Download Simple identification of dominant p53 mutants by

Carcinogenesis vol.18 no.10 pp.2019–2021, 1997
SHORT COMMUNICATION
Simple identification of dominant p53 mutants by a yeast
functional assay
Alberto Inga1, Sara Cresta1, Paola Monti1, Anna Aprile,
Gina Scott2, Angelo Abbondandolo1,3, Richard Iggo4 and
Gilberto Fronza1,5
1CSTA-Mutagenesis
Laboratory, National Institute for Research on Cancer
(IST), Genova, Italy, 2School of Medicine, University of Leeds, Leeds, UK,
3Chair of Genetics, University of Genova, Italy and 4Swiss Institute for
Experimental Cancer Research (ISREC), Epalinges, Switzerland
5To
whom correspondence should be addressed
Analysis of families with germline p53 mutations shows
that the mutant p53 allele behaves as a dominant oncogene
at the genetic level, although it behaves as a recessive
oncogene at the cellular level, since tumours invariably
show mutation or loss of both wild-type alleles. At the
biochemical level it is possible that some clinically important mutant p53 proteins may be carcinogenic through a
dominant mechanism. We show that p53 mutants can be
readily classified according to their dominant potential
using a simple yeast functional assay. Wild-type p53 is
constitutively expressed from a TRP1 vector, p53 mutants
are expressed from an otherwise identical LEU2 vector
and net transcriptional activity is scored using an ADE2based reporter. Twenty seven p53 mutants were tested: 19
were recessive, i.e. gave white colonies, and eight showed
dominant activity, i.e. gave pink/red colonies. This simple
assay should facilitate studies on p53 dominance.
p53 alleles inactivated by missense mutations are seen in
~50% of all human tumours (1,2). Tumour-derived mutants
are unable to function as sequence-specific transcription factors,
generally because they contain mutations in the DNA binding
domain which reduce the affinity for DNA (3,4). The high
level of p53 protein commonly seen in tumours probably
reflects the presence of a persistent p53 activating signal and
loss of the mdm2-induced p53 degradation feedback loop
(5,6). In a normal cell p53 activation leads to cell cycle arrest
or apoptosis, but in a tumour containing mutant p53 the cells
can continue to divide because mutant p53 is unable to
transactivate its target genes. While this abortive activation/
impaired degradation model provides a satisfactory explanation
for the high level p53 expression seen in tumours (5,6), it
does not rule out the possibility that a selective growth
advantage may be conferred by mutated p53 proteins, either
because dominant inhibition of wild-type protein could facilitate progression from the heterozygous to the homozygous
mutant state or because the mutant protein has a novel growth
promoting activity (7). Although superficially similar, there is
an important biological distinction between these two models,
since the former implies only loss of p53 function as a
transcription factor, whereas the latter suggests that mutants
have gained new functions.
Tumour development is marked by a long period of coexistence of wild-type and mutant protein in the same cell,
either following an initial somatic mutagenic event or from
© Oxford University Press
birth in the case of germline mutations (2). Since p53 exists
as a stable tetramer in the cell (8) and since in most cases both
alleles are equally expressed and equally stable, heterozygous
mutant cells should contain a mixture of wild-type, mixed and
mutant tetramers in the ratio 1:14:1, provided the mutant
protein is able to form oligomers with the wild-type. The
extent to which p53 mutants really do abrogate wild-type p53
function in the heterozygous state is controversial and depends
on the particular assay employed (4,9–12), but there is certainly
the potential for strong dominant inhibition of wild-type p53
through tetramerization with mutant protein in heterozygous
cells.
It is possible to classify p53 mutations into those that
introduce adverse contacts and actively prevent DNA binding
by changing residues that physically interact with DNA (contact
mutants) and those that interfere with the orientation of the
contact residues by changing the overall structure of the DNA
binding domain (structural mutants) (8). Since physiological
p53 binding sites in genomic DNA frequently contain imperfect
matches to the p53 consensus, some subunits of a tetramer
frequently do not make a full set of bonds with DNA. Indeed,
a difference in affinity between the various natural p53 targets
has been described (13). Hence, a reduction in the number of
available protein contact sites is not a priori grounds for mixed
tetramers being inactive in the cell. It has been shown that
wild-type p53 can form stable tetramers with DNA contact
mutant proteins and that mixed tetramers can bind specific
DNA sequences. For example, mixed tetramers containing
248W mutant p53 bind DNA well (10), which illustrates how
a reduced affinity, due to loss of the four hydrogen bonds
between Arg248 and the minor groove, need not detract from
high affinity binding by remaining wild-type subunits in the
tetramer, provided the mutant residue (tryptophan) can be
accommodated in the structure.
