A serine 37 mutation associated with two missense mutations ... - Nature

ã

Oncogene (2000) 19, 5098 ± 5105 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc

SHORT REPORT

A serine 37 mutation associated with two missense mutations at highly conserved regions of p53 aect pro-apoptotic genes expression in a T-lymphoblastoid drug resistant cell line Caterina Cinti*,1,2, Pier Paolo Claudio2,3, Antonio De Luca4, Monica Cuccurese5, Candace M Howard2, Maurizio D'Esposito5, Marco G Paggi4, Dario La Sala8, Livio Azzoni7, Thanos D Halazonetis6, Antonio Giordano2 and Nadir Mario Maraldi1,8 1

Institute of Normal and Pathologic Cytomorphology, CNR, c/o IOR, 40136 Bologna, Italy; 2Department of Pathology, Anatomy and Cell Biology, Thomas Jeerson University, Philadelphia, Pennsylvania, PA 19107, USA; 3Department of `Scienze Odontostomatologiche e Maxillo-Facciali', University of Naples `Federico II', 80131 Napoli, Italy; 4Laboratory for Cell Metabolism and Pharmacokinetics, Center for Experimental Research, Regina Elena Institute, 00158 Roma, Italy; 5Genome Research and Sequencing Laboratory, International Institute of Genetics and Biophysics, CNR, 80125 Napoli, Italy; 6Departement of Molecular Oncology, The Wistar Institute, Philadelphia, Pennsylvania, PA 19104-4268, USA; 7Kimmel Cancer Institute, Thomas Jeerson University, Philadelphia, Pennsylvania, PA 19107, USA; 8Laboratory of Cell Biology and Electron Microscopy, IOR, 40136 Bologna, Italy

The p53 protein accumulates rapidly through posttranscriptional mechanisms following cellular exposure to DNA damaging agents and is also activated as a transcription factor leading to growth arrest or apoptosis. Phosphorylation of p53 occurs after DNA damage thereby modulating its activity and impeding the interaction of p53 with its negative regulator oncogene Mdm2. The serines 15 and 37 present in the amino terminal region of p53 are phosphorylated by the DNAdependent protein kinase (DNA-PK) in response to DNA damage. In order to verify if speci®c p53 mutations occur in the multi-drug resistance phenotype, we analysed the p53 gene in two T-lymphoblastoid cell lines, CCRFCEM and its multi-drug-resistant clone CCRF-CEM VLB100, selected for resistance to vinblastine sulfate and cross-resistant to other cytotoxic drugs. Both cell lines showed two heterozygous mutations in the DNA binding domain at codons 175 and 248. The multi-drug resistant cell line, CCRF-CEM VLB100, showed an additional mutation that involves the serine 37 whose phosphorylation is important to modulate the protein activity in response to DNA damage. The eects of these mutations on p53 transactivation capacity were evaluated. The activity of p53 on pro-apoptotic genes expression in response to DNA damage induced by (-irradiation, was aected in the vinblastine (VLB) resistant cell line but not in CCRF-CEM sensitive cell line resulting in a much reduced apoptotic cell death of the multi-drug resistant cells. Oncogene (2000) 19, 5098 ± 5105. Keywords: drug resistance; p53 mutations; apoptotic genes; T-lymphoblastoid cells Wild-type p53 is a negative regulator of cell proliferation, and its overexpression inhibits both

*Correspondence: C Cinti, Istituto di Citomorfologia Normale e Patologica, CNR, c/o IOR, via di Barbiano 1/10, 40136 Bologna, Italy Received 22 September 1998; revised 30 July 2000; accepted 7 August 2000

normal and tumor cell growth. p53 is phosphorylated at sites within its N-terminal and C-terminal regions, and several protein kinases have been shown to phosphorylate p53 in vitro. There is evidence that phosphorylation and dephosphorylation may be the key regulatory steps of the p53 function to mediate signal transduction from the cell surface to the nucleus (Karin and Hunter, 1995; Meek, 1997; Milczarek et al., 1997). Because a rapid increase in p53 levels has been observed following exposure of cells to g-irradiation or other DNA damaging agents, it is believed that p53 arrests cells in the G1 phase in response to DNA damage (Fritsher et al., 1993; Levine, 1997). Large spectrums of tumors have inactivated wild-type p53, and accumulation of p53 is absent or delayed in some types of cells defective in DNA repair (Lu and Lane, 1993; Nelson and Kastan, 1994). Wild-type p53 activates expression of genes acting as mediators of its function after DNA damage such as the Gadd45, p21/WAF1 and Bax causing G1 arrest and/or apoptosis (Sang et al., 1995). An important clue for understanding the mechanism underlying the tumor suppressor activity of p53 was provided by the indication that p53 aects apoptosis. In lymphocytes, DNA damage induces apoptosis after exposure to ionizing radiation or cytotoxic drugs (Cohen et al., 1992). The `guardian of the genome,' p53, plays a central role in radiation responses including cell growth arrest and apoptosis, and its loss induces the progression into S phase without repairing DNA damage (Kastan et al. 1992; Levine 1997). There is evidence that cells with wild-type p53 show increased apoptosis after g-irradiation. Cells harboring homozygous p53 mutants are completely resistant to induction of apoptosis, while the heterozygous p53 mutants show an intermediate response (Hooper, 1994). Indeed, many studies suggest that the presence of p53 mutations may determine the lack of tumors response to chemotherapy (Houldsworth et al. 1998), increased tumor aggressiveness, shortened periods of remission and reduced overall survival rates (Shelling, 1997).

