Expression of the mutated p53 tumor suppressor protein and ... - Nature

5 downloads 0 Views 610KB Size Report
Multidrug resistance in MCF-7/Adr human breast cancer cells is mediated by several mechanisms including over- expression of the MDR1 gene product, ...
Oncogene (1997) 14, 499 ± 506  1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00

Expression of the mutated p53 tumor suppressor protein and its molecular and biochemical characterization in multidrug resistant MCF-7/Adr human breast cancer cells Besim Ogretmen and Ahmad R Safa Department of Experimental Oncology, Hollings Cancer Center, Medical University of South Carolina, Charleston, SC 29425, USA

Multidrug resistance in MCF-7/Adr human breast cancer cells is mediated by several mechanisms including overexpression of the MDR1 gene product, P-glycoprotein and glutathione-related detoxifying enzymes. Mutations in the p53 tumor suppressor protein have been reported to play a role in the development of resistance to DNA damaging agents in several human cancer cells. In the present study we have assessed the mutational status of the p53 protein and its expression levels, degree of stability and cellular localization to investigate whether it is involved in modulating multidrug resistance in MCF-7/Adr cells compared to sensitive MCF-7 cells. As revealed by immuno¯uorescence microscopy using the anti-p53 mouse monoclonal antibody DO-1, wild-type p53 is sequestered in the cytoplasm of MCF-7 cells, whereas in MCF-7/Adr cells, the protein is localized in the nucleus. The sequencing of full-length p53 cDNA revealed a 21 bp deletion in its one of the four conserved regions within the conformational domain, spanning codons 126-133 at exon ®ve, in MCF-7/Adr cells. Moreover, detection of ThaI polymorphism of codon 72 showed that MCF-7 cells predominantly express wild-type p53 with proline, while mutated p53 in MCF-7/Adr cells contains an arginine residue at codon 72. In addition, we demonstrate that the half-life of p53 in MCF-7 cells is less than 30 min while the mutated protein is more stable; its half-life is about 4 h in MCF-7/Adr cells. Thus, this study demonstrates that the deletion of codons 126-133 in p53 causes increased stability, overexpression and nuclear localization of the protein in multidrug resistant MCF-7/Adr cells, and further suggests that mutated p53 might be involved in the development of multidrug resistance in this cell line. Keywords: deleted p53; multidrug resistance; MCF-7/ Adr

Introduction Wild-type nuclear phosphoprotein p53 plays an essential role in the negative regulation of cell growth (Casey et al., 1991), is involved in regulating apoptosis (Wosikowski et al., 1995; Haldar et al., 1994) and acts as a tumor suppressor protein in a cell cycle-dependent manner (Takahashi et al., 1993). Although the mechanisms of action of p53 in the control of cell growth are not clearly known, there is evidence that it is involved in the transcriptional regulation of certain genes and DNA replication (Maxwell and Roth, 1994). Correspondence: AR Safa Received 22 July 1996; revised 30 September 1996; accepted 1 October 1996

Moreover, studies showing the overexpression of p53 in cells after exposure to DNA-damaging agents suggest that it might be involved in facilitating DNA repair (Wosikowski et al., 1995; Fan et al., 1995). The antiproliferative and anti-transforming actions of p53, however, can be altered by loss of p53 alleles, cytoplasmic sequestration or point mutations (Chen et al., 1990); the mutant p53 can act like oncoproteins that immortalize and transform primary cells in vitro (Maxwell and Roth, 1994). Many studies have reported that alterations of the p53 gene, located on chromosome 17p 13.1 (Negrini et al., 1994), is the most common event identi®ed in at least 50% of all cancers (Landesman et al., 1994; Runneabaum et al., 1991). In most cases, the point mutations, usually single missense mutations that occur in the conserved region of the protein, cause overexpression, a€ect the conformation and prolong the half-life of p53 resulting in nuclear accumulation (Maxwell and Roth, 1994; Moll et al., 1995). Recent studies demonstrate that mutations in p53 can increase resistance to ionizing radiation and other DNA-damaging agents in vivo and in vitro (Lee and Bernstein, 1993; Lowe et al., 1994). In addition, as reported by Lowe et al. (1994), inactivation of p53 by mutations can cause resistance to doxorubicin and gamma radiation treatments in vivo and the mutational status of p53 might be associated with drug resistance in human tumors. The doxorubucin-resistant human breast cancer cell line MCF-7/Adr displays crossresistance to many chemotherapeutic agents such as Vinca alkaloids, anthracyclines, colchicine, taxol, actinomycin D and some other drugs (Fairchild et al., 1987). Multidrug resistance in MCF-7/Adr cells is associated with several di€erent mechanisms including overexpression of P-glycoprotein (P-gp), glutathione Stransferase (GST-Pi) (Fairchild et al., 1990) and glutathione peroxidase (GST-Px) (Wells et al., 1995). In the present study, we have determined the molecular and biochemical characteristics of p53 protein in the MCF-7 human breast cancer cell line and its multidrug resistant variant MCF-7/Adr cells to investigate whether changes in the p53 status are critical in the development of multidrug resistant phenotype in MCF7/Adr cells.