Despite the fact that only one out of 16 tetramers in a
heterozygous cell are fully wild-type, some studies have shown
near normal p53 responses in heterozygous cells. This may
partly reflect the ability of p53 to accumulate to high levels
after activation, an effect which could negate even a substantial
reduction in the starting level of wild-type tetramers. Many of
the problems with mammalian studies derive from the difficulty
of achieving stable low level p53 expression in a reproducibly
defined genetic background and results are frequently clouded
by concerns over the amount of p53 produced, which often
far exceeds any relevant physiological level. For instance,
differences with transcription assays and transformation assays
probably reflect the fact that very small amounts of p53 give
saturating levels of transcriptional activation, whilst CMV
promoter-derived p53 overexpression is necessary for growth
inhibition. Dominance is an important issue in transformation
assays, where incoming p53 mutants must inactivate both
resident wild-type alleles. A further problem with mammalian
transfection studies is that it is difficult to achieve exactly
equal expression of both alleles. This can be overcome by use
2019
A.Inga et al.
Table I. Dominance analysis of p53 mutants in strain yIG397 of
Saccharomyces cerevisiae
Mutantb
Mutation
Codon
Amino acid
Dominancec Domaind
CCNU- and UV-induced mutations never found in human tumoursa
1
CCC→ATC
72
Pro→Ilee
No
Out
2
CCT→CTT
98
Pro→Leu
No
Preceeds
S1
3
GGG→GAG 117
Gly→Asp
No
L1
4
AAG→GAG 139
Lys→Glu
No
S29–S3
5
ACG→AGG 170
Thr→Arg
No
L2
6
GAG→AAG 224
Glu→Lys
No
S7–S8
7
GAC→GTC
228
Asp→Val
No
S7-S8
8
ACC→TCC
230
Thr→Ser
No
S8
9
GGT→GAT
262
Gly→Asp
No
S9–S10
10
CTG→CCG
323
Leu→Pro
No
Out
CCNU-induced and UV-induced mutations already found in human tumours
11
ACC→ATC
155
Thr→Ile
No
S3–S4
12
CGC→CCC
156
Arg→Pro
No
S4
13
CCC→CAC
177
Pro→His
Yes
H1
14
CGA→TGA
213
Arg→Gln
No
S6–S7
15
CCC→CTC
219
Pro→Leu
No
S7
16
TCC→TTC
241
Ser→Phe
Yes
L3
17
TGC→TAC
242
Cys→Tyr
Yes
L3
18
GGC→AGC
245
Gly→Ser
Yes
L3
19
GGC→GAC
245
Gly→Asp
Yes
L3
20
GAA→GAT
258
Glu→Lys
No
S9–S10
21
GAC→AAC
259
Asp→Asn
No
S9–S10
22
GGA→AGA 266
Gly→Arg
No
S10
23
CGT→TGT
273
Arg→Cys
Yes
S10
24
CCT→TCT
278
Pro→Ser
No
H2
25
GGG→GAG 279
Gly→Glu
Yes
H2
26
GAG→AAG 281
Asp→Asn
Yes
H2
27
GAA→AAA 286
Glu→Lys
No
H2
aAccording
to EMBL p53 database, May 13 1996, 5174 mutants entered
(18).
bNot
Fig. 1. Schematic representation of dominance analysis of p53 mutants in
strain yIG397. Double transformant clones (Leu1, Trp1), where only wildtype p53 homotetramers are present (A), express the ADE2 gene, giving rise
to white (Ade1) colonies on minimal plates lacking leucine and tryptophan
but containing sufficient adenine for adenine auxotrophs to grow and turn
red. Double transformants clones expressing wild-type p53 and mutant p53
proteins form white (Ade1) or pink/red (Ade–) colonies and were
interpreted as expressing recessive (B) or dominant (C) p53 mutant proteins
respectively. For simplicity, the situation in which mixed heterotetramers
containing three wild-type and one mutant protein (3:1) is represented.
of bicistronic vectors and studies using this approach have
shown that wild-type p53 is generally dominant over mutant
p53 (12).
