¹Divion of Oncology, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada

²Divisions of Human Cancer Genetics and Population Sciences, Dana-Farber Cancer Institute, 44 Binney St, Boston, Massachussetts, MA 02115, USA

³OncorMed Inc., 205 Perry Parkway, Gaithersburg, Maryland MD 20877, USA

⁴Laboratory of Molecular Oncology, Institut Gustave-Roussy, Villejuif 75016, France

Correspondence to: David Malkin, Divion of Oncology, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada

The Li-Fraumeni Syndrome (LFS) is a rare, dominantly inherited syndrome that features high risk of cancers in childhood and early adulthood. Affected families tend to develop bone and soft tissue sarcomas, breast cancers, brain tumors, leukemias, and adrenocortical carcinomas. In some kindreds, the genetic abnormality associated with this cancer phenotype is a heterozygous germline mutation in the p53 tumor suppressor gene. Recently, we identified one patient who presented in early childhood with multiple primary cancers and who harbored three germline p53 alterations (R156H and R267Q on the maternal allele and R290H on the paternal allele). To classify the biologic effects of these alterations, functional properties of each of the p53 mutants were examined using in vitro assays of cellular growth suppression and transcriptional activation. Each amino acid substitution conferred partial or complete loss of wild-type p53 function, but the child completed normal embryonic development. This observation has not been previously reported in a human, but is consistent with observations of normal embryogenesis in p53-deficient mice.

Li-Fraumeni Syndrome; p53; germline mutations; functional assays

Introduction

The Li-Fraumeni Syndrome (LFS) is an autosomal dominant familial cancer disorder characterized by the diagnosis of bone or soft tissue sarcoma at an early age in an individual from a family with multiple members affected by a wide array of other cancers (Li and Fraumeni 1969a,b; Williams and Strong, 1985; Li, 1988). The component tumors of LFS are bone and soft tissue sarcomas, breast cancers, brain tumors, leukemias, and adrenocortical carcinoma (Kleihues et al., 1997). An LFS family is classically defined by the inclusion of one individual, the proband, diagnosed with sarcoma before 45 years of age, who has one first degree relative with cancer before 45 years of age and another first or second degree relative in the same parental lineage with any cancer diagnosed before 45, or with sarcoma at any age (Li, 1988). The estimated disease penetrance is 50% by 30 years and 90% by 60 years of age (Williams and Strong, 1985).

Many families have been studied that resemble LFS but do not fit the original definition. These LFS-like (LFS-L) families either fail to meet one particular criterion of the classic definition or present with other less frequently observed malignancies, such as germ cell tumors, melanoma, prostatic or pancreatic cancer (Strong et al., 1987; Li, 1988; Hartley et al., 1989; Garber et al., 1990). These tumors are now considered to be associated with LFS and others are being evaluated. Efforts have been made to define LFS-L, but this term has been used by some to describe families with individuals who are first or second degree relatives affected at any age (Eeles, 1995) and by others to describe LFS with an increased age restriction of 60 years (Birch et al., 1994). Further molecular and epidemiologic studies are needed to clarify these distinctions.

Inactivion of the p53 tumor suppressor gene has been found in a wide variety of sporadic tumors, including rhabdomyosarcoma, osteosarcoma (Mulligan et al., 1990), leukemia (Ahuja et al., 1989), carcinomas of the lung (Takahashi et al., 1989) and breast (Osborne et al., 1991), brain tumors (Mashiyama et al., 1991), and colorectal cancer (Baker et al., 1989). Germline p53 alterations have been observed in the majority of LFS families and in some LFS-L families (Malkin et al., 1990; reviewed in Malkin, 1998). Although deletions, insertions and splice mutations have been observed, the majority of the p53 mutations reported in LFS and sporadic tumors are missense, giving rise to an altered protein (Hollstein et al., 1991; reviewed in Malkin, 1998). Most germline mutations are found in four highly conserved regions of the protein: residues 117 - 142, 171 - 181, 234 - 258 and 270 - 286. The particularly high frequency of mutations reported at codons 175, 245, 248, 273 and 282 has led to their reference as `hot spots' and are similar to those found in sporadic tumors.

Because of its apparent involvement in the pathogenesis of numerous sporadic and hereditary cancers, the functions of the p53 protein have been widely studied. p53 plays a central role in the cellular response to DNA damage (Levine, 1997). Wild-type p53 has been shown to be an important player in the G₁/S cell cycle checkpoint (Kastan et al., 1991; Kuerbitz et al., 1992), which involves induction of p21^WAF1/CIP1 (El-Deiry et al., 1993; Xiong et al., 1993), an inhibitor of cyclin-dependent kinases (Cdks) (Harper et al., 1993; Dulic et al., 1994). A link between the p53-dependent cell cycle checkpoint and DNA repair was established when it was demonstrated that ionizing radiation induction of GADD45, which binds PCNA to induce excision repair of damaged DNA (Smith et al., 1994), is dependent on wild-type p53 (Kastan et al., 1992). Recent evidence indicates that p53 may be involved in blocking cells at G_2/M (Paules et al., 1995; Cross et al., 1995; Wang and Prives, 1995; Fukasawa et al., 1996). In addition to its role in cell cycle regulation, p53 is also able to induce apoptosis (Yonish-Rouach et al., 1991; Shaw et al., 1992; Lowe et al., 1993) by regulating the expression of other target genes, such as bax and bcl-2 (Miyashita et al., 1994a; Miyashita and Reed, 1995).