Results Analysis of resistance to doxorubucin in MCF-7 and MCF-7/Adr cells The degree of resistance to doxorubucin in MCF-7 and MCF-7/Adr cells was determined by cell survival assay

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

and the concentrations of doxorubucin that inhibited cell survival by 50% (IC50) were extrapolated from cell survival plots (Figure 1a). As seen in Figure 1a, the IC50 concentrations of doxorubucin for MCF-7 and MCF-7/Adr cells were 0.003 and 20 mM, respectively. These results indicate that MCF-7/Adr cells are approximately 6000-fold more resistant to doxorubucin than MCF-7 cells. It is known that this cell line displays resistance to many MDR-related drugs and that its multidrug resistance is mediated by several di€erent mechanisms (Fairchild et al., 1990) including overexpression of P-gp, as shown by Western blotting using the C219 mouse monoclonal antibody which recognizes P-gp (Figure 1b). Detection of expression levels, stability and cellular localization of p53 protein The protein levels of p53 in MCF-7 and MCF-7/Adr cells were measured by Western blot analysis using the DO-1 mouse monoclonal antibody speci®c for human p53. As seen in Figure 2a, p53 protein is overexpressed (about four to sixfold) in MCF-7/Adr (lane 2) a

b kDa

1

2

220 —



500

p-gp

97.4 — 66 —

compared to MCF-7 cells (lane 1). This overexpression of p53 protein was also evident when a range of protein concentrations (20 ± 80 mg) was used in immunoblotting assays (data not shown). The electrophoretic mobility of p53 on 10 ± 15% SDS-polyacrylamide gels in these cells were di€erent; p53 from MCF7/Adr migrates slightly faster than that of MCF-7 cells (Figure 2a, lanes 2 and 1, respectively). p53 mRNA levels were measured by RT ± PCR (Figure 2b, lanes 2 and 3) and normalized against the rRNA levels (lanes 5 and 6). These results showed that the p53 mRNA levels are almost the same in MCF-7 and MCF-7/Adr cells, suggesting that the overexpression of p53 protein in MCF-7/Adr cells is controlled by post-transcriptional mechanisms such as increased protein stability. Therefore, to determine whether di€erences in the p53 protein levels in these two cell lines are the result of protein stability di€erences, we treated the cells with the protein synthesis inhibitor cycloheximide (CHX) at various time points, detected the p53 levels by Western blotting as described above and compared the results with samples from untreated cells. As Figure 2c illustrates, the half-life of p53 protein in MCF-7 cells is less than 30 min and it is completely degraded after 4 h CHX treatment (lanes 1 ± 5). Similar results were obtained by pulse-chase labelling the MCF-7 cells with [35S]methionine, followed by CHX treatment and p53 immunoprecipitation using DO-1 mouse monoclonal antibody (data not shown). Immunoblotting of MCF-7/Adr cells in the presence of CHX showed that mutated p53 protein is stable up to 12 h and has a half-life of approximately 24 h (Figure 2c, lanes 6 ± 10). However, pulse-chase labelling of these cells using [35S]methionine, without subsequent CHX treatment, followed by immunoprecipitation with DO-1 mouse monoclonal antibody revealed that the half-life of the mutated p53 protein is approximately 4 h and that it is mostly degraded after 12 h (Figure 2d, lanes 1 ± 5). Pulse-chase labelling of MCF-7/Adr cells using [35S]methionine with CHX treatment revealed similar results, showing that the half-life of the mutated p53 protein is about 4 h. Moreover, our immunostaining demonstrated that p53 protein, detected using the primary DO-1 mouse monoclonal antibody, is mainly sequestered in the cytoplasm in MCF-7 cells (Figure 3a). In contrast, as seen in Figure 3b, p53 is predominantly localized in the nucleus in MCF-7/Adr cells; the amount of p53 protein detected in the cytoplasm in this cell line is relatively small. Detection of codon 72 polymorphism and sequencing of the p53 gene