To get over some of the problems with mammalian studies
we have modified the ADE2-based yeast p53 functional assay
(14) to allow straightforward characterization of the dominance
potential of p53 mutants. In this assay, human p53 cDNA is
cloned by homologous recombination in vivo (gap repair) into
a yeast expression vector. To determine whether the p53
clone can activate transcription, the yeast strain contains a
transcription reporter (Figure 1A) where the ADE2 gene is
regulated by a p53-responsive promoter. With this system,
wild-type p53 gives white colonies and mutant p53 gives red
colonies. To test dominance, two p53 cDNA must be expressed
together in the reporter strain. The existing expression vector
has a LEU2 marker; a new p53 expression vector, named
pTS76, was therefore constructed which is identical with the
first but has a TRP1 marker. To do this, the 4.9 kb PvuI
fragment (bp 2968–7964) containing the pADH1 wild-type
p53 cDNA from pLS76 (15) was ligated to the 2.9 kb PvuI
2020
underlined, CCNU-induced mutants (17); underlined, UV-induced
mutants (unpublished data).
cTransformants (Leu1Trp1) forming white (Ade1) or red/pink (Ade–)
colonies were classified as recessive or dominant mutants respectively.
dLocalization of the mutation according to the topological diagram of the
core domain of p53 (19).
eAmino acid change never found, but this is a site of a known
polymorphism.
fragment (bp 5973–1736) containing the CEN/ARS TRP1
region from pLS89 (16). pTS76 was used to express wildtype p53, while different p53 mutants were expressed from
pLS76-derived LEU2 vectors. The haploid strain yIG397
(MATa ade2-1 leu2-3,112 trp1-1 his3-11,15 can1-100, ura3-1
URA3 3xRGC::pCYC1::ADE2), containing the ADE2 reporter
gene under p53 control, was transfected with equimolar
amounts of both vectors by electroporation. Double transformants were selected on minimal plates lacking leucine and
tryptophan but containing sufficient adenine (5 mg/l) for
adenine auxotrophs to grow and turn red. Double transformant
clones (Leu1, Trp1) giving rise to white (Ade1) or pink/
red (Ade–) colonies were interpreted as expressing recessive
(Figure 1B) or dominant (Figure 1C) mutations respectively.
Generally, .100 independent transformants were analysed and
their phenotype was invariably homogeneous.
A panel of p53 mutations obtained through application of
the yeast p53 functional assay to the field of mutagenesis (17)
was used to validate the dominance assay. This included 18
CCNU-induced and nine UV-induced mutants (Table I). About
20% of the experimentally induced p53 mutants selected so
far in yeast (19 out of 82) (17; unpublished data) contain base
pair substitutions never found in human tumours or cell lines
Dominant p53 mutatants in yeast
(18; EMBL database, May 13 1996, 5174 mutations entered).
Such mutations, although completely inactive for transactivation (red colonies), are apparently not selected for in vivo
during the carcinogenic process. This could reflect mutagen
specificity, repair specificity or some biological activity of the
mutants. In particular, if dominance plays an important part
in vivo, one would not expect to see recessive mutations
in tumours.
Dominance analysis revealed that all the CCNU- and UVinduced mutations not found in human tumours were indeed
recessive (Table I). Double transformants of eight out of the
17 mutants already found in tumours showed instead pink/red
colonies and were considered dominant. However, colony size
and staining intensity suggested the presence of a minimal
wild-type p53 activity. All dominant mutants affected key
amino acids that are essential for stabilization of the DNA
binding surface of the p53 core domain or for direct interaction
between p53 and DNA (19).
Recently another selection scheme (URA3-based) for spontaneous dominant p53 mutations in yeast was described (20).
The analysis of that spectrum showed that p53 dominant
mutations correlate well with mutational hotspots in human
cancer. Only two out of the 41 dominant mutations described
(20) have not been observed in tumours, a number significantly
lower than that found by selecting mutants irrespective of their
dominant phenotype (19 out of 82, P 5 0.01, Fisher’s exact
test) (17; unpublished data). The results with the ADE2
selection method are generally consistent with those obtained
by the URA3-based selection. Dominant mutations were found
only in the group of mutants already found in tumours (Table
I; P 5 0.012, Fisher’s exact test).
Selection for dominance thus may play a role in determining
the mutation spectrum seen in human tumours. It should be
noted, however, that this is not evidence for ‘gain of function’
in the sense of acquisition of novel biochemical activities
favouring tumour cell growth, since it can be adequately
explained by mutant-induced loss of function of the wild-type
allele. The ability to classify mutants simply according to their
dominant potential at the biochemical/cellular level should
facilitate studies on Li–Fraumeni syndrome and tumour progression from the heterozygous to the homozygous state and
it may in future provide important information to the clinician
about the likelihood of response to chemotherapy and p53
gene therapy.
5. Haupt,Y., Maya,R., Kazaz,A. and Oren,M. (1997) Mdm2 promotes the
rapid degradation of p53. Nature, 387, 296–299.
6. Kubbutat,M.H.G., Jones,S.N. and Vousden,K.N. (1997) Regulation of p53
stability by mdm2. Nature 387, 299–303.