Some p53 mutants have simply lost their growth suppressive properties, but others are thought to inhibit the functions of the wild-type protein in a `dominant negative' fashion through oligomerization (Farmer et al., 1992; Kern et al., 1992). Furthermore, some mutants may have acquired a `gain of function', demonstrated by the conferring of increased tumorigenicity upon introduction of mutant p53 constructs into p53-deficient cells (Dittmer et al., 1993). Therefore, both dominant negative and gain of function mutations would be more carcinogenic than the loss of one p53 allele. This is well illustrated in the analysis of mice containing a mutant p53 transgene (Harvey et al., 1995).

The relationship between the LFS phenotype and genotype has yet to be fully elucidated. The reported proportion of patients with `classic' LFS (Li, 1988) harboring germline p53 mutations has varied from 60 - 80% (Birch et al., 1994; Frebourg et al., 1995; Varley et al., 1997). Furthermore, the degree of tumorigenicity associated with different p53 mutants is poorly understood. As part of initial steps to better understand the mechanisms of action of p53 mutants in LFS and to permit predictive testing, we regularly evaluate DNA obtained from blood samples of LFS patients for the presence and type of p53 mutation. We report here an intriguing p53 genotype in a classical LFS family.

Results

Identification of germline p53 mutations

DNA sequence analysis of the p53 gene in peripheral blood of an LFS patient revealed the presence of R156H, R267Q and R290H mutations. Plasmids containing exon 8 PCR products, amplified from DNA from the child (the proband), obtained from several bacterial clones harbored either an R267Q mutation or an R290H mutation; no clones contained both mutations and no clones contained only wild-type DNA sequence (Figure 1). Blood samples from his parents showed co-segregation of the R156H and R267Q mutant alleles in the mother, whereas the father harbored the R290H mutant allele. The extended pedigree of this LFS family is shown in Figure 2a. The proband is a child who developed a rhabdomyosarcoma in the right neck at age 2, a brain tumor at age 10, and has died at age 12. His mother had bilateral breast cancer at ages 35 and 43, and his father was unaffected. No family history is available for the father. Figure 2b showed the pedigree of an LFS-L family, in which a R290H alteration has also been found in the germline of the proband, who developed a brain tumor at age 9. Her parents were unaffected but the paternal grandfather died of brain tumor at age 40 and her paternal first cousin died of a rhabdomyosarcoma at age 4. No samples were available for analysis from the relatives, but the pattern of cancers in the three affected family members suggests that the R290H mutation might be incompletely penetrant.

Effect of p53 variants on activation of transcription from RGC promoter in yeast

To examine the extent of loss of wild-type function, each of the three missense alterations were further characterized using in vitro functional assays. cDNA vectors were constructed by site-directed mutagenesis PCR to express R156H, R267Q and R290H mutants, as well as one that expresses both the R156H and R267Q mutations (R156H/R267Q - to represent the corresponding allele of the proband). The effect of these alterations on protein function were studied using in vitro assays that assess the ability of p53 to suppress cellular growth and activate transcription. The mutants were first assessed for their ability to bind DNA and activate transcription of a reporter gene, ADE2. Binding of the wild-type p53 to an RGC binding site upstream of its promoter drives the expression of this gene and corrects the red phenotype of yeast defective in their ability to produce adenine (colonies will become white) (Flaman et al., 1995). As shown in Figure 3, R267Q and R156H/R267Q transformants retained the red mutant phenotype, indicating that these substitutions represent loss of function mutations for the ability of p53 to activate transcription from the RGC sequence. In contrast, the R290H and R156H transformants were white, indicating that these amino acid substitutions did not affect p53 activity. These results were confirmed in a second laboratory (JF) that performed this assay using mRNA from the parents.

Expression of p53 and p21^WAF1/CIP1 in transfected Saos-2 cells

Since the DNA-binding properties of the p53 variants may be influenced by the nature of the binding site, each variant was also tested for the ability to induce the expression of p21^Waf1/Cip1, an important downstream target of p53 in inducing G₁/S arrest (El-Deiry et al., 1993; Dulic et al., 1994). Protein expression in p53-deficient Saos-2 osteosarcoma cells (American Type Cell Culture) was analysed 48 h following transfection with the various p53 cDNA constructs and each transfection was performed in triplicate. Western blot analysis (Figure 4) demonstrates that p53 protein is expressed by each of the p53 cDNA vectors in Saos-2 cells, but not in untransfected cells or in those transfected with the cDNA3 vector alone (with no insert). This confirms that Saos-2 cells do not express p53 and therefore any that is detected is exogenously expressed from the transfected construct. The relative amount of p53 expressed by each construct was assessed by correcting for loading (% actin) and transfection efficiency (% -gal activity). To facilitate comparison between experiments, values are expressed as a percentage of the wild-type. For all analyses, a two-tailed t-test was used to determine whether each of the p53 variants is significantly different from the wild-type. As shown in Table 1, R267Q and R156H/R267Q variants result in levels of p53 protein that are similar to wild-type, while R156H and R290H variants give rise to significantly lower levels of protein (P<0.005 and P<0.001, respectively).