p53-Ser 37 mutation affecting apoptotic response C Cinti et al

In the present study we investigated the presence of dierent p53 mutations in two T-lymphoblastoid cell lines, CCRF-CEM and its multi-drug resistant variant, CCRF-CEM VLB100, respectively sensitive and resis-

Figure 1 (a) SSCP analysis of exon 4 of the p53 gene: 1 wildtype p53, 2 CCRF-CEM and 3 CCRF-CEM VLB 100. Genomic DNA was ampli®ed by exon to exon PCR. The primer panels to amplify exons 2 ± 11 but exon 4 of p53 and the PCR reaction conditions were those recommended by Clontech Laboratories. Exon 4 was ampli®ed with the following set of primers: P4ca 5'GGGAAGCGAAAATTCATGGGAC-3' and P4cb 5'-AGACTTCAATGCCTGGCCGTAT-3'. PCR products were electrophoresed into an MDE gel at 8 W constant power for 8 h at room temperature, in 0.66 TBE running buer and the SSCP analysis was performed as recommended by FMC BioProducts. (b) Restriction enzyme analysis of exons 5 and 7: 1 DNA molecular weight marker VI (Boehringer Mannheim); 2 Wild-type PCR product of exon 5 after HpaI digestion: four fragments of 112, 49, 29 and 18 bp were visible; 3 and 4 CCRF-CEM and CCRF-CEM VLB100 cells PCR products of exon 5, respectively showing an extra fragment of 67 bp derived by lack of the HpaI restriction site; 5 Wild-type PCR product of exon 7 after MspI digestion: two fragments of 83 and 56 bp were visible; 6 and 7 CCRF-CEM and CCRF-CEM VLB100 cells PCR products of exon 7, respectively showing an extra fragment of 139 bp derived by lack of the MspI restriction site. (c,d) Sequence analysis of exon 4 PCR products of CCRF-CEM (c) and CCRF-CEM VLB 100 cells (d) in which is visible the substitution of thymine to cytosine at codon 37. Sequences of PCR products were carried out by automated DNA sequencing, using the dideoxy-terminator reaction chemistry for sequence analysis on the Applied Biosystem model 373A DNA sequencer

tant to cytotoxic drugs: vinblastine sulfate, actinomycin-D and adriamycin (Cianfriglia et al., 1991) by SSCP and sequence analysis of exon PCR products. The transactivating activity of p53 mutants on proapoptotic genes was investigated before and after exposure to g-irradiation, to which CCRF-CEM VLB 100 cells developed an increased resistance against radio-induced apoptosis. The transcriptional activity of the wild-type p53 and of the Pro37 mutant combined with His175 or Gln248 were examined by transient transfection in Saos-2 osteosarcoma cells, which lack endogenous p53 (Diller et al., 1990). We demonstrated that the mutation found at serine 37 resulted in diminished p53 transactivating capacity and therefore in its altered function. p53 exons (2 ± 11) were ampli®ed and analysed by SSCP to detect the presence of point mutations (Moyret et al., 1994). Exon 4 PCR products of sensitive and resistant cell lines, CCRF-CEM and CCRF-CEM VLB 100, showed dierent SSCP migration patterns when compared to the control suggesting the presence of further mutations (Figure 1a). Direct sequencing of exon 4 revealed in the CCRF-CEM cells a CGC?CCC transversion at codon 72, while the CCRF-CEM VLB100 cells showed the same polymorphism at codon 72 and a mutation TCC?CCC at codon 37. This mutation results in the substitution of serine to proline (Figure 1c,d, Table 1). Since in CCRF-CEM cells two heterozygous mutations were found at codon 175 of exon 5 and at codon 248 of exon 7, causing amino acid substitutions of arginine to histidine and arginine to glutamine, abolishing the HpaI and the MspI restriction site, respectively (Cheng and Haas, 1990); we digested the exon 5 and exon 7 PCR products of CCRF-CEM and CCRF-CEM VLB100 cells to verify the presence of mutations in these regions. The results of the restriction enzyme analysis are shown in Figure 1b, and con®rm the presence of heterozygous mutations at codons 175 and 248 in both cell lines. Human CCRF-CEM T-lymphoblastoid cell line and its multidrug-resistant variant, CCRF-CEM VLB100 cells, were examined for their sensitivity to g-irradiation. The proportion of surviving cells at the dose of 5 Gy, 48 h after irradiation, was 52.2% and 77.1% for sensitive and resistant cells, respectively. Three days later, the corresponding percentages were 39.5% and 63.9%, respectively, indicating that the CCRF-CEM VLB100 cells presented an increased radioresistance when compared to CCRF-CEM cells (Figure 2a, Table 2). In order to determine whether the observed decrease in viable cell counts was due to cell death through apoptosis, cells were ®xed 48 h post-treatment,