Figure 1 (a) Cell survival analysis of MCF-7 and MCF-7/Adr cells treated with doxorubucin. MCF-7 (open circles) and MCF-7/ Adr (closed circles) cells were grown in 24-well plates in the absence or presence of increasing concentrations of doxorubucin for 96 h at 378C in 5% CO2. Untreated cells were used as controls and each treatment (including controls) was performed in triplicate wells. The IC50 value was the doxorubucin concentration required to decrease cell survival by 50%. Error bars represent standard deviations. (b) Detection of P-gp by Western blotting. Total proteins (100 mg/lane) from MCF-7 (lane 1) and MCF-7/Adr (lane 2) cells were separated on 5 ± 15% SDSpolyacrylamide gels. The mouse monoclonal antibody C219 was used to detect human P-gp which migrates at Mr 185 000. Molecular weight markers are indicated on the left

Previous studies have shown that the wild-type p53 gene has a single nucleotide substitution at codon 72 in several di€erent genomic clones (Ara et al., 1990; Buchman et al., 1988). The single base change at the codon 72, CCC to CGC, causes proline to arginine replacement without any known functional alteration of the wild-type p53 protein (Chen et al., 1990). As seen in Figure 4, this amino acid polymorphism also represents ThaI restriction site polymorphism at codon 72 of the p53 gene, and it can be detected easily by PCR or RT ± PCR followed by ThaI digestion; the ThaI restriction site is present at codon 72 that codes

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

501

a

b kDa

1

2

97.4 —

1 2 3

4

5

6

66 —

— — —

603 —



46 —

— — — —



➝ ➝

310 271 234 194

p53



p53

1353 — 1078 — 872 —

— — — —

rRNA

c kDa 97.4 —

1

2

3

4

5

6

7

8

9

10

66 —

p53

➝ 46 — CHX ➝

0

0.5

1

2

0

4

d

1

4

2

8

12

24

3

4

5

time (hr)

➝ pulse-chase ➝

0

2

4

12

24

p53

time (hr)

Figure 2 (a) The levels of p53 protein in MCF-7 (lane 1) and MCF-7/Adr (lane 2) were detected by Western blot analysis as described in Materials and methods. The protein samples were resolved on 10 ± 15% SDS-polyacrylamide gels and the p53 protein was detected using the mouse monoclonal antibody DO-1. Molecular weight markers are indicated on the left. The autoradiogram of the p53 proteins after Western blotting was obtained after 50 s exposure at room temperature. (b) The mRNA levels of p53 in MCF-7 (lane 2) and MCF-7/Adr (lane 3) were detected by RT ± PCR and normalized against their rRNA levels (lanes 5 and 6, respectively). Lanes 1 and 4 contain HaeIII cut fX174 DNA fragments used as molecular weight markers. (c) Analysis of p53 protein stability by Western blotting after CHX treatments. MCF-7 cells treated for 0.5, 1, 2 and 4 h (lanes 2 ± 5) and MCF-7/Adr cells treated for 4, 6, 10 and 24 h (lanes 7 ± 10) with 80 mg/ml cycloheximide (CHX) were grown at 378C in 5% CO2. The autoradiograms of the p53 proteins after Western blotting were obtained after 4 min (lanes 1 ± 5) or 50 s (lanes 6 ± 10) exposure at room temperature. The proteins were isolated from each sample, separated on 10 ± 15% SDS-polyacrylamide gels and p53 was detected by Western blot analysis as described in Materials and methods. Proteins isolated from untreated MCF-7 (lane 1) and MCF-7/Adr (lane 6) were used as controls. (d) Analysis of the mutated p53 protein stability by pulse-chase labelling with [35S]methionine. MCF-7/Adr cells were labelled in 5 ml of the methionine-free RPMI-1640 medium containing 15 mCi/ml [35S]methionine and, after several washes, they were further incubated in the RPMI-1640 medium containing L-methionine for 0, 2, 4, 12 and 24 h (lanes 1 ± 5, respectively). After that, p53 protein was immunoprecipitated, resolved in 10% SDS ± PAGE and exposed to Kodak-XAR-5 ®lm as described in Materials and methods. Molecular weight markers are indicated on the left. The time of treatments are indicated on the bottom of the autoradiograms. The results shown are representative of two independent experiments