As shown in Figure 4, p21^WAF1/CIP1 is only induced in Saos-2 cells that have been transfected with p53 cDNA. p21^WAF1/CIP1 levels were analysed in two ways: first, to assess the efficiency of each mutant p53 protein to activate transcription, relative p21^WAF1/CIP1 levels were measured by optical densitometry and calculated as a fraction of p53 protein levels; second, to compare the actual amounts of p21^WAF1/CIP1 that result from each transfection, p21^WAF1/CIP1 levels were corrected for loading and transfection efficiency. The latter analysis takes into account the ability to activate transcription as well as differences in the amount of p53 present due to variations in protein stability. For both analyses, values were expressed as a percentage of the p21^WAF1/CIP1 found in wild-type transfectants. All p53 variants have a significantly reduced ability to induce p21^WAF1/CIP1 (Table 1). Differences between the R156H and R290H variants and the wild-type were further increased when p21^WAF1/CIP1 values were not normalized for p53 levels.

Effect of p53 variants on growth of Saos-2 cells

In activating the transcription of several target genes, the ultimate effect of p53 is the suppression of tumor cell growth. Therefore, the p53 variants were assessed for the ability to inhibit the proliferation of Saos-2 cells. Growth suppression is manifested by a reduction in colony formation by cells transfected with p53. Expression vectors were stably transfected into Saos-2 cells by lipofection and selection in G418 for 3 weeks. Transfection of Saos-2 cells with the wild-type p53 cDNA vector resulted in a reduced plating efficiency as compared to cells transfected with the cDNA₃ vector alone (Figure 5 and Table 2), indicating that growth suppression is due to the expression of p53. The p53 mutants displayed a variation in their ability to suppress the growth of Saos-2 cells. To account for differences in conditions for each experiment, the results were expressed as the number of colonies as a percentage of the control (cDNA₃ vector alone) and the average of this value was calculated for three different experiments (Figure 6 and Table 3). Stable transfection of Saos-2 cells with the wild-type p53 construct resulted in 15.38±3.16% of the number of colonies formed by cDNA₃ transfectants. Stable transfection of the p53 V143A mutant and the R156H/R267Q mutant vector resulted in substantially higher numbers of colonies (103.65±8.86% and 97.12±11.78%, respectively), indicating that these mutants have lost the ability to suppress growth. The other p53 variants demonstrated a stronger ability to suppress the growth of Saos-2 cells: R156H was similar to the wild-type p53 (17.96±5.39%), R290H led to slightly lower growth suppression than the wild-type p53 (23.87±3.04%), and R267Q displayed a greater drop in the ability to suppress growth (37.87±3.82%).

A two-tailed t-test was used to determine whether each of the p53 variants were significantly different from the wild-type in their ability to suppress the growth of Saos-2 cells (P values are listed in Table 2). In summary, only the V143A, R267Q and R156H/R267Q mutants displayed a significantly reduced growth suppressive ability (P<0.001, P<0.02 and P<0.005, respectively) and the R290H variant demonstrated a reduction that was only significant at p<0.2. The R156H variant does not demonstrate results that are significantly different from the wild-type (P>0.5).

Discussion

Germline p53 mutations have been observed in the majority of families with Li-Fraumeni Syndrome. This report describes an affected child in an LFS family with germline alterations in both p53 alleles (R156H and R267Q on the maternal allele and R290H on the paternal allele). All three alterations are located in the highly conserved DNA binding domain, although residue 290 is found at the extreme 3'-end of this region. The R290H mutation has been observed previously on three occasions: (1) in a colorectal carcinoma (Lang et al., 1993); (2) in acute myelogenous leukemia cells (Hu et al., 1992), in which the protein was shown to be recognized by the PAb 240 antibody that detects mutant p53 (Gannon et al., 1990; Milner and Medcalfe, 1991); and (3) in the germline of our previously unreported LFS family (Figure 2b). The R267Q mutation has been reported in a breast cancer (Coles et al., 1992) and in the germline of a non-LFS cancer family in which a variety of cancers (mostly breast cancer) were observed over three generations (Prosser et al., 1992). The R156H alteration has not been reported previously.

Transient transfection of Saos-2 cells with mutant p53 constructs resulted in significantly lower levels of activated p21^WAF1/CIP1 than in cells transfected with the wild-type cDNA. In contrast, the R156H and R290H variants demonstrate the normal functional activation of transcription from the RGC sequence in yeast. Another study has also demonstrated that some p53 mutants retain wild-type activity for transcriptional activation using a similar yeast assay as described here, yet have lost the ability to undergo G₁ arrest and to induce p21^WAF1/CIP1 or GADD45 following ionizing irradiation (O'Connor et al., 1997). However, a comparison between the two transcriptional assays is difficult because the yeast assay is not quantitative (results are all or nothing), whereas the Western analysis in Saos-2 cells can detect subtle differences in various mutant forms of p53. It is possible that lower levels of p21^WAF1/CIP1 induction observed for the R156H and R290H variants would be sufficient to cause the phenotypic color change in the yeast. However, if this were true, then a color change would also be expected in yeast transformed by the R267Q mutant (16% of wild-type).