5099

Table 1 p53 gene mutations in human T-lymphoblastoid cells Cell lines

Codons Mutation site sequenced (codon)

CCRF-CEM

33 ± 125 133 ± 287

CCRF-CEM VLB 100

26 ± 125 133 ± 287

72 175 248 37 72 175 248

Mutation nucleotide

Amino acid substitutions

CGC?CCC CGC?CAC CGG?CAG TCC?CCC CGC?CCC CGC?CAC CGG?CAG

Arg?Pro Arg?His Arg?Glyn Ser?Pro Arg?Pro Arg?His Arg?Gln Oncogene


5100

A

B

C

Figure 2 (a) Viability of T-lymphoblastoid cell lines treated with 5 Gy of g-irradiation. I=error bars. CCRF-CEM human Tlymphoblastoid cell lines, sensitive and resistant to chemotherapeutic drugs, were cultured in RPMI 1640 medium supplemented with 10% FCS. The multidrug-resistant cell line was cultured in the presence of 100 ng/ml of vinblastine sulfate (Lilly), by which the name CCRF-CEM VLB100. The amount of resistance was determined by evaluating IC50 with respect to that of the parental cell line to the following drugs: vinblastine, actinomycin-D and adriamycin (Cianfriglia et al. 1991). Cells were plated in a 4 cm2 cell culture dish at a density of 16106 cells/dish, and irradiated using a Cs source for an appropriate length of time to deliver a preselected dose of 5 Gy. The viability of the irradiated cells was determined by erytrosin B dye exclusion within 48 h to 1 week after irradiation. The experiment was repeated three times. (b) Apoptotic cells 48 h after g-irradiation were examined by electron microscopy. The compact chromatin organization into cap-shaped (?) and/or micronuclei (?) structures are visible. Cell pellets were ®xed with 1% glutaraldehyde in 0.1 M phosphate buer, post-®xed in 1% osmium tetroxide, dehydrated in ethanol and embedded in Epon. Thin sections were stained with uranyl acetate and lead citrate and examined in a Zeiss EM 109 electron microscope to detect the apoptotic nuclear characteristics of irradiated cells. (c) Typical oligonucleosome DNA fragmentation Oncogene


Table 2 Quantitative data regarding the percentage and standard errors of cell proliferation with and without g-irradiation treatment and statistical signi®cance of dierence in radiation sensitivity calculated by chi-test CCRF-CEM 48 h 72 h 96 h

93+1.7 91+0.9 96.1+2 CCRF-CEM*

% VE+SE CCRF-CEM VLB100 94.8+5.1 93.75+0.7 96+1.4 % V+SE CCRF-CEM* VLB100

CCRF-CEM 7.0+1.7 9+1.0 3.9+1.3 CCRF-CEM*

5101

% M+SE CCRF-CEM VLB100 5.2+1.7 6.25+1.0 4+0.7 % M+SE CCRF-CEM* VLB100

48 h

52.2+3.1

77.1+2.4

47.8+1.5

22.9+1.6

72 h

41.6+0.5

64+0.9

58.4+0.8

36+0.8

96 h

39.5+1.6

63.9+1.9

60.5+3.2

36.1+5.8

w2=10.647 P=0.001 w2=22.076 P=0.001 w2=3.726 P=0.05

*=irradiated cells; V=viability; M=mortality; SE=standard error

and thin sections were examined by electron microscopy. A typical chromatin margination organized into cap-shaped compact structures or into micronuclei was visible (Figure 2b). Additional markers such as DNA ladder and in situ Nick Translation were performed at 48, 72 and 96 h to detect DNA cleavage due to apoptosis. Figure 2c shows a typical oligonucleosomic DNA ladder, characteristic of apoptosis, at 48, 72 and 96 h after irradiation. In the in situ Nick Translation the propidium iodide (PI) signal (red) demonstrates the areas of double-stranded DNA, while the FITC signal (green), shows the distribution of nicked areas corresponding to the sites of cleaved DNA (Figure 2d,f). The percentage of cells exhibiting apoptotic bodies was markedly higher in CCRF-CEM cells when compared to CCRF-CEM VLB 100 cells. Quantitative data regarding both cell lines undergoing apoptosis after g-irradiation and the statistical signi®cance of the dierences in radiation sensitivity between both cell lines was calculated by the chi-test and indicated that all data were statistically signi®cant (P40.05) (Table 2). These results suggested that g-irradiation induced cell death through an apoptotic mechanism in the sensitive cell line to a greater extent than that observed in the resistant cell line. The failure of the CCRF-CEM VLB 100 cells to undergo apoptosis may be due to the additional mutation of the Ser at codon 37, which determines the lack of functional p53. To verify whether or not the p53 mutant form, present in the resistant cell line, may aect the transactivating activity of pro-apoptotic and antiapoptotic genes, we analysed the p53, Mdm2, p21/ WAF1, Bax and bcl-2 proteins' expression before and 48 h after 5 Gy of g-irradiation (Figure 3a). In the parental cell line, the p53 level was increased markedly by g-irradiation 48 h post-treatment, which

