Download Definition of a p53 transactivation function-deficient mutant

Oncogene (1999) 18, 2405 ± 2410
ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00
http://www.stockton-press.co.uk/onc
SHORT REPORT
De®nition of a p53 transactivation function-de®cient mutant and
characterization of two independent p53 transactivation subdomains
Corinne Venot1, Michel Maratrat1, VeÂronique Sierra1, Emmanuel Conseiller1 and
Laurent Debussche*,1
1
Centre de Recherche de Vitry-Alfortville, RhoÃne-Poulenc Rorer, 13 quai Jules Guesde, 94403 Vitry sur Seine Cedex, France
The wild-type protein product of the p53 tumor
suppressor gene can activate transcription of genes which
are involved in mediating either growth arrest, e.g.
WAF1 or apoptotis, e.g. BAX and PIG3. Additionally,
p53 can repress a variety of promoters, which, in turn,
may be responsible for the functional activities exhibited
by p53. This study shows that the Q22, S23 double
mutation, which is known to inactivate a p53 transactivation subdomain located within the initial 40
residues of the protein, while abrogating transactivation
from the WAF1 promoter, only attenuates apoptosis
triggering, transactivation from other p53-responsive
promoters and repression of promoters by p53. The
Q53, S54 double mutation, which inactivates another p53
transactivation subdomain situated between amino acids
43 and 73 results in attenuation of all of the aforementioned p53 activities. In contrast to the Q22, S23
double mutation, this latter mutation set does not alter
mdm-2-mediated inhibition and degradation of p53.
Finally, mutation of all four residues results in complete
abrogation of every p53 activity mentioned above.
Keywords: p53; apoptosis; transactivation; repression;
PIG3
p53 is a critical tumor suppressor gene, which can
respond to multiple signals of cellular alarm, by
inducing either growth arrest or apoptosis (reviewed
in Bates and Vousden, 1996; Gottlieb and Oren, 1996;
Ko and Prives, 1996; Hansen and Oren, 1997; Levine,
1997). p53 protein is a transcriptional transactivator
and induces speci®c target genes by binding to short
genomic sequences, called p53-responsive elements
(p53RE), which ®t a consensus sequence to variable
extents (Funk et al., 1992; El-Deiry et al., 1992).
Sequence speci®c transactivation (SST) by p53 has
been shown to be dependent upon three major
independent domains: (i) a central core domain
responsible for sequence-speci®c DNA binding (amino
acid residues 102 ± 292) (Bargonetti et al., 1993;
Pavletich et al., 1993; Cho et al., 1994); (ii) a Nterminal transactivation domain responsible for recruitment of transcription machinery (amino acid residues
1 ± 40) (Unger et al., 1992; Lin et al., 1994; Thut et al.,
1995; Lu and Levine, 1995); and (iii) a C-terminal
*Correspondence: L Debussche
Received 23 June 1998; revised 28 October 1998; accepted 29 October
1998
oligomerization domain (amino acid residues 319 ± 360)
(Sturzbecher et al., 1992; Clore et al., 1994; Je€rey et
al., 1995). The importance of SST in p53 tumor
suppressor activity is underscored by the fact that the
vast majority of p53 mutations derived from tumors
map within the sequence-speci®c DNA binding domain
(Hollstein et al., 1991; Pavletich et al., 1993; Cho et al.,
1994). Growth arrest by p53 seems clearly due to SST
of several p53 target genes, such as WAF1, GADD45
and cyclin G (reviewed in Bates and Vousden, 1996;
Gottlieb and Oren, 1996; Ko and Prives, 1996; Hansen
and Oren, 1997; Levine, 1997). However, many reports
suggest that p53 can mediate apoptosis through either
transcriptional regulation-dependent or independent
pathways. The transcriptional regulation-dependent
pathway involves induction by p53 of a growing list
of p53 target genes which are involved in apoptosis
triggering, such as BAX (Miyashita and Reed, 1995),
FAS/APO1 (Owen-Schaub et al., 1995), IGF-BP3
(Buckbinder et al., 1995), Cathepsin D (Wu et al.,
1998), Killer/DR5 (Wu et al., 1997), PAG608 (Israeli et
al., 1997) and PIG3 (Polyak et al., 1997). On the other
hand, the transcriptional regulation-independent pathway remains largely obscure (Caelles et al., 1994;
Wagner et al., 1994) and much of the evidence
describing its existence stems from studies with SSTde®cient p53 mutants, most especially the p53
transactivation domain mutant Q22, S23 (M22/23)
which showed them to be still capable of inducing
apoptosis in certain tumor-derived cell lines (YonishRouach et al., 1994; Haupt et al., 1995, 1996; Chen et
al., 1996).