The colony-forming assay, which measures the ability of p53 to suppress the growth of Saos-2 cells, reveals that the p53 variants studied here display a range of maintenance of functional integrity. Such a range for different p53 mutants has been observed by others (Diller et al., 1990; Frebourg et al., 1992; Crook et al., 1994; Ishioka et al., 1997). Interestingly, transfection of the construct containing both the R156H and R267Q mutants, as on the maternal allele, resulted in a much higher number of colonies than with either mutant alone or what would be expected if the effects of each were additive. These mutants appear to enhance synergistically the mutant properties of the protein product, possibly through a conformational effect. One corollary of this conclusion has been previously observed in studies in which a second p53 mutation is able to work in cis to suppress the mutant phenotype (Brachman et al., 1997).

Although the colony-forming assay is quantitative, the results differ slightly from those observed for the up-regulation of p21^WAF1/CIP1. The R156H and R290H variants demonstrated no significant reduction in the ability of wild-type p53 to suppress Saos-2 cell growth. Others have also observed discordance in retention of function as measured by the two assays. One study demonstrated that p53 cDNA constructs lacking domains necessary for growth suppression of H1299 exhibit wild-type transcriptional activation of a reporter gene using a CAT assay (Walker and Levine, 1996). Another study demonstrated that a R273H/P309S p53 variant lost the ability to suppress the growth of several types of cell lines in spite of its wild-type ability to activate transcription from the RGC sequence, while a R273L variant maintained the ability to suppress growth in spite of having lost the ability to activate transcription (Kawamura et al., 1996). Furthermore, oligomerization domain p53 mutants L344P and K351Ter, which are defective for transcriptional activation in Saos-2 cells, displayed the ability to inhibit colony formation in these cells (Ishioka et al., 1995, 1997).

Perhaps the ability of p53 to activate transcription upon binding to one specific responsive element (RGC or p21^WAF1/CIP1 promoter) does not represent potential interactions with other p53 binding elements that may have important functional consequences for growth suppression, such as those upstream of target genes whose products are involved in apoptosis. Alternatively, it is possible that p53-dependent growth suppression of Saos-2 cells involves a mechanism that is independent of transcriptional activation, such as the repression of transcription.

It is difficult to compare the biological importance of results from the different assays used in this study. Although the RGC sequence exhibits strong binding to p53 (Kern et al., 1992; Kim et al., 1997), the relevance of this interaction in vivo is not known. In contrast, the upregulation of p21^WAF1/CIP1 by binding of p53 to sequences upstream of its promoter is critical for its induction of G₁/S arrest (El-Deiry et al., 1993; Dulic et al., 1994). However, it is difficult to predict the biological consequence of the partial activity demonstrated by the R156H and R290H variants. The assays that measure activation of transcription from the RGC sequence and p21^WAF1/CIP1 promoter examine specific mechanisms of p53 action, whereas the growth suppression assay may represent other mechanisms and analyses a cellular phenotype. Thus, each assay has important, but different biological relevance.

This study demonstrates that the R156H/R267Q p53 allele inherited by the proband encodes a protein that is defective for transcriptional activation as well as growth suppression of Saos-2 cells. The other allele (R290H) gives rise to a protein that has maintained the ability to induce growth suppression of Saos-2 and activate transcription from the RGC sequence, but this ability is reduced for the p21^WAF1/CIP1 promoter. Although the R290H mutant may have less severe functional consequences than the R156H/R267Q allele, the relative severity of the consequences of these different mutations in this family are difficult to assess. It is possible that the aggressive cancer phenotype of the proband may be attributed to this unusual p53 genotype, in which each allele has lost partial or complete function. Also, the observation that the R290H alteration is less detrimental to p53 function may explain the pattern of incomplete penetrance found in the LFS-like cancer family (Figure 2b).

The fetal development of the proband is consistent with similar observations in p53-deficient mice, which are able to complete gestation and are usually born without any observable gross defects, but then rapidly develop a variety of tumors, including sarcomas and others commonly seen in LFS (Donehower et al., 1992; Purdie et al., 1994; Sah et al., 1995). Therefore, although p53 plays a central role in cell cycle arrest and apoptosis in the cellular response to DNA damage, perhaps these functions are not essential in human embryogenesis. Alternatively, as with the Rb family, there may exist other proteins sharing properties similar to p53, such as p73, which has been shown to activate the transcription from p53 binding sequences and induce apoptosis in transiently transfected Saos-2 cells (Jost et al., 1997; Kaghad et al., 1997). p53CP (p53 competing protein), another protein that has been recently isolated and characterized, demonstrates homology to p53 and the ability to bind specifically to the consensus p53 binding sites found in several p53 downstream target genes (Bian and Sun, 1997). Perhaps there exist other homologous proteins or even pathways that have yet to be discovered that are redundant for the important functions of p53.

Materials and methods

Isolation of DNA from peripheral blood

Upon obtaining informed consent all eligible adults and from the parents of all eligible children, genomic DNA was extracted at the DFCI from peripheral blood leukocytes, using a Qiagen blood DNA isolation kit.

PCR - SSCP and sequencing

Fluorescent automated sequencing (ABI) was performed at DFCI. As described previously (Peng et al., 1993), PCR - SSCP was performed on the isolated DNA spanning p53 exons 2 - 11 at the HSC, and samples demonstrating reproducible band shifts were sequenced by the Sanger dideoxy method to characterize the precise base pair alterations.