may depend on an increase in transcriptional rate as reported in other cells after genotoxic insults (Tishler et al., 1995). These cells showed unmodi®ed levels of Bax and Mdm2 proteins and increased levels of p21/WAF1 48 h after irradiation suggesting that the heterozygous p53 mutant maintained its transactivation capability for these genes as reported by Park et al. (1994) and Chen and Haas (1990). On the contrary, CCRF-CEM VLB100 cells showed decreased levels of all proapoptotic proteins after g-irradiation suggesting that the substitution of Ser with Pro at codon 37, in association with the heterozygous mutations at codons 175 and 248 caused a reduction in the p53 transactivating activity and its growth suppressive ability. Levels of anti-apoptotic bcl-2 protein were instead unchanged in both cell lines upon irradiation. The Ser 37 to Pro substitution targets the transactivation domain of p53 (Fields and Jang 1990; Unger et al., 1992), raising the possibility that this substitution could aect the proteins' transcriptional activity. In order to further clarify the mechanism by which the heterozygous Ser37 to Pro mutation alone or the combination of this with the other two mutations (Arg175 to His and Arg248 to Gln) may result in abolished function of the p53 alleles, the transcriptional activities of wild-type p53 and the mutants were examined by transient transfection in Saos-2 osteosarcoma cells, which lack endogenous p53 (Diller et al., 1990). A series of luciferase assays were performed using the minimal p21/WAF1 and full-size Bax promoters responsive to p53 and the p53 promoter itself, to evaluate the transactivation activity of p53 mutants. As shown in Figure 3b,c, the Pro37 mutant activated the p21/WAF1 and p53 promoters transcriptional activity with reduced potency with respect to wild-type

detected after irradiation. Genomic DNAs of treated and untreated cells were extracted at 48, 72 and 96 h following a standard protocol (Maniatis). Puri®ed DNAs were run on a 1% agarose gel, and stained with ethidium bromide. (d ± f) In situ Nick Translation of apoptotic cells 48 (d), 72 (e) and 96 h (f) after g-irradiation. Pellets of both irradiated cell lines were ®xed in 3 : 1 methanol-acetic acid at 48, 72 and 96 h after radiation treatment. Cells were spread on glass slides and an in situ Nick Translation was performed using a mix of dNTP digoxigenin complexed. A ¯uoresceinated isothiocyanate (FITC) complexed anti-digoxigenin antibody diluted 1 : 500 was used to detect the DNA nicked areas. Finally, nuclei were counterstained with 1 mg/ml propidium iodide (PI) and observed by confocal laser scanning microscopy (CSLM) equipped with an argon ion laser attached to a Nikon Optiophot-2 camera. Confocal microscopy images were elaborated and reconstructed as three-dimensional (3D) look-through projections. Optical sections are obtained as increments of 0.3 mm in the Z-axis with a scanning mode format of 5126512 pixels. Normal non-apoptotic cells served as an internal control Oncogene