Firstly, we con®rmed that M22/23 retains growth
suppression and apoptotic activities in tumor cells. In a
colony formation assay performed with H1299 (Figure
1) and Saos-2 (data not shown) cell lines, which are
both p537/7, M22/23 displays a phenotype which is
intermediate between the inactive tumor-derived
mutant, D281G, and p53 wild-type (p53 WT). This
assay enables growth suppression in its most general
sense to be measured as it scores for either p53dependent growth arrest and/or apoptosis in tumor
cells that receive the p53 protein. In the case of M22/
23, the remaining growth suppression e€ect results in
great part from apoptosis as this mutant retains
signi®cant apoptotic activity (Figure 2, columns 1 ± 3)
but no detectable growth arrest activity (data not
shown), which is consistent with previous reports
(Chen et al., 1996, Walker and Levine, 1996).
Recent studies have demonstrated the existence of
functionally di€erent p53RE, which could be classi®ed
according to their anities for p53 WT and certain
non-apoptotic p53 mutants (Rowan et al., 1996;
Characterization of p53 transactivation domains
C Venot et al
2406
Figure 1 M22/23 and M53/54 retains growth suppression activity while p53QM was completely lost it. p537/7 human lung
carcinoma cell line H1299 were seeded at 1.5 105 cells per multi-well six culture testplate and transfected with 1.5 mg of p53
expression plasmid (upper panel) or 0.75 mg of p53 expression plasmid+0.75 mg of pcDNA3 empty vector (lower panel), preincubated for 30 min with 5 ml lipofectAMINE reagent (GIBCO ± BRL, Gaithersburg, MD, USA). Selection of drug-resistant
colonies was performed as follows: 24 h after transfection, cells were seeded in 10-cm plates, cultured in 400 mg/ml G418 containing
medium for 2 ± 3 weeks and then colored with fuchsin
Figure 2 FACS analysis of apoptotic activity of p53 mutants.
Cells were analysed as described previously (Conseiller et al.,
1998) with some modi®cations. Brie¯y, 48 h after transfection,
supernatants were collected and anchored cells were trypsinized.
Floating and trypsinized cells were combined, ®xed in 1%
formalin and permeabilized with Triton 0.1%. They were then
incubated with PAb 240 antibody (which recognized both wild
type and mutants p53 under our denaturating ®xation conditions,
data not shown) and subsequently with Goat anti-mouse FITC
(Immunotech, Marseille, France). They were ®nally counterstained with 10 mg/ml Propidium iodide (PI) in 1 mg/ml boiled
Rnase and analysed on a FACS Epics Elite ESP II (Coulter,
Miami, FL, USA). Transfection eciencies were evaluated as the
percentages of positive cells for PAb 240. In each independent
experiment, percentages were very similar for all ®ve p53 proteins
and ranged between 30 and 60% from one to another experiment.
Percentage of apoptotic cells were quanti®ed on the basis of their
positivity for Pab 240 and their relative DNA content (sub G1).
