2000 (2 AL) mixturein 1 mL of DMEM without serum at 37jC in 5% CO2. When
cisplatin was used, it was added at the time the medium with
serum was added back to the transfected cells. The cells were
split the day before the treatment at a concentration of 1.5 � 105
per well. U2OS cells were transfected using the calcium/phos-
phate precipitation procedure (7, 43) using 0.1 Ag pRL-TK, avector encoding the Renilla reniformis luciferase from Promega,
and 0.5 Ag p53.luc, a reporter plasmid bearing the firefly luci-ferase cDNA under the control of p53-responsive elements (44).
Lentiviral InfectionThe lentiviral vectors expressing wild-type and mutant p53
proteins have been described previously (44). Lentiviruses were
produced as described (5). Lentivirus-containing supernatants
were collected 36 h after transfection, 0.45 Am filtered, and
frozen at �80jC. The amount of virus leading to infection of30% of the HCT116 cells was chosen. The cells were further
selected in puromycin-containing medium to ensure that each
cell expressed the lentivirus-encoded proteins.
Peptide SynthesisThe HIV-TAT48-57 (GRKKRRQRRR) and TAT-RasGAP317-326
(GRKKRRQRRRGGWMWVTNLRTD) peptides were synthe-
sized at the Institute of Biochemistry (University of Lausanne,
Lausanne, Switzerland) using N-(9-fluorenyl)methoxycar-
bonyl technology, purified by high-performance liquid chro-
matography, and tested by mass spectrometry. Peptides were
diluted at a concentration of 1 mmol/L in H2O and stored
at �20jC.
ChemicalsCisplatin was from Sigma and was diluted in PBS at a final
concentration of 1 mmol/L and stored at �80jC. Hoechst
FIGURE 9. TAT-RasGAP317-326 does not modulate the expression of PUMA to sensitize the HCT116 cancer cell line to cisplatin-induced apoptosis.A. HCT116 PUMA+/+ or HCT116 PUMA�/� cells were treated with 20 Amol/L cisplatin in the presence or absence of 20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326. The cells were lysed 16 h later, and the expression of PUMA was assessed by Western blot analysis. B. Corresponding quantitativeanalyses done on three independent Western blots and normalized against the values obtained in untreated control cells. Columns, mean; bars, SD.C. U2OS cells were treated with 15 Amol/L cisplatin in the presence or absence of 20 Amol/L of HIV-TAT48-57 or TAT-RasGAP317-326. The cells were lysed16 h later, and the expression of Mcl-1 was assessed by Western blot analysis. Quantitations of the bands are presented below the Western blots. Columns,mean of three independent experiments normalized to the values obtained in untreated cells; bars, SD.
and total ERK (Cell Signaling Technology) were diluted
1:1,000 in 5% bovine serum albumin in TBS. The mono-
clonal antibody against p53 DO1 (gift from Dr. Richard
Iggo, Bute Medical School, University of St. Andrews, Fife,
Scotland), the polyclonal antibody against p53 (Cell
Signaling Technology), the antibody against p21 (Cell
Signaling Technology), and the antibody against PUMA
(Axxora) were diluted 1:1,000 in 5% nonfat dry milk in
TBS. The antibody recognizing the active, cleaved, form of
caspase-3 was from Cell Signaling Technology. It was used
at a 1:1,000 dilution in 5% milk in TBS. The antibody
specific for Mcl-1 was from Sigma-Aldrich. It was used at a
1:1,000 dilution in 5% milk in TBS. The antibodies specific
for the various p53 serine phosphorylation sites and for the
acetylated form of p53 were from Cell Signaling Technology.
They were used at a 1:1,000 dilution in 5% bovine serum
albumin in TBS. Quantitation was done using the Odyssey IR
imaging software.
Mitochondrial Membrane Potential MeasurementMitochondrial membrane depolarization was assessed
using the JC-1 mitochondrial membrane potential sensor
(Sigma). This compound exhibits potential-dependent accu-
mulation in mitochondria. On excitation at 488 nm, JC-1
fluoresces at 529 nm (green) in the cytoplasm and at 590 nm
(red) in the mitochondria. Mitochondrial depolarization is
indicated by a decrease in the red/green fluorescence
intensity ratio. The cells were processed as follows. The
medium of the cell culture (with its associated floating cells)
was collected. The remaining adherent cells were washed
with 1 mL PBS and collected. The adherent cells were
detached with 500 AL of a trypsin/EDTA solution (5 g/L
FIGURE 10. TAT-RasGAP317-326 increases the apoptotic signalsdownstream of PUMA. U2OS cells were treated with 20 Amol/L cisplatin inthe presence or absence of 20 Amol/L HIV-TAT48-57 or TAT-RasGAP317-326for 16 h. They were then processed as follows. A. The mitochondrialpotential of the cells was measured as described in Materials and Methods.Insets, mean F SD (from three independent experiments) of the percen-tage of cells having lost their mitochondrial potential. The difference betweencells treated with cisplatin alone or in combination with HIV-TAT48-57 withcells treated with cisplatin and TAT-RasGAP317-326 was statisticallysignificant. B. The cells were lysed, and the cleavage of caspase-3 wasassessed by Western blot analysis. The experiment was done two moretimes with similar results.
porcine trypsin and 2 g/L EDTA; Sigma-Aldrich) that was
pooled with the collected medium and PBS. The cells were
pelleted by centrifugation and resuspended in 1 mL culture
medium and incubated with 2 Amol/L JC-1 for 30 min at
37jC in the dark. The cells were then kept on ice before
being analyzed with a FACScan apparatus (BD Biosciences).