for arginine but not for proline. To detect polymorphism of the p53 gene, genomic DNA or total RNA were isolated from MCF-7 and MCF-7/Adr cells and then used for ampli®cation of DNA or cDNA fragments spanning codon 72 (Figure 4a, b, lanes 2 and 3, respectively). After the ampli®ed DNA

fragments were puri®ed, they were digested with ThaI and analysed on agarose gels by ethidium bromide staining (Figure 4a, b, lanes 5 and 6). Our results demonstrate that two variant forms of the p53 gene are present in these two cell lines; however, p53 with proline residue at codon 72 is predominantly

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

502

Figure 3 Subcellular localization of p53 protein by indirect immuno¯uorescence. MCF-7 (a) and MCF-7/Adr (b) cells grown on cover slips were reacted with the primary mouse monoclonal antibody DO-1 speci®c for human p53, then p53 was localized by immunostaining as described in Materials and methods. The results shown are representative of two independent

a

4 1

2

5

6

3

1353 — 1078 — 872 —

—u

603 — 310 — 271 — 234 —

p53



b

c 4

1

2

5

6

3

1353 — 1078 — 872 — 603 —



310 — 271 — 234 —

p53

—u c

Figure 4 Detection of codon 72 polymorphism in the p53 gene. Genomic DNA or total RNA samples were isolated from MCF-7 and MCF-7/Adr cells and used for the ampli®cation of DNA (a) or cDNA (b) fragments spanning codon 72 of the p53 gene by PCR and RT ± PCR, respectively. After isolation and puri®cation, the ampli®ed DNA fragments from MCF-7 and MCF-7/Adr (lanes 5 and 6) were digested with ThaI and analysed on 2% agarose gels by ethidium bromide staining. The DNA fragments designated as u and c correspond to the uncut and the cut forms, respectively. Undigested DNA fragments from MCF-7 and MCF-7/Adr cells were used as controls (lanes 2 and 3). HaeIII cut fX174 DNA fragments were used as molecular weight markers (lanes 1 and 4)

expressed in MCF-7 cells and in contrast, its counterpart with arginine residue at codon 72 is highly expressed in MCF-7/Adr cells. In order to determine whether mutation of the p53

gene is associated with its increased stability and nuclear localization, the sequencing of full-length p53 cDNA from MCF-7/Adr was performed and the results were compared to the sequence of wild-type p53 in MCF-7 cells. For sequencing full-length p53 cDNA, total RNA was extracted from the cells, and cDNA was synthesized by reverse transcription. Then, p53 speci®c DNA fragments were ampli®ed by PCR and subsequently sequenced. Consistent with other reports (Takahashi et al., 1993), our results demonstrated that MCF-7 has the wild-type p53 gene containing CCC for proline at codon 72 (Figure 5a). However, as shown in Figure 5b, our sequencing analysis of the p53 gene in MCF-7/Adr cells revealed a 21 bp deletion at the upstream end of exon 5, and codon 72 was CGC coding for arginine. In addition, these results showed that sequences of the p53 cDNAs corresponding to the nuclear localization signal (NLS) at the carboxy terminus (Shay et al., 1992) and known phosphorylation sites (serine at codon 9, 15, 37, 315 and 392) (reviewed by Maxwell and Roth, 1994; Ulrich et al., 1992), are the same in MCF-7 and MCF-7/Adr cells (Figure 5a and b, respectively). Discussion Several mechanisms including overexpression of P-gp, GST-Pi and GST-Px contribute to the degree of drug resistance in MCF-7/Adr cells (Wells et al., 1995; Fairchild et al., 1990). Since many studies have reported that alterations in the p53 status by mutations can cause resistance to some DNAdamaging agents including doxorubicin and u.v.irradiation in vivo and in vitro (Lowe et al., 1994), the main goal of this study was to determine whether