5102

Figure 3 (a) Western blot analysis of Mdm2, p53, bcl-2, Bax and p21/WAF1 before and after 48 h of g-irradiation treatment. R=CCRF-CEM VLB100; Rg=CCRF-CEM VLB100 treated with 5 Gy of g-irradiation; S=CCRF-CEM; Sg=CCRF-CEM treated with 5 Gy of g-irradiation. Cells were lysed and fractionated into a 15% sodium dodecyl sulfate-polyacrylamide gel, followed by electrophoretic transfer of the proteins to a PVDF membrane (Millipore). Immunoblotting was performed following manufacturer's instructions with the following antibodies: Waf1/p21, p53/AB6 and Mdm2 monoclonal antibodies (Oncogene Science), Bax and bcl2 polyclonal antibodies (Santa Cruz Biotechnology). An equal amount of 30 mg of total cell lysate was used for each sample. (b,c) Luciferase assay of the minimal p21/WAF1 promoter, p53 and full-size Bax promoters to evaluate the transactivating activity of p53 mutants. Saos 2 cells were transiently transfected using a standard calcium phosphate method (Claudio et al., 1994). Luciferase assays were performed according to manufacturer's instructions (Promega). Transfections were normalized by cotransfecting the pSVbGAL plasmid (Promega) and by performing a standard b-galactosidase assay. One mg of pSVbGAL was co-transfected along with 2 mg of the luciferase p21/WAF1 (p21 Luc) or p53 (PG13Pyluc) promoters or of the luciferase Bax (pHBaxPF) promoter and 5 mg of either pCDNA3 p53 wild-type, p53S37P, p53-S37P/R175H, or p53-S37P/R248Q. The total amount of CMV promoter was maintained constant in each transfection. There was no signi®cant dierence in the transfection eciencies among the various constructs as determined by the b-gal activity. Luciferase assays were performed in triplicates and were repeated three times. Standard deviation values are also indicated. Mutations were introduced in p53 using a PCR based protocol. Two dierent mutated proteins were designed either mutated in Ser 37 (Ser to Pro; TCC?CCC) and Arg 175 (Arg to His; CGC?CAC) or in Ser 37 and Arg 248 (Arg to Gln; CGG?CAG). The following sets of primers were used, and the mutations introduced are underlined and in bold. For plasmid p53-S37P: p53/1 (forward) 5'CGT TCT GTC CCC CTT GCC GCC CCA 3'; p53/2 (reverse) 5'TGG GGC GGC AAG GGG GAC AGA ACG 3'; T7L (external) 5'TAA TAC GAC TCA CTA TAG GGA GAC 3'; p53/3 (external) 5'GTG CTG TGA CTG CTT GTA GAT GGC 3'. For plasmid p53-S37P/R175H: p53/5 (forward) 5'GAC GGA GGT TGT GAG GCA CTG C 3'; p53/6 (reverse) 5'GCA GTG CCT CAC AAC CTC CGT C 3'; T7L (forward) 5'TAA TAC GAC TCA CTA TAG GGA GAC 3'; p53/9 (external) 5'TTT TAT GGC GGG AGG TAG ACT G 3'. For plasmid p53-S37P/ R248Q: p53/7 (external) 5'ATG GGC GGC ATG AAC CAG AGG C 3'; p53/8 (reverse) 5'GCC TCT GGT TCA TGC CGC CCA T 3'; T7L (external) 5',TAA TAC GAC TCA CTA TAG GGA GAC 3'; p53/9 (forward) 5'TTT TAT GGC GGG AGG TAG ACT G 3'. PCR products were subcloned into the pRC/CMV expression vector (InVitrogen) and mutations were con®rmed by direct sequencing of the constructs. (b) Eect of wild-type p53 (column 3) and p53-S37P, p53-S37P/R175H, p53-S37P/R248Q mutants (columns 4, 5, 6) on the activity of the p21/WAF1 (Luc p21) promoter. (c) Eects of wild-type p53 (columns 3, 8) and p53-S37P, p53-S37P/R175H, p53-S37P/ R248Q mutants (columns 4, 5, 6, 9, 10, 11) on the activity of the p53 (luc p53) and Bax (Luc Bax) promoters Oncogene


p53, and this activity was completely abolished by the Pro37 mutation combined with His175. The double mutation Pro37/Gln248 also reduced the activity of the p53 mutant with respect to wild-type but at a lesser extent than that of the Pro37/His175 mutant on the p21/WAF1 promoter. The double mutants Pro37/ Gln248 and Pro37/His175 were equally eective in abolishing the transcriptional activity of the p53 promoter. These results indicate that the single serine substitution at the DNA binding domain is sucient to aect the transcriptional activity of p53 and that multiple mutations can be synergistic in aecting p53 transcriptional activity. Overexpression of wild-type p53 did not transactivate the full-size human Bax promoter in Saos-2 cells. This is in agreement with previously published data (Igata et al., 1999). Furthermore the double mutants did not have any signi®cant eect on the transcriptional activity of the Bax promoter. The tumor suppressor protein p53 modulates cell proliferation and plays a crucial role in suppressing tumor development (Ullrich et al., 1992; Levine, 1993; Ko and Prives, 1996). Additionally, p53 is an important component of the apoptotic pathway in response to DNA damage induced by therapeutic agents such as anti-metabolites and radiation (Williams and Smith, 1993). Loss of p53 suppressor function through mutations or deletions is the most common genetic change associated with a wide variety of human cancers (Hollstein et al., 1991; Levine et al., 1991). The role of mutant p53 in determining the sensitive/ resistant status appears to be cell-type dependent (Delia et al., 1996; Righetti et al., 1996) and dierent mutations may vary in their ability to aect p53 function, including the capability of the protein to induce apoptosis after DNA damage (Hooper, 1994). Drug resistant cells are often resistant to dierent chemotherapeutic agents and g-irradiation. Gene ampli®cation, translocation or rearrangement of chromosomes and single mutations ensure ecient selection and overgrowth of drug resistant tumors' cells during and after chemotherapy (Lowe et al., 1993; Harris and Hollstein, 1993; Kerbel, 1997). In this study we identi®ed an additional mutation of p53, which is present in the resistant CCRF-CEM VLB 100 cell line but not in the parental sensitive Tlymphoblastoid cell line. The mutation found involved one of the serines present in the N-terminal region of p53. This is one of the sites that becomes phosphorylated after DNA damage and putatively disrupt p53mdm2 binding leading to accumulation of p53 and enabling transactivation of downstream target genes. To further clarify the mechanism by which this mutation may result in altered p53 protein function, we engineered a series of mutants reproducing the naturally occurring mutations in the multi-drug resistant cell line. Ectopic expression of the Pro37 mutant or of the double mutants, Pro37/His175 or Pro37/Gln248, in p53 null cells, resulted in the downregulation of the p53 and p21/WAF1 promoters' activity. In particular, the serine 37 mutation alone is able to reduce the transcriptional activity of p53 on the p53 and p21/WAF1 promoters. This eect is enhanced by the double mutations. This data suggest that the serine 37 mutation may be responsible for the clonogenic survival of the multi-drug resistant cells