This ®gure is a compilation of three independent experiments
Ludwig et al., 1996; Friedlander et al., 1996; Venot et
al., 1998). We thus reinvestigated SST by M22/23 on a
larger set of promoters than the ones used by others to
establish its SST-de®ciency. By transient transfection
experiments using luciferase reporter constructs (Figure
3, columns 1 ± 3), we ®rstly showed that M22/23 was
unable to transactivate a construct containing the
WAF1 promoter. This inactivity was observed at
optimal plasmid concentration as indicated in the
titration curve of p53 WT inputs inserted in the upper
right corner of Figure 3, but also at largely higher
M22/23 plasmid concentrations (data not shown). On
the other hand, M22/23 induced a low but signi®cant
expression of reporter gene from the BAX, MDM2
and, interestingly, PIG3 promoters. Transactivation
from this last promoter is strongly correlated with p53dependent apoptosis in colon and lung cancer cells
and, especially, in H1299 cells (Polyak et al., 1997;
Venot et al., 1998). These results indicate that
apoptosis by M22/23 is most probably SST-dependent, since this mutant is not totally de®cient but
rather attenuated in its ability to induce p53-responsive
apoptotic genes. This remaining SST activity could be
due to the second transactivation subdomain within
p53 N-terminal domain. This second subdomain, called
subdomain 2, has been localized to a region
encompassing residues 43 ± 73 by deletion analysis
(Chang et al., 1995) and functionally characterized in
yeast (Candau et al., 1997). Within subdomain 2, two
hydrophobic residues W53 and F54 play a critical role,
suggesting thus that it is structurally similar to both
VP16 and the ®rst p53 transactivation domains (Cress
and Treizenberg, 1991; Regier et al., 1993; Lin et al.,
1994; Candau et al., 1997). We therefore mutated both
residues W53 and F54 together in p53 WT and M22/
23, in order to generate the mutant Q53, S54 (M53/54)
or the quadruple mutant at positions 22, 23, 53 and 54
(p53QM), respectively. These mutants were evaluated
in the three above-mentioned functional assays: colony
formation assay, apoptosis and transactivation of
di€erent functional promoters. As shown in Figure 1,
p53QM has totally lost growth suppression activity and
this is correlated with a corresponding loss of apoptotic
Characterization of p53 transactivation domains
C Venot et al
activity (Figure 2, column 4). Moreover, this mutant is
completely de®cient for transactivation of all promoters tested (Figure 3, column 4). This mutant can be
thus de®ned as a p53 transactivation function-de®cient
mutant.
On the other hand, functional characterization of
M53/54 revealed that it displays a phenotype which is
intermediate to that displayed by p53 WT and M22/23.
It exhibits slightly stronger growth suppression and
apoptotic activities than M22/23 (Figures 1 and 2). In
transactivation assays, a similar quantitative di€erence
was observed when the MDM2, BAX and PIG3
promoters were used. Results obtained with the
WAF1 promoter are consistent with the above-
mentioned inactivity of M22/23 on this promoter,
since the activation due to M53/54 is nearly identical to
the activation due to p53 WT. This led us to conclude
that transactivation subdomain 1 but not subdomain 2
is critical for p53-dependent activation of the WAF1
promoter. A functional di€erence between both
subdomains is not only observed with WAF1
promoter activation. As shown by transient cotransfection with a MDM2 expressing vector (Figure
4a), only transactivation by subdomain 1 is sensitive to
MDM2-mediated inactivation, thus con®rming previous reports (Lin et al., 1994; Candau et al., 1997). As
expected and shown in Figure 4b, mutations in
subdomain 2 do not a€ect the recently described
Figure 3 Transcriptional activity of p53 mutants on various promoters. The transcriptional activity was measured at optimal
plasmid concentration as indicated in the titration curves of p53 WT inputs inserted in the upper right corner of each barchart.
0.3 mg of luciferase reporter gene under the control of di€erent p53 speci®c promoters and 0.1 mg of p53 expression vectors were
used in each transfection. For titration curves, p53 WT plasmid concentrations used were 1, 3, 10, 30, 100 and 300 ng. The
concentration used for the experiments (0.1 mg) is indicated by the arrow. Twenty-four hour after transfection, H1299 cells were
rinsed with phosphate-bu€ered saline (PBS), resuspended in cell lysis bu€er (Promega), and incubated for 20 min at room
temperature. Samples were centrifuged and luciferase activity of the cleared supernatant was determined in the presence of luciferin
(Promega) and ATP. Each experiment was repeated at least three times to calculate the means and standard deviations
2407
Characterization of p53 transactivation domains
C Venot et al
2408
Figure 4 (a) MDM2 can reduce SST by p53 WT and p53 M53/54 on PIG3 promoter but not SST by p53 M22/23. H1299 cells
were transiently transfected with 0.3 mg of PIG3 promoter (upstream from the luciferase gene), 0.1 mg of p53 expression vectors and
in presence or absence of the MDM2 expression plasmid (Dubs-Poterszman et al., 1995). Luciferase assay was performed as
described in Figure 3. (b) Expression levels of p53 mutant proteins in absence or presence of MDM2 protein. H1299 cells were
transiently co-transfected with 1 mg of p53 mutant expression vectors and with either MDM2 expression vector or empty vector.