In the experiments shown, FL-1 corresponds to the green
channel and FL-2 to the red channel.
Statistical AnalysisAll the statistical analyses were done with Microsoft Excel
(XP edition) using the two-tailed unpaired Student’s t test.
Significance is indicated by an asterisk when P < 0.05/n , where
P is the probability derived from the t test analysis and n is the
number of comparisons done (Bonferroni correction).
AcknowledgmentsWe thank Dr. Bert Vogelstein for the gift of the various HCT116 cell lines, Dr.Richard Iggo and Dr. Mathias Kaeser for suggestions and comments and for thegift of the plasmids encoding p53 and its mutants, and Dr. Peter Clarke for hiscritical reading of the manuscript.
References1. Chabner BA, Roberts TG, Jr. Timeline: chemotherapy and the war on cancer.Nat Rev Cancer 2005;5:65 –72.
2. Widmann C, Gibson S, Johnson GL. Caspase-dependent cleavage of signalingproteins during apoptosis. A turn-off mechanism for anti-apoptotic signals. J BiolChem 1998;273:7141– 7.
3. Wen LP, Madani K, Martin GA, Rosen GD. Proteolytic cleavage of RasGTPase-activating protein during apoptosis. Cell Death Differ 1998;5:729 –34.
4. Yang J-Y, Widmann C. Antiapoptotic signaling generated by caspase-inducedcleavage of RasGAP. Mol Cell Biol 2001;21:5346–58.
5. Yang J-Y, Michod D, Walicki J, et al. Partial cleavage of RasGAP by caspasesis required for cell survival in mild stress conditions. Mol Cell Biol 2004;24:10425 –36.
6. Yang J-Y, Walicki J, Abderrahmani A, et al. Expression of an uncleavable N-terminal RasGAP fragment in insulin-secreting cells increases their resistancetoward apoptotic stimuli without affecting their glucose-induced insulin secretion.J Biol Chem 2005;280:32835 –42.
7. Michod D, Yang JY, Chen J, Bonny C, Widmann C. A RasGAP-derived cellpermeable peptide potently enhances genotoxin-induced cytotoxicity in tumorcells. Oncogene 2004;23:8971–8.
8. Norbury CJ, Zhivotovsky B. DNA damage-induced apoptosis. Oncogene2004;23:2797–808.
9. Zabkiewicz J, Clarke AR. DNA damage-induced apoptosis: insights from themouse. Biochim Biophys Acta 2004;1705:17 –25.
10. Michalak E, Villunger A, Erlacher M, Strasser A. Death squads enlistedby the tumour suppressor p53. Biochem Biophys Res Commun 2005;331:786– 98.
11. Kulkarni SV, Gish G, Van der Geer P, Henkemeyer M, Pawson T. Role ofp120 ras-GAP in directed cell movement. J Cell Biol 2000;149:457 –70.
12. Tocque B, Delumeau I, Parker F, Maurier F, Multon MC, Schweighoffer F.Ras-GTPase activating protein (GAP): a putative effector for Ras. Cell Signal1997;9:153 –8.
13. Yang J-Y, Widmann C. The RasGAP N-terminal fragment generated bycaspase cleavage protects cells in a Ras/PI3K/Akt-dependent manner that does notrely on NFnB activation. J Biol Chem 2002;277:14641–6.
14. Widmann C, Gibson S, Jarpe MB, Johnson GL. MAPK pathways:conservation of a three-kinase module from yeast to man. Physiol Rev 1999;79:143 –80.
15. Pearson G, Robinson F, Beers GT, et al. Mitogen-activated protein (MAP)kinase pathways: regulation and physiological functions. Endocr Rev 2001;22:153– 83.
16. Sears RC, Nevins JR. Signaling networks that link cell proliferation and cellfate. J Biol Chem 2002;277:11617–20.
17. Hoshino R, Chatani Y, Yamori T, et al. Constitutive activation of the 41-/43-kDa mitogen-activated protein kinase signaling pathway in human tumors.Oncogene 1999;18:813–22.
18. Vousden KH. Activation of the p53 tumor suppressor protein. BiochimBiophys Acta 2002;1602:47 –59.
19. Bunz F, Dutriaux A, Lengauer C, et al. Requirement for p53 and p21 tosustain G2 arrest after DNA damage. Science 1998;282:1497 –501.