Mutagenesis and construction of p53 expression vectors

Mutations were inserted into a p53 wild-type cDNA sequence using a PCR-based method of site-directed mutagenesis. Two PCR reactions were carried out, amplifying sequences both upstream and downstream of each mutation. For one reaction, a primer complementary to the antisense strand containing the mutation of interest [in bold] (5'-CACCCACGTCCGCGCCATGGCC-3' for R156H, 5'-GGGACAGAACAGCTTTGAGG-3' for R267Q, and 5'-GCACAGAGGAAGAGAATCTCCACAAG-3' for R290H) were used with a primer complementary to the sense strand downstream of the mutation (5'-AATGTCAGTCTGAGTCAGGC-3' for all). For the other reaction, a primer complementary to the sense strand containing the mutation (5'-GGACGTGGGTGCCGGGCGGGGGTG-3' for R156H, 5'-CTGTTCTGTCCCAGTAGA-3' for R267Q), and a primer downstream of the mutation (5'-CTTGAGTTCCAAGGCCTCATTCAGC-3' for R290H) were used with a primer complementary to the antisense strand upstream of the mutation (5'-AAGTCTAGAGCCACCGTCCA-3' for all). Each 100 l PCR reaction consisted of 10 ng of p53 cDNA template, 500 ng of each primer, 200 M dNTP's, 1.5 mM MgCl₂, and 2.5 units of Taq polymerase (AmpliTaq; Cetus). The reaction conditions for the Perkin-Elmer 480 thermocycler (Norwalk, CT, USA) were: 94°C for 45 s, 60°C for 5 min and 72°C for 2 min for 5 cycles, followed by: 94°C for 45 s, 60°C for 2 min, 72°C for 2 min for 20 cycles, and then 72°C for 10 min. Products from both reactions were purified from agarose using Qiaex extraction (Qiagen) and 10 l of each were then used together as the template in a second PCR reaction with the flanking oligonucleotides primers used in first PCR step (5'-AATGTCAGTCTGAGTCAGGC-3' and 5'-AAGTCTAGAGCCACCGTCCA-3'). Other conditions remain the same as above except all 30 cycles followed the profile: 94°C for 45 s, 60°C for 2 min, 72°C for 2 min, and then 72°C for 10 min. The final PCR product was purified from agarose and each digested with different restriction enzymes: PvuII+Bsu36I for R156H, Bsu36I+StuI for R267Q, and StuI+EarI for R290H. Each fragment was subcloned into an appropriately digested WT p53cDNA₃ expression vector. The WT p53cDNA₃ was created by inserting WT p53 cDNA (a gift of Dr Stephen Friend) into the BamHI site of the pcDNA₃ pcDNA₃ plasmid (In Vitrogen). 1 l calf intestinal phosphatase (CIP) was also included in the pcDNA₃ digest to prevent self-annealing. V143A cDNA was provided by Dr Sam Benchimol (Ontario Cancer Institute) and was inserted into cDNA3 in the same way. The R156H/R267Q double mutant was created by subcloning the R267Q mutant cDNA into a R156H p53cDNA₃ backbone that had been digested with Bsu36I+StuI. The pcDNA₃ plasmid carries a CMV promoter, driving the expression of the upstream p53 cDNA, as well as Ampicillin and Neomycin resistance genes. The presence of mutations in all constructs were verified by sequence analysis.

Activation of transcription in yeast

All p53 cDNA constructs were analysed for their ability to bind to the ribosomal gene cluster (RGC) sequence and activate transcription of the ADE2 reporter gene in yeast. This was performed by OncorMed as described previously (Flaman et al., 1995).

Cell culture and transfection

Saos-2 osteosarcoma cells were purchased from American Type Cell Culture and grown as monolayers in -MEM with 10% fetal calf serum in a humidified incubator at 37°C and 5% CO₂. Protein expression was analysed following transient transfection of Saos-2. 5´10⁵ cells were seeded onto 100 mm plates, incubated overnight at 37°C, and washed with serum-free -MEM without antibiotics. For each plate, 18.75 l of Lipofectin Reagent was diluted in 375 l of Opti-Mem I Medium and 7.5 g each of p53 cDNA₃ and cDNA3/LacZ (In Vitrogen) were diluted together in 375 l of Opti-Mem I Medium and incubated at room temperature for 45 min. These were then mixed together and incubated further for 15 min at room temperature. Three ml of serum-free -MEM was added to the above mixture, which was then added to washed cells, incubated overnight at 37°C, and then replaced with 10 ml of medium containing 10% serum. Forty-eight hours later, cells were washed twice with PBS, and incubated in 100 l modified RIPA lysis buffer (50 mM Tric-Cl pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.5% sodium deoxycholate, 1% Triton X-100, 0.1% SDS, 1 mM PMSF, 2 g/ml each leupeptin, pepstatin, aprotinin) on ice for 10 min. Cells were harvested by scraping them into a 1.5 ml microfuge tube and then sonicated with an XL-2020 ultrasonic processor (Misonix) using 20% power for 1 min. Transfection efficiency was determined by measuring -gal activity of each protein sample. A 20 l aliquot of each cell lysate was removed prior to sonication, 10 g of total protein was diluted to 20 l in modified RIPA buffer, and transferred to a microtiter plate. The assay buffer consisted of 1 ml of a 2 mg/ml solution of the substrate ONPG (O-nitrophenyl -D-galactopyranoside), in freshly prepared 0.1 M NaPO₄, added to 5 ml of -gal buffer (60 mM Na₂HPO₄, 40 mM NaH₂PO₄, 10 mM KCl, 1 mM MgCl₂, 50 mM -mercaptoethanol). 208 l of the assay buffer was mixed into each of the wells containing cell lysates and incubated at 37°C until a yellow color developed (~1 h). 80 l of 1 M Na₂CO₃ was then mixed into each well and the absorbance was measured at 410 nm.