Cells (56105) were collected 24 h after transfection and lysed for 15 min at 48C with 100 ml of RIPA bu€er and centrifuged at 15
000 r.p.m. for 15 min. The supernatants were collected, 1/3 loaded onto 10% SDS ± PAGE gels, and blotted onto PVDF membranes
(New England Nuclear, Boston, MA, USA). p53 WT and p53 mutants were detected with Pab240 anti-p53 monoclonal antibody
and with anti-mouse horseradish peroxidase-conjugated secondary antibody. Grb2 protein content was used as loading control by
using anti-grb2 monoclonal antibody (Transduction Laboratories) and anti-mouse horseradish peroxidase-conjugated secondary
antibody, after SDS ± PAGE migration of samples on 14% gels. Proteins were visualized by chemiluminescent detection (ECL)
(Amersham)
MDM2-mediated degradation of p53 (Haupt et al.,
1997; Kubbutat et al., 1997). Besides it appeared that
M53/54 and mainly D281G were more strongly
degraded by MDM2 than p53 WT.
Both amino-terminal and C-terminal regions of p53
have been shown to be involved in repression by p53
(reviewed in Gottlieb and Oren, 1996; Ko and Prives,
1996). This activity may be in¯uential in apoptosis
triggering (Murphy et al., 1996; Venot et al., 1998 and
references therein). Within the amino-terminal region,
both subdomain 1 and proline-rich domain of p53 have
been shown to be required for the repression function
of p53 (Roemer and Mueller-Lantzsch, 1996; Venot et
al., 1998). We have also investigated the potential role
of subdomain 2 in mediating repression by using M53/
54 and p53QM in transient co-transfection experiments
Characterization of p53 transactivation domains
C Venot et al
with a RSV LTR promoter construct. Results
presented in Figure 5 show that p53QM, like the
tumor-derived p53 mutant D281G, is unable to repress
RSV LTR promoter. On the other hand, M53/54, like
M22/23, retains partial repression activity on RSV
LTR. This led us to conclude that both transactivation
subdomains are required for full repression of
promoters by p53.
The 12 residue overlap between the end of
subdomain 2 and the start of the so-called prolinerich domain which has been localized between residues
62 and 91 (Walker and Levine, 1996) raises the
question of possible functional interdependence between the two regions. In fact, these domains seem to
be totally independent, since inactivating mutations of
either domain cause opposite phenotypes linked to
apoptotic activity. Whereas the proline-rich domaindeletion mutant is totally silent for transactivation of
PIG3 promoter, has lost repression activity, and is
de®cient for apoptosis triggering (Venot et al., 1998),
M53/54 retains signi®cant levels of all three activities
(this study). This, in turn, shows that the phenotype
displayed by the proline-rich domain-deletion mutant is
not attributable to the impairment of any transactivation function of p53 subdomain 2, e.g. on the PIG3
promoter.
At this stage, we can conclude that the region
encompassing the ®rst hundred residues of p53
contains three independent functional domains: transactivation subdomain 1 from residues 1 ± 40, transactivation subdomain 2 from 43 ± 73, and proline-rich
domain from 62 ± 91. The proline-rich domain mediates
some cellular activation of DNA binding on low
anity p53RE, e.g. PIG3 promoter and is also
required for repression of promoters (Venot et al.,
1998). Subdomains 1 and 2 cooperate with respect to
the transcriptional regulation activities of p53, i.e. SST
and repression of promoters, subdomain 1 being more
critical for transactivation of WAF1. Transactivation
and repression by p53 may operate through direct
contacts with components of the general transcription
machinery (reviewed in Bates and Vousden, 1996;
Gottlieb and Oren, 1996; Ko and Prives, 1996; Hansen
and Oren, 1997; Levine, 1997). It would be of utmost
interest to study the speci®city of each subdomain for
these components. Importantly, in vivo studies in yeast
have shown that the two subdomains require the yeast
adaptor complex ADA2/ADA3/GCN5 to varying
extents (Candau et al., 1997).