(CCRF-CEM VLB 100), which are more refractory to the cytotoxic eects of chemotherapeutics. Loss of wild-type p53 activity could potentially account for the increased radio- and chemo-resistance of tumors bearing p53 mutations. The radiation viability curve, quantitative and qualitative determination of apoptosis by electron microscopy, DNA ladder, in situ Nick Translation, and Western blot analysis of p53dependent proteins' expression, supported this data. In fact, the CCRF-CEM VLB 100 cells exhibit a reduced degree of apoptotic cell death and lack of induction of p53-responsive genes when compared to sensitive parental cells. Since p53 is a transcriptional activator factor, the expression or loss of expression of p53 dependent gene products might aect drug sensitivity and apoptotic response induced by ionizing radiation or by chemotherapeutic drugs. The determinants of a cell's decision to enter apoptosis or G1 arrest appear to involve a number of factors among which include the bcl2/bax ratio, p21/WAF1 and other transactivating factors (Miyashite and Reed, 1995; Oltvai et al., 1993; Polyak et al., 1996; White, 1996). In CCRF-CEM VLB 100 cells p53-responsive genes such as Mdm2, p21/WAF1, p53 and Bax were down regulated at the protein level after g-irradiation. For the p53 and p21/WAF1 promoters, the p53 Pro37 mutant and the double mutants, Pro37/His175 and Pro37/Gln248 drastically inhibited p53 transcriptional activity in a synergistic manner (Figure 3b,c). Although transcriptional regulation has been proposed as one of the most important factors regulating the expression of the Bax protein and the p53 protein has been reported to transactivate fragments of the human Bax promoter (Miyashite and Reed, 1995), we found that neither wild-type p53 nor the mutants transactivated the Bax promoter in Saos-2 cells, which agrees with previously published data (Igata et al., 1999). Furthermore, others have shown that the p53-binding element in the Bax promoter has a lower anity for p53 than that of other p53 responsive promoters (Friedlander, 1996). Additionally, other groups have reported that human p53 positively or negatively transactivates speci®c human gene promoters, such as the retinoblastoma gene (RB), depending on the size and location of the promoter region (Osifchin et al., 1994; Shiio et al., 1992). For this reason we choose the full-length human Bax promoter for our studies to re¯ect physiologic conditions. Even though Bax is one of the key proteins in initiating the apoptotic cascade, a full apoptotic response requires other p53-dependent factors (McCurrach et al., 1997). The decrease in Bax protein expression (Figure 3a) in the resistant cells (CCRFCEM VLB 100) following g-irradiation may possibly be due to post-translational changes in the stability of the protein. Recent results indicate that the regulation of the DNA-binding activity of p53 is modulated by phosphorylation at speci®c sites (Hecker et al., 1996; Shieh et al., 1997). In particular, the serines 15 and 37, located into the transcriptional activation domain of p53, are phosphorylated by double stranded DNA dependent protein kinase, DNA-PK. The phosphorylation of these two serines alters the protein's antiproliferative activity (Lees-Miller et al., 1992; Ullrich et al., 1993). Hyperphosphorylation may contribute to the acidic nature of this domain, thus enhancing the