Western blotting

Forty g total protein from each lysate separated on a 12.5% acrylamide gel, electroblotted onto nitrocellulose, and blocked in 5% Blotto for 1 h. The membrane was incubated overnight at 4°C with the following mouse anti-human monoclonal antibodies: two p53 antibodies pAb 1801 and pAb DO-1 (1 : 200 for each, Oncogene Science, Ab-2 and Ab-6, respectively), a p21^Waf1/Cip1 antibody (1 : 100, Ab-1, Oncogene Science) and an actin antibody (1 : 400, Sigma). After washing and incubation with peroxidase-labeled goat anti-mouse IgG (H+L) secondary antibody (1 : 5000, KPL) for 1 h at room temperature, protein bands were detected using ECL (Enhanced Chemiluminescence, Amersham) according to the manufacturer's instructions. The signal for each protein band was quantitated using optical densitometry.

Stable transfection of Saos-2 and colony-forming assay

Transfections were performed on the same day using Saos-2 cells plated from the same stock. Two ´ 10⁵ cells were seeded onto 60 mm plates, incubated overnight at 37°C, and washed with serum-free -MEM without antibiotics. Transfections were carried out using Lipofectin (Gibco - BRL). 10 l of Lipofectin Reagent was diluted in 100 l of Opti-Mem I medium and incubated at RT for 45 min. This was then mixed with 100 l of Opti-Mem I medium containing 4 g of p53 mutant construct and incubated further for 15 min at RT. 1.8 ml of serum-free -MEM was added to above mixture, which was then incubated overnight at 37°C with washed cells and then replaced with 4 ml of medium containing 10% serum. Forty-eight hours later, cells were split 1 : 3 into selection medium (-MEM containing 10% serum and 500 g/ml G418). After 3 weeks, colonies were stained with methylene blue and counted. Each set of transfections was done in triplicate on separate days yielding reproducible relative colony numbers.

Acknowledgements

This work was funded in part by a grant from the National Cancer Institute of Canada (S Quesnel, C Portwine, D Malkin) with funds from the Terry Fox Foundation; and the Starr Foundation, Liberty Mutual Group and the Boston Foundation (S Verselis, J Garber, P FP Li). D Malkin is a Research Scientist of the NCIC supported with funds provided by the Canadian Cancer Society. Technical assistance with the yeast assay was kindly provided by Christopher Alvares and Tracy Staton. Thanks for critical comments from Drs Alan Bernstein, Yaacov Ben-David and Andre Schuh.

Ahuja H, Bar-Eli M, Advani SH, Benchimol S and Cline MJ. (1989). Proc. Natl. Acad. Sci. USA 86, 6783-6787. MEDLINE

Baker SJ, Fearon ER, Nigro JM, Hamilton SR, Preisinger AC, Jessup JM, vanTuinen P, Ledbetter DH, Barker DF, Nakamura Y, White R and Vogelstein B. (1989). Science 244, 217-221. MEDLINE

Bian J and Sun Y. (1997). Proc. Natl. Acad. Sci. USA 94, 14753-14758. MEDLINE

Birch JM, Hartley AL, Tricker KJ, Prosser J, Condie A, Kelsey AM, Harris M, Morris Jones PH, Binchy A, Crowther D, Craft AW, Eden OB, Evans GR, Thompson E, Mann JR, Martin J, Mitchell ELD and Santibanez-Koref MF. (1994). Cancer Res. 54, 1298-1304. MEDLINE

Brachmann RK, Yu K, Eby Y, Pavletich NP and Boeke JD. (1997). EMBO J. 17, 1847-1859.

Coles C, Condie A, Chetty U, Steel CM, Evans HJ and Prosser J. (1992). Cancer Res. 52, 5291-5298. MEDLINE

Crook T, Marston NJ, Sara EA and Vousden KH. (1994). Cell 79, 817-827. MEDLINE

Cross SM, Sanchez CA, Morgan CA, Schimke MK, Ramel S, Idzerda RL, Raskind WH and Reid BJ. (1995). Science 267, 1353-1356. MEDLINE

El-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein B. (1993). Cell 75, 817-825. MEDLINE

Diller L, Kassel J, Nelson CE, Gryka MA, Litwak G, Gebhardt M, Bressac B, Ozturk M, Baker SJ, Vogelstein B and Friend SH. (1990). Mol. Cell Biol. 10, 5772-5781. MEDLINE

Dittmer D, Pati S, Zambetti G, Chu S, Teresky AK, Moore M, Finlay C and Levine AJ. (1993). Nature Genet. 4, 42-46. MEDLINE

Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery CA, Butel JS and Bradley A. (1992). Nature 356, 215-221. MEDLINE

Dulic V, Kaufman WK, Wilson SJ, Tlsty TD, Lees E, Harper JW, Elledge SJ and Reed SI. (1994). Cell 76, 1013-1023. MEDLINE

Eeles RA. (1995). Cancer Surveys 25: Genetics and Cancer: A Second Look. Ponder BAJ, Cavenee WK, Solomon E. (eds). Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 101.