The observation that both M22/23 and M53/54 but
not p53QM retain apoptotic activity, does not favor
the hypothesis that either transactivation subdomain
would be a speci®c docking site for proteins involved in
transcriptional regulation-independent apoptosis by
p53. We thus infer that apoptosis by p53 in H1299
2409
Figure 5 Repression activity of p53 mutants. 0.3 mg of plasmid
containing luciferase reporter gene under the control of RSV LTR
promoter (Venot et al., 1998) and 0.1 mg of p53 expression vectors
were used in each transfection. Luciferase assay was performed as
described in Figure 3. The transcriptional activity was measured
at optimal plasmid concentration (0.1 mg) as indicated by the
arrow in the titration curves of p53 WT inputs inserted in the
upper left corner of this barchart. Concentrations used for the
titration curve were 5, 10, 50, 100 and 500 ng
cells is strictly dependent on its transcriptional
activities. This also implies that previous reports,
which suggested a transcriptional regulation-independent p53-mediated apoptotic pathway in other cell
lines, must be reinterpreted, keeping in mind the
relevancy of the p53 mutant(s) used.
Note added in proof
A recently published study (Zhu et al., 1998) focused on
p53 transactivation subdomains 1 and 2 corroborates the
conclusions of our work that subdomain 1 is speci®cally
required for activation of WAF1 and that apoptosis by p53
in H1299 cells is strictly dependent on its transcriptional
activities.
Acknowledgements
We gratefully acknowledge B Vogelstein and K Polyak for
providing WWP-luc and 0.7 kb PIG3 promoter constructs,
as well as W Gallagher for critical reading of the
manuscript.
References
Bargonetti J, Manfredi J, Chen X, Marshak D and Prives C.
(1993). Genes Dev., 7, 2565 ± 2574.
Bates S and Vousden K. (1996). Curr. Opin. Genet. Dev., 6,
12 ± 18.
Buckbinder L, Talbott R, Velasco-Miguel S, Takenaka I,
Faha B, Seizinger B and Kley N. (1995). Nature, 377,
646 ± 649.
Caelles C, Helmberg A and Karin M. (1994). Nature, 370,
220 ± 223.
Candau R, Scolnick D, Darpino P, Ying C, Halazonetis T
and Berger S. (1997). Oncogene, 15, 807 ± 816.
Chang J, Kim D, Lee S, Choi K and Sung Y. (1995). J. Biol.
Chem., 270, 25014 ± 25019.
Characterization of p53 transactivation domains
C Venot et al
2410
Chen X, Ko L and Prives C. (1996). Genes Dev., 10, 2438 ±
2451.
Cho Y, Gorina S, Je€rey P and Pavletich N. (1994). Science,
265, 346 ± 355.
Clore G, Omichinski J, Sakaguchi K, Zambrano N,
Sakamoto H, Appella E and Gronenborn A. (1994).
Science, 265, 386 ± 391.
Conseiller E, Debussche L, Landais D, Venot C, Maratrat
M, Sierra V, Tocque B and Bracco L. (1998). J. Clin.
Invest., 101, 123 ± 127.
Cress WD and Triezenberg SJ. (1991). Science, 251, 87 ± 90.
Dubs-Poterszman M, Tocque B and Wasylyk B. (1995).
Oncogene, 11, 2445 ± 2449.
El-Deiry W, Kern S, Pietenpol J, Kinzler K and Vogelstein
B. (1992). Nat. Genet., 1, 45 ± 49.
Friedlander P, Haupt Y, Prives C and Oren M. (1996). Mol.
Cell. Biol., 16, 4961 ± 4971.
Funk W, Pak D, Karas R, Wright W and Shay J. (1992).
Mol. Cell. Biol., 12, 2866 ± 2871.