7. Hann,B.C. and Lane,D.P. (1995) The dominating effect of mutant p53.
Nature Genet., 9, 221–222.
8. Arrowsmith,C.H. and Morin,P. (1996) New insights into p53 function from
structural studies. Oncogene, 12, 1379–1385.
9. Waterman,J.L.F., Waterman,M.J.F. and Halazonetis,T.D. (1996) An
engineered four-stranded coil substitutes for the tetramerization domain
of wild-type p53 and alleviates transdominant inhibition by tumor-derived
p53 mutants. Cancer Res., 56, 158–163.
10. Bargonetti,J., Reynisdottir,I., Friedman,P.N. and Prives,C. (1992) Sitespecific binding of wild-type p53 to cellular DNA is inhibited by SV40 T
antigen and mutant p53. Genes Dev., 6, 1886–1898.
11. Kern,S.E., Pietenpol,J.A., Thiagalingam,S., Seymour,A., Kinzler,K.W. and
Vogelstien,B. (1992) Oncogenic forms of p53 inhibits p53-regulated gene
expression. Science, 265, 827–830.
12. Frebourg,T., Sadelain,M., Ng,Y.-S., Kassel,J. and Friend,S.H. (1994) Equal
transcription of wild-type and mutants p53 using bicistronic vectors results
in the wild-type phenotype. Cancer Res., 54, 878–881.
13. Wang,Y. and Prives,C. (1995) Increased and altered DNA binding of
human p53 by S and G2/M but not G1 cyclin-dpendent kinases. Nature,
376, 88–91.
14. Flaman,J.M. et al. (1995) A simple p53 functional assay for screening cell
lines, blood, and tumors. Proc. Natl Acad. Sci. USA, 92, 3963–3967.
15. Ishioka,C., Frebourg,T., Yan,Y-X., Vidal,M., Friend,S.H., Schmidt,S. and
Iggo,R. (1993) Screening patients for heterozygous p53 mutations using
a functional assay in yeast. Nature Genet., 5, 124–129.
16. Scharer,E. and Iggo,R. (1992) Mammalian p53 can function as a
transcription factor in yeast. Nucleic Acids Res., 7, 1539–1545.
17. Inga,A., Iannone,R., Monti,P., Molina,F., Bolognesi,M., Abbondandolo,A.,
Iggo,R. and Fronza,G. (1997) Determining mutational fingerprints at the
human p53 locus with a yeast functional assay: a new tool for molecular
epidemiology. Oncogene, 14, 1307–1313.
18. Hollstein,M., Shomer,B., Greenblatt,M., Soussi,T., Hovig,E., Montesano,R.
and Harris,C.C. (1996) Somatic point mutation in the p53 gene of human
tumours and cell lines updated compilation. Nucleic Acids Res., 24,
141–146.
19. Cho,Y., Gorina,S., Jeffrey,P.D. and Pavletich,N.P. (1994) Crystal structure
of a p53 tumor suppressor–DNA complex: understanding tumorigenic
mutations. Science, 265, 346–355.
20. Brachmann,R.K., Vidal,M. and Boeke,J.D. (1996) Dominant-negative p53
mutations selected in yeast hit cancer hot spots. Proc. Natl Acad. Sci.
USA, 93, 4091–4095.
Received on April 21, 1997; revised on June 18, 1997; accepted on June
18, 1997
Acknowledgements
We thank Drs P.Menichini, P.Campomenosi and L.Ottaggio for helpful
discussions. This work was partially supported by contract CHRX-CT94-0581
from the Commission of the European Communities and by the Italian
Association for Cancer Research (AIRC).
References
1. Greenblatt,M.S., Bennett,W.P., Hollstein,M. and Harris,C.C. (1994)
Mutations in the p53 tumor suppression gene: clues to cancer etiology
and molecular pathogenesis. Cancer Res., 54, 4855–4878.
2. Ko,L.J. and Prives,C (1996) p53: puzzles and paradigm. Genes Dev., 10,
1054–1072.
3. Ory,K., Legros,Y., Auguin,C. and Soussi,T. (1994) Analysis of the most
representative tumour-derived p53 mutants reveals that changes in protein
conformation are not correlated with loss of transactivation or inhibition
of cell proliferation. EMBO J., 13, 3496–3504.
4. Pietenpol,J.A., Tokino,T., Thiagalingam,S., El-Deiry,W.S., Kinzler,K.W.
and Vogelstein,B. (1994) Sequence-specific transcriptional activation is
essential for growth suppression by p53. Proc. Natl Acad. Sci. USA, 91,
1998–2002.
2021