Farmer G, Bargonetti J, Zhu H, Friedman P, Prywes R and Prives C. (1992). Nature 358, 83-86. MEDLINE

Flaman JM, Frebourg T, Moreau V, Charbonnier F, Martin C, Chappuis P, Sappino A, Limacher J-M, Bron L, Benhatter J, Tada M, Van Meir EG, Estreicher A and Iggo RD. (1995). Proc. Natl. Acad. Sci. USA 92, 3963-3967. MEDLINE

Frebourg T, Kassel J, Lam KT, Gryka MA, Barbier N, Andersen TI, Borresen AL and Friend SH. (1992). Proc. Natl. Acad. Sci. USA 89, 6413-6417. MEDLINE

Freboug T, Barbier N, Yan YX, Garber JE, Dreyfus M, Fraumeni J, Li FP and Friend SH. (1995). Am. J. Hum. Genet. 56, 608-615. MEDLINE

Fukasawa K, Choi T, Kuriyama R, Rulong S and Vande Woude GF. (1996). Science 271, 1744-1747. MEDLINE

Gannon JV, Greaves R, Iggo R and Lane DP. (1990). EMBO J. 9, 1595-1602. MEDLINE

Garber JE, Liepman MK, Gelles EJ, Coson JM and Antman KH. (1990). Cancer 66, 2432-2434. MEDLINE

Harper JW, Adami GR, Wei N, Keyomarsi K and Elledge SJ. (1993). Cell 75, 805-816. MEDLINE

Hartley AL, Birch JM, Kelsey AM, Marsden HB, Harris M and Teare MD. (1989). Cancer Genet. Cytogenet. 42, 221-226. MEDLINE

Harvey M, Vogel H, Morris D, Bradley A, Bernstein A and Donehower LA. (1995). Nature Genet. 9, 305-311. MEDLINE

Hollstein M, Sidransky D, Vogelstein B and Harris CC. (1991). Science 253, 49-54. MEDLINE

Hu G, Zhang W and Deisseroth AB. (1992). Br. J. Haematology 81, 489-494.

Ishioka C, Englert C, Winge P, Yan Y-X, Engelstein M and Friend SH. (1995). Oncogene 10, 1485-1492. MEDLINE

Ishioka C, Shimodaira H, Englert C, Shimada A, Osada M, Jia L-Q, Suzuki T, Gamo M and Kanamaru R. (1997). Biochem. Biophys. Res. Comm. 232, 54-60. MEDLINE

Jost CA, Marin MC and Kaelin WG. (1997). Nature 389, 191-194. Article MEDLINE

Kaghad M, Bonnet H, Yang A, Creancier L, Biscan JC, Valent A, Minty A, Chalon P, Lelias JM, Dumont X, Ferrara P, McKeon F and Caput D. (1997). Cell 90, 809-819. MEDLINE

Kastan MB, Onyekwere O, Sidransky D, Vogelstein B and Craig RW. (1991). Cancer Res. 51, 6304-6311. MEDLINE

Kastan MB, Zhan Q, el-Deiry WS, Carrier F, Jacks T, Walsh WV, Plunkett BS, Vogelstein B and Fornace Jr AJ. (1992). Cell 71, 587-597. MEDLINE

Kawamura M, Yamashita T, Segawa K, Kaneuchi M, Shindoh M and Fujinaga K. (1996). Oncogene 12, 2361-2367. MEDLINE

Kern SE, Pietenpol JA, Thiagalingham S, Seymour A, Kinzler KW and Vogelstein B. (1992). Science 256, 827-830. MEDLINE

Kim E, Albrechtsen N and Deppert W. (1997). Oncogene 15, 857-869. MEDLINE

Kleihues P, Schauble B, zur Hausen A, Esteve J and Ohgaki H. (1997). Am. J. Pathol. 150, 1-10. MEDLINE

Kuerbitz SJ, Plunkett BS, Walsh WV and Kastan MB. (1992). Proc. Natl. Acad. Sci. USA 9, 7491-7495.

Lang P, Maurer J, Haher D, Bornhoft G, Montenarh M, Berdel WE and Thiel E. (1993). Int. J. Oncology 2, 347-355.

Levine AJ. (1997). Cell 88, 323-331. MEDLINE

Li FP and Fraumeni Jr JF. (1969a). J. Natl. Cancer Inst. 43, 1365-1373. MEDLINE

Li FP and Fraumeni Jr JF. (1969b). Ann. Intern. Med. 71, 747-752. MEDLINE

Li FP. (1988). Cancer Res. 48, 5358-5362. MEDLINE

Lowe SW, Ruley HE, Jacks T and Housman DE. (1993). Cell 74, 957-967. MEDLINE

Malkin D, Li FP, Strong LC, Fraumeni Jr. JF, Nelson CE, Kim DH, Kassel J, Gryka MA, Bischoff FZ, Tainsky MA and Friend SH. (1990). Science 250, 1233-1238. MEDLINE

Malkin D. (1998). The Genetic Basis for Human Cancer. 1st ed. Vogelstein B, Kinzler KW (eds). McGraw-Hill, New York, pp. 393-408.