Gottlieb T and Oren M. (1996). Biochim. Biophys. Acta.,
1287, 77 ± 102.
Hansen R and Oren M. (1997). Curr. Opin. Genet. Dev., 7,
46 ± 51.
Haupt Y, Rowan S, Shaulian E, Vousden K and Oren M.
(1995). Genes Dev., 9, 2170 ± 2183.
Haupt Y, Barak Y and Oren M. (1996). EMBO J., 15, 1596 ±
1606.
Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature,
387, 296 ± 299.
Hollstein M, Sidransky D, Vogelstein B and Harris C.
(1991). Science, 253, 49 ± 53.
Israeli D, Tessler E, Haupt Y, Elkeles A, Wilder S, Amson R,
Telerman A and Oren M. (1997). EMBO J., 16, 4384 ±
4392.
Je€rey P, Gorina S and Pavletich N. (1995). Science, 267,
1498 ± 1502.
Ko L and Prives C. (1996). Genes Dev., 10, 1054 ± 1072.
Kubbutat M, Jones S and Vousden K. (1997). Nature, 387,
299 ± 303.
Levine A. (1997). Cell, 88, 323 ± 331.
Lin J, Chen J, Elenbaas B and Levine A. (1994). Genes Dev.,
8, 1235 ± 1246.
Lu H and Levine A. (1995). Proc. Natl. Acad. Sci. USA, 92,
5154 ± 5158.
Ludwig R, Bates S and Vousden K. (1996). Mol. Cell. Biol.,
16, 4952 ± 4960.
Miyashita T and Reed J. (1995). Cell, 80, 293 ± 299.
Murphy M, Hinman A and Levine A. (1996). Genes Dev., 10,
2971 ± 2980.
Roemer K and Mueller-Lantzsch N. (1996). Oncogene, 12,
2069 ± 2079.
Owen-Schaub L, Zhang W, Cusack J, Angelo L, Santee S,
Fujiwara T, Roth J, Deisseroth A, Zhang W and Kruzel E
et al. (1996). Mol. Cell. Biol., 15, 3032 ± 3040.
Pavletich N, Chambers K and Pabo C. (1993). Genes Dev., 7,
2556 ± 2564.
Polyak K, Xia Y, Zweier J, Kinzler K and Vogelstein B.
(1997). Nature, 389, 300 ± 305.
Regier JL, Shen F and Triezenberg SJ. (1993). Proc. Natl.
Acad. Sci. USA, 88, 7605 ± 7609.
Rowan S, Ludwig R, Haupt Y, Bates S, Lu X, Oren M and
Vousden K. (1996). EMBO J., 15, 827 ± 838.
Sturzbecher H, Brain R, Addison C, Rudge K, Remm M,
Grimaldi M, Keenan E and Jenkins J. (1992). Oncogene, 7,
1513 ± 1523.
Thut C, Chen J, Klemm R and Tjian R. (1995). Science, 267,
100 ± 104.
Unger T, Nau M, Segal S and Minna J. (1992). EMBO J., 11,
1383 ± 1390.
Venot C, Maratrat M, Dureuil C, Conseiller E, Bracco L and
Debussche L. (1998). EMBO J., 17, 4668 ± 4679.
Wagner A, Kokontis J and Hay N. (1994). Genes Dev., 8,
2817 ± 2830.
Walker K and Levine A. (1996). Proc. Natl. Acad. Sci. USA,
93, 15335 ± 15340.
Wu G, Burns T, McDonald III E, Jiang W, Meng R, Krantz
I, Kao G, Gan D, Zhou J, Muschel R, Hamilton S,
Spinner N, Markowitz S, Wu G and El-Deiry W. (1997).
Nature Genetics, 17, 141 ± 143.
Wu G, Saftig P, Peters C and El-Deiry W. (1998). Oncogene,
16, 2177 ± 2185.
Yonish-Rouach E, Borde J, Gotteland M, Mishal Z, Viron A
and May E. (1994). Cell Death and Di€er., 1, 39 ± 47.
Zhu J, Zhou W, Jiang J and Chen X. (1998). J. Biol. Chem.,
21, 13030 ± 13036.