20. Saller E, Tom E, Brunori M, et al. Increased apoptosis induction by 121Fmutant p53. EMBO J 1999;18:4424–37.
21. Shao RG, Shimizu T, Pommier Y. 7-Hydroxystaurosporine (UCN-01)induces apoptosis in human colon carcinoma and leukemia cells independentlyof p53. Exp Cell Res 1997;234:388 –97.
22. Honda R, Tanaka H, Yasuda H. Oncoprotein MDM2 is a ubiquitin ligase E3for tumor suppressor p53. FEBS Lett 1997;420:25–7.
23. Haupt Y, Maya R, Kazaz A, Oren M. Mdm2 promotes the rapid degradationof p53. Nature 1997;387:296 –9.
24. Kubbutat MH, Jones SN, Vousden KH. Regulation of p53 stability byMdm2. Nature 1997;387:299 –303.
25. Lane DP, Hall PA. MDM2—arbiter of p53’s destruction. Trends Biochem Sci1998;22:372–4.
26. Li Y, Jenkins CW, Nichols MA, Xiong Y. Cell cycle expression and p53regulation of the cyclin-dependent kinase inhibitor p21. Oncogene 1994;9:2261– 8.
27. Saito S, Yamaguchi H, Higashimoto Y, et al. Phosphorylation siteinterdependence of human p53 post-translational modifications in response tostress. J Biol Chem 2003;278:37536–44.
28. Chao C, Herr D, Chun J, Xu Y. Ser18 and 23 phosphorylation isrequired for p53-dependent apoptosis and tumor suppression. EMBO J 2006;25:2615–22.
29. Bode AM, Dong Z. Post-translational modification of p53 in tumorigenesis.Nat Rev Cancer 2004;4:793 –805.
30. El Deiry WS, Tokino T, Velculescu VE, et al. WAF1, a potential mediator ofp53 tumor suppression. Cell 1993;75:817–25.
31. Nijhawan D, Fang M, Traer E, et al. Elimination of Mcl-1 is required for theinitiation of apoptosis following ultraviolet irradiation. Genes Dev 2003;17:1475– 86.
32. Zhuang J, Brady HJ. Emerging role of Mcl-1 in actively counteracting BH3-only proteins in apoptosis. Cell Death Differ 2006;13:1263 –7.
33. Herrant M, Jacquel A, Marchetti S, et al. Cleavage of Mcl-1 by caspasesimpaired its ability to counteract Bim-induced apoptosis. Oncogene 2004;23:7863– 73.
34. Michels J, O’Neill JW, Dallman CL, et al. Mcl-1 is required for Akata6B-lymphoma cell survival and is converted to a cell death molecule byefficient caspase-mediated cleavage. Oncogene 2004;23:4818 – 27.
35. Clohessy JG, Zhuang J, Brady HJ. Characterisation of Mcl-1 cleavage duringapoptosis of haematopoietic cells. Br J Haematol 2004;125:655 –65.
36. Weng C, Li Y, Xu D, Shi Y, Tang H. Specific cleavage of Mcl-1 by caspase-3in tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-inducedapoptosis in Jurkat leukemia T cells. J Biol Chem 2005;280:10491 –500.
37. Yu J, Zhang L, Hwang PM, Kinzler KW, Vogelstein B. PUMA induces therapid apoptosis of colorectal cancer cells. Mol Cell 2001;7:673– 82.
38. Trahey M, McCormick F. A cytoplasmic protein stimulates normal N-ras p21GTPase, but does not affect oncogenic mutants. Science 1987;238:542 –5.
39. Chipuk JE, Bouchier-Hayes L, Kuwana T, Newmeyer DD, Green DR.PUMA couples the nuclear and cytoplasmic proapoptotic function of p53.Science 2005;309:1732–5.
40. Beuvink I, Boulay A, Fumagalli S, et al. The mTOR inhibitor RAD001sensitizes tumor cells to DNA-damaged induced apoptosis through inhibition ofp21 translation. Cell 2005;120:747– 59.
41. Yu J, Wang Z, Kinzler KW, Vogelstein B, Zhang L. PUMA mediates theapoptotic response to p53 in colorectal cancer cells. Proc Natl Acad Sci U S A2003;100:1931–6.
42. Waldman T, Kinzler KW, Vogelstein B. p21 is necessary for the p53-mediated G1 arrest in human cancer cells. Cancer Res 1995;55:5187 –90.
43. Jordan M, Schallhorn A, Wurm FM. Transfecting mammalian cells:optimization of critical parameters affecting calcium-phosphate precipitateformation. Nucleic Acids Res 1996;24:596–601.
44. Kaeser MD, Pebernard S, Iggo RD. Regulation of p53 stability and functionin HCT116 colon cancer cells. J Biol Chem 2004;279:7598–605.