Mashiyama S, Murakami Y, Yoshimoto T, Sekiya T and Hayashi K. (1991). Oncogene 6, 1313-1318. MEDLINE

Milner J and Medcalf EA. (1991). Cell 65, 765-774. MEDLINE

Miyashita T, Harigai M, Hanada M and Reed JC. (1994a). Cancer Res. 54, 3131-3135. MEDLINE

Miyashita T, Krajewski S, Krajewska M, Wang HG, Lin HK, Liebermann DA, Hoffman B and Reed JC. (1994b). Oncogene 9, 1799-1805. MEDLINE

Miyashita T and Reed JC. (1995). Cell 80, 293-299. MEDLINE

Mulligan LM, Matlashewski GJ, Scrable HJ and Cavanee WK. (1990). Proc. Natl. Acad. Sci. USA 87, 5863-5867. MEDLINE

O'Connor PM, Jackman J, Bae I, Myers TG, Fan S, Mutoh M, Scudiero DA, Monks A, Sausville EA, Weinstein JN, Friend S, Fornace AJ and Kohn KW. (1997). Cancer Res. 57, 4285-4300. MEDLINE

Osborne RJ, Merlo GR, Mitsudomi T, Venesio T, Liscia DS, Cappa AP, Chiba I, Takahashi T, Nau MM, Callahan R and Minna JD. (1991). Cancer Res. 51, 6194-6198. MEDLINE

Paules RS, Levedakou EN, Wilson S, Innes CL, Rhodes N, Tlsty TD, Galloway DA, Donehower LA, Tainsky MA and Kaufmann WK. (1995). Cancer Res. 55, 1763-1773. MEDLINE

Peng H-Q, Hogg D, Malkin D, Bailey D, Gallie BL, Bulbul M, Jewett M, Buchanan J and Goss PE. (1993). Cancer Res. 53, 3574-3578. MEDLINE

Prosser J, Porter D, Coles C, Condie A, Thompson AM, Chetty U, Steel CM and Evans HJ. (1992). Br. J. Cancer 65, 527-528. MEDLINE

Purdie CA, Harrison DJ, Peter A, Dobbie L, White S, Howie SE, Salter DM, Bird CC, Wyllie AH, Hooper ML and Clarke AR. (1994). Oncogene 9, 603-609. MEDLINE

Sah VP, Attardi LD, Mulligan GJ, Williams BO, Bronson RT and Jacks T. (1995). Nature Genet. 10, 175-180. MEDLINE

Shaw P, Bovey R, Tardy S, Sahli R, Sordat B and Costa J. (1992). Proc. Natl. Acad. Sci. USA 89, 4495-4499. MEDLINE

Smith ML, Chen I-T, Zhan Q, Bae I, Chen CY, Gilmer TM, Kastan MB, O'Connor PM and Fornace Jr AJ. (1994). Science 266, 1376-1380. MEDLINE

Strong LC, Stine M and Norsted TL. (1987). J. Natl. Cancer Inst. 79, 1213-1220. MEDLINE

Takahashi T, Nau MM, Chiba I, Birrer MJ, Rosenberg RK, Vinocour M, Levitt M, Pass H, Gazdar AF and Minna JD. (1989). Science 246, 491-494. MEDLINE

Varley JM, McGown G, Thorncroft M, Santibanez-Koref MF, Kelsey AM, Tricker KJ, Evans DG and Birch JM. (1997). Cancer Res. 57, 3245-3252. MEDLINE

Walker KK and Levine AJ. (1996). Proc. Natl. Acad. Sci. USA 93, 15335-15340. Article MEDLINE

Wang Y and Prives C. (1995). Nature 376, 88-91. MEDLINE

Williams WC and Strong LC. (1985). Familial Cancer. Muller HR and Weber W (eds.) Karger, Basel pp. 151-153.

Xiong Y, Hannon GJ, Zhang H, Casso D, Kobayashi R and Beach D. (1993). Nature 366, 701-704. MEDLINE

Yonish-Rouach E, Resnitzky D, Lotem J, Sachs L, Kimchi A and Oren M. (1991). Nature 352, 345-347. MEDLINE

Figure 1 (a) Pedigree of LFS family in which the proband harbors germline p53 mutations on both alleles. (b) Pedigree of an LFS-like family with the R290H germline mutation. BB, bilateral breast cancer at age 30; CNS-brain tumor; ME-melanoma; NS-neurofibrosarcoma; OS, osteosarcoma; RMS, rhabdomyosarcoma; SS, soft tissue sarcoma. , unaffected female; male with cancer; Ø-deceased female

Figure 2 Sequence of clones obtained from exons 5 and 8 from the proband. R, arginine; H, histidine; Q, glutamine; C, cytosine; T, thymine; A, adenine; , direction of sense strand

Figure 3 Colonies formed by yeast transformed by p53 constructs. (a) R156H transformants are white; (b) R267Q transformants are red; (c) R290H transformants are white; (d) R156H/R267Q transformants are red

Figure 4 Protein expression in Saos-2 cells 48 h following transfection with various p53 constructs. Expression of p53 and p21^WAF1/CIP1 result from transfection with wild-type p53 and levels vary among transfectants of different p53 constructs harboring different alterations

Figure 5 Colonies formed by 2´10⁵ Saos-2 cells transfected with various p53 cDNA expression constructs. Each set of transfected cells was split 1 : 3, grown in 500 g/ml G418 for 3 weeks, and stained with methylene blue

 Expression of p53 and p21^WAF1/CIP1 in Saos-2 transfectants

 Colonies formed by Saos-2 transfectants

 Summary of functional properties for p53 variants