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

Figure 5 Comparison of the full-length p53 cDNA sequences in MCF-7 and MCF-7/Adr cells. One mg total RNA was isolated from the cells and used in RT ± PCR to generate cDNA fragments that overlap the full-length p53 sequence. After isolation and puri®cation, the ampli®ed DNA fragments were directly used for sequencing as described in Materials and methods. (a) Our results demonstrated that MCF-7 cells predominantly express wild-type p53 with CCC for proline at codon 72. (b) We detected a 21 bp deletion (marked box) at the upstream end of exon 5, spanning codons 126 ± 133, in the p53 gene in MCF-7/Adr cells; the deleted p53 in this cell line has CGC for arginine at codon 72. Nuclear localization signal (NLS) at the carboxy terminus, the known phosphorylation sites of the protein (serines at codon 9, 15, 37, 315 and 392), and four conserved regions (I-IV) within the conformational domain are indicated for both wild-type and mutated p53 from MCF-7 and MCF-7/Adr in Figure 5a and b. The sequencing of p53 in MCF-7/Adr cells was performed using both strands of the cDNA fragments separately

or not the alterations in the p53 protein might be one of the mechanisms involved in the development of multidrug resistance in MCF-7/Adr cells. In this study, we have characterized the p53 protein expression levels, mutation status, stability and cellular localization in multidrug resistant MCF-7/ Adr and sensitive MCF-7 cells. Consistent with the published data (Takahashi et al., 1993), our sequencing results showed that MCF-7 cells have wild-type p53 with proline at codon 72. In contrast, we detected a 21 bp deletion in one of the four conserved regions within the conformational domain of the p53 gene (Harris et al., 1986; Lamb and Crawford, 1986; Maxwell and Roth, 1994), beginning at exon 5 spanning codons 126-133, in MCF-7/Adr cells. While this work was in progress Wosikowski et al. (1995) also reported this deletion in their resistant MCF-7 clone, independently selected by doxorubucin treatments. In addition, detection of codon 72 ThaI polymorphism by PCR and RT ± PCR demonstrated that although both of the polymorphic p53 alleles are present in the two cell lines, wild-type p53 with proline and mutated p53 with arginine at codon 72 are predominantly expressed in MCF-7 and MCF-7/ Adr cells, respectively; the biological signi®cance of this, however, is not known. Consistent with the previously published data (Reich et al., 1983), both our Western blot analysis of samples in the presence of CHX and pulse-chase labelling with [35S]methionine followed by CHX treatment showed that the half-life of the wild-type