5103

Oncogene


5104

transactivation function of p53, while mutation of the serine 15 phosphorylation site reduces the growth suppressive ability of the protein by about 50% (Fiscella et al., 1993). It has been recently demonstrated that a single point mutation can cause signi®cant conformational changes of p53 resulting in loss of p53 availability as a substrate for a number of kinases (Adler et al., 1997). The activation of DNA-PK requires DNA double strand breaks induced by ionizing radiation, and it has been speculated that this kinase plays a role in DNA repair (Finnie et al., 1995; Lees-Miller et al., 1995). Since the DNA-PK sites are located in the vicinity of the Mdm2 interaction motif, it has been suggested that serine 15 and 37 phosphorylation alters the tertiary structure of p53 and aects the interaction with Mdm2. These serines negatively regulate the interaction between p53 and Mdm2 reducing their association and concomitantly increase the transcriptional activity of p53 (Mayr et al., 1995; Shieh et al., 1997). It is reasonable to suppose that the serine 37 mutation alone or combined with the two other heterozygous mutations that we found aect, in the resistant cells, the p53 growth suppressive activity with consequential down-regulation of apoptotic related genes. Whereas, heterozygous mutations at codons 175 and 243 in CCRF-CEM cells have been reported to aect but not abolish p53 DNA binding ability and to maintain its transactivating activity on pro-apoptotic genes (Park et al., 1994). Site directed mutagenesis experiments demonstrate that mutation of serine 37 is sucient to alter p53 DNA-binding properties and transactivating function. The addition of this mutation to the two other heterozygous mutations, that alone are unable to abolish the activity of the protein, are necessary to

aect p53 functions. These ®ndings suggest that some forms and/or combinations of p53 mutants may directly enhance the resistance of tumor cells to anti cancer agents in a manner that depends on the particular mutation as shown by Blandino et al. (1999). The data reported here stress the need for evaluating the possible combinatorial eects of p53 multiple mutations that may aect radio- and chemo resistance. It is apparent that dierent mutations may vary in their ability to aect p53 function. p53 functional inactivation by mutations could explain the lack of response by tumors to chemotherapeutic treatments. Further functional and structural studies are required to determine the physiological role of these regions as a target of chemo-therapeutics.

Acknowledgments The authors thank Dr WS El-Deiry of Howard Hughes Medical Institute, Philadelphia, USA, for providing the luciferase p21/WAF1 promoter (P21Luc) and luciferase p53 promoter (PG13PYLuc); Prof T Sakai of Kyoto Prefectural University of Medicine, Kyoto, Japan for providing the luciferase human Bax promoter (PhBaxPF). We also thank Dr M Cianfriglia for the CCRF-CEM VLB100 line and Dr N Zini for providing the electron microscopy images. This work was supported by 40 ± 60% MURST, by `Piani potenziamento della rete scienti®ca e tecnologica' MURST, by `Funds for Selected Research Topics' from University of Bologna by `IOR Ricerca Corrente' and CNR grants to C Cinti; by NIH RO1 CA 60999-01A1 and P01 NS 36466 grants to A Giordano; by AIRC and Ministero SanitaÁ grants to MG Paggi; by F.I.R.C. grant to A De Luca; by C.N.R. `Target Project on Biotechnology' grant to M D'Esposito. PP Claudio is the recipient of a fellowship from the `Associazione Leonardo di Capua', Naples, Italy.

References Adler V, Pincus MR, Minamoto T, Fuchs SY, Bluth M, Brandt-Rauf PW, Friedman FK, Robinson RC, Chen JM, Wang XW, Harris CC and Ronai Z. (1997). Proc. Natl. Acad. Sci. USA, 94, 1686 ± 1691. Blandino G, Levine AJ and Oren M. (1999). Oncogene, 18, 477 ± 485. Cheng J and Haas M. (1990). Mol. Cell. Biol., 10, 5502 ± 5509. Cianfriglia M, Yassen A, Tombesi M, Samoggia P, Barca S and Caserta M. (1991). Int. J. Cancer, 49, 394 ± 397. Claudio PP, Howard CM, Baldi A, De Luca A, Condorelli G, Sun Y, Colburn N, Calabretta B and Giordano A. (1994). Cancer Res., 54, 5556 ± 5560. Cohen JJ, Duke RC, Fadok VA and Sellins KS. (1992). Annu. Rev. Immunol., 10, 267 ± 293. Delia D, Mizutani S, Lamorte G, Goi K, Iwata S and Pierotti MA. (1996). Nat. Med., 2, 724 ± 725. Diller L, Kassel J, Nelson CE, Gryka MA, Litwak G, Gebhardt M, Bressac B, Ozturk M, Baker SJ, Volgestein B and Friend SH. (1990). Mol.Cell. Biol., 10, 5772 ± 5781. Fields S and Jang SK. (1990). Science, 249, 1046 ± 1049. Finnie NJ, Gottlieb TM, Blunt T, Jeggo PA and Jackson SP. (1995). Proc. Natl. Acad. Sci. USA, 92, 320 ± 324. Fiscella M, Ullrich SJ, Zambrano N, Shield MT, Lin D, Lees-Miller SP, Anderson CW, Mercer WE and Appella E. (1993). Oncogene, 8, 1519 ± 1528. Friedlander P, Haupt Y, Prives C and Oren M. (1996). Mol. Cell. Biol., 16, 4961 ± 4971. Fritsher M, Haessler C and Brandner G. (1993). Oncogene, 8, 307 ± 318. Oncogene

Harris CC and Hollstein M. (1993). N. Engl. J. Med., 329, 1318 ± 1327. Hecker D, Page G, Lohrum M, Weiland S and Scheidtmann KH. (1996). Oncogene, 12, 953 ± 961. Hollstein M, Sidransky D, Volgestein B and Harris CC. (1991). Science, 253, 49 ± 52. Hooper ML. (1994). J. Cell. Sci. Supp., 18, 13 ± 17. Houldsworth J, Xiao H, Murty VVVS, Chen W, Ray B, Reuter VE, Bosl GJ and Chaganti RSK. (1998). Oncogene, 16, 2345 ± 2349. Igata E, Inoue T, Ohtani-Fujita N, Sowa Y, Tsujimoto Y and Sakai T. (1999). Gene., 238, 407 ± 415. Karin M and Hunter T. (1995). Curr. Biol., 5, 747 ± 757. Kastan MB, Zhan Q, El-deiry WS, Carrier F, Jacks T, Walsh WV, Plunkett BS, Volgelstein B and Fornace JR. AJ. (1992). Cell, 71, 589 ± 597. Kerbel RS. (1997). Nature, 390, 335 ± 336. Ko LJ and Prives C. (1996). Genes Dev., 10, 1054 ± 1078. Lees-Miller SP, Godbout R, Chan DW, Weinfeld M, Day III RS, Barron GM and Allalunis-Turner J. (1995). Science, 267, 1183 ± 1185. Lees-Miller SP, Sakaguchi K, Ullrich SJ, Appella E, Anderson CW. (1992). Mol. Cell. Biol., 12, 5041 ± 5049. Levine A. (1993). Annu. Rev. Biochem., 62, 623 ± 651. Levine AJ. (1997). Cell, 88, 323 ± 331. Levine AJ, Momand J and Finday CA. (1991). Nature, 351, 453 ± 456. Lowe SW, Ruley HE, Jack T and Housman DE. (1993). Cell, 74, 957 ± 967. Lu X and Lane DP. (1993). Cell, 75, 765 ± 778.


Mayr GA, Reed M, Wang P, Wang Y, Schwedes JF and Tegtmeyer P. (1995). Cancer. Res., 55, 2410 ± 2417. McCurrach ME, Connor TM, Knudson CM, Korsmeyer SJ and Lowe SW. (1997). Proc. Natl. Acad. Sci. USA, 94, 2345 ± 2349. Meek DW. (1997). Semin. Cancer Biol., 5, 203 ± 210. Milczarek GJ, Martinez J and Bowden GT. (1997). Life Sci., 60, 1 ± 11. Miyashite T and Reed JC. (1995). Cell, 80, 293 ± 299. Moyret C, Theillet C, Puig PL, MoleÂs JP, Thomas G and Hamelin R. (1994). Oncogene, 9, 1739 ± 1743. Nelson WG and Kastan MB. (1994). Mol. Cell. Biol., 14, 1815 ± 1823. Oltvai ZN, Milliman CL and Korsmeyer SJ. (1993). Cell, 74, 609 ± 619. Osifchin NE, Jiang D, Ohtani-Fujita N, Fujita T, Carroza M, Kim SJ, Sakai T and Robbins PD. (1994). J. Biol. Chem., 269, 6383 ± 6389. Park D, Nakamura H, Chumakov AM, Said JW, Miller CW, Chen DL and Koeer HP. (1994). Oncogene, 9, 1899 ± 1906. Polyak K, Waldman T, He TC, Kinzler KW and Volgelstein B. (1996). Genes Dev., 10, 1945 ± 1952.

Righetti SC, Della Torre G, Pilotti S, MeÂnard S, Ottone F, Colnaghi MI, Pierotti MA, Lavarino C, Cornarotti M, Oriana S, Bohm S, Bresciani GL, Spatti G and Zunino F. (1996). Cancer Res., 56, 689 ± 693. Sang N, Baldi A and Giordano A. (1995). Mol. Cell. Dif., 3, 1 ± 29. Shelling AM. (1997). Lancet, 349, 744 ± 745. Shieh SY, Ikeda M, Taya Y and Prives C. (1997). Cell, 91, 325 ± 334. Shiio Y, Yamamoto T and Yamaguchi N. (1992). Proc. Natl. Acad. Sci. USA, 89, 5206 ± 5210. Tishler RB, Lamppu DM, Park S and Price BD. (1995). Cancer Res., 55, 6021 ± 6025. Ullrich SJ, Anderson CW, Mercer WE and Appella E. (1992). J. Biol. Chem., 267, 15259 ± 15262. Ullrich SJ, Sakaguchi K, Lees-Miller S, Fiscella M, Mercer WE, Anderson CW and Appella E. (1993). Proc. Natl. Acad. Sci. USA, 90, 5954 ± 5958. Unger T, Nau MN, Segal S and Minna JD. (1992). EMBO J., 11, 1383 ± 1392. White B. (1996). Genes Dev., 10, 1 ± 15. Williams GT and Smith CA. (1993). Cell, 74, 777 ± 779.

5105

Oncogene