p53 protein in MCF-7 cells is very short, less than 30 min. Similar experiments with CHX treatment of MCF-7/Adr cells followed by Western blotting, however, showed a longer half-life of the mutated protein (approximately 24 h). Pulse-chase labelling using [35S]methionine with or without CHX, on the other hand, revealed that the half-life of the mutated p53 protein was approximately 4 h. The di€erence between the half-life of the mutated p53 protein in MCF-7/Adr cells determined by Western blotting with CHX treatment (approximately 24 h) and pulse-chase labelling using [35S]methionine with or without CHX treatment (about 4 h) suggests that CHX may be e‚uxed by P-gp in MCF-7/Adr cells, and therefore is less e€ective in inhibiting protein synthesis in these cells. Interestingly, CHX has been shown to increase the mRNA levels of MDR1 gene in mouse, rat and human cell lines (Gant et al., 1992). These results are consistent with the previous studies which showed that the mutation in the conformational domain of the p53 protein increased its stability and extended its half-life compared to wild-type p53 (Maxwell and Roth, 1994). In addition, as revealed by immuno¯uorescence microscopy, mutated p53 was detected in the nucleus of MCF-7/Adr cells, while wild-type p53 was localized mainly in the cytoplasm of MCF-7 cells. Although the biochemical basis of increased stability and nuclear localization of the mutated p53 protein in MCF-7/Adr cells is not clearly known, it may be due to alterations in the conformation of the protein caused by the deletion of codons 126 ± 133. In fact, mutations in the conserved region between amino acids 100 ± 300 (known as the conformational domain) of the p53 protein have been reported to a€ect its immunologic con®guration (Maxwell and Roth, 1994). We have observed low levels of the anti-apoptotic protein bcl-2 in MCF-7/Adr cells as compared to the MCF-7 cell line (Ogretmen and Safa, 1996). Down-regulation of bcl-2 protein might also be involved in the nuclear localization of mutated p53 in MCF-7/Adr cells. In fact, previous studies have shown that bcl-2 and c-myc proteins may play roles in the control of cellular and nuclear tracking of p53 since they were detected mainly at the outer nuclear membrane in the cell and their levels and that of nuclear p53 are inversely correlated (Ryan et al., 1994; Krajewski et al., 1993). The nuclear localization signal (NLS) at the carboxy terminus of the p53 (Shay et al., 1992) is reported to play a role in regulating cellular localization of the protein depending on cell type and growth conditions, but no di€erences in the carboxy terminus of the p53 in MCF-7 and MCF-7/Adr cells were detected in this study. The degree of phosphorylation of the p53 protein was reported to cause changes in its conformation (Ullrich et al., 1992). However, Fuchs et al. (1995) have recently shown that alterations on the phosphorylation sites of rat p53 protein did not change its transcription and growth suppression activities. Our sequencing results demonstrated that the known phosphorylation sites of the p53 protein, serines at positions 9, 15, 37, 315 and 392 (Maxwell and Roth, 1994; Ullrich et al., 1992) are the same in both cell lines. In addition, comparison of the 32P-orthophosphate-labeled and 35 S-methionine-labeled p53 protein levels in MCF-7 and MCF-7/Adr cells following immunoprecipitation

503

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

504

revealed that p53 proteins in both cell lines were phosphorylated to the same extent (data not shown), suggesting that nuclear localization of the deleted p53 in MCF-7/Adr cells is independent of its phosphorylation level. The p53 protein is known to modulate the transcription of some genes (Nguyen et al., 1994). However, there are con¯icting reports about the e€ects of wild-type and mutated p53 proteins on the transcriptional activation of the MDR1 gene. Nguyen et al. (1994) and Zastawyn et al. (1995) demonstrated that mutated p53 enhances the MDR1 promoter activity; in contrast, Goldsmith et al. (1995) recently reported that wild-type p53 stimulates MDR CAT activity in a human lung cancer cell line. These con¯icting results could be due to the di€erences in the plasmid constructs, the method of transfections or the cell lines used by di€erent researchers (Goldsmith et al., 1995). The MDR1 promoter does not contain the TATA box sequence and initiation of transcription occurs by some other mechanisms (Nguyen et al., 1994). Several studies have shown that p53 protein binds directly to the TATA sequence to modulate transcription (Segawa et al., 1993). The p53 responsive regions of the MDR1 promoter have been de®ned as sequences spanning from 75 to +25 (Zastawny et al., 1992) and 739 to +50 (Goldsmith et al., 1995) relative to initiation of transcription at +1. The absence of known p53 binding sequences in the MDR1 promoter region indicate that mechanisms by which p53 modulates the activity of this promoter does not require its direct interactions with cis-acting DNA regulatory sequences in this promoter. Therefore, p53 or some other factors induced by p53 may interact with the proteins or transcription factors which are involved in the transcriptional activation of the MDR1 promoter. In light of this, several cellular and viral proteins including MDM2, TATA binding protein, CCAAT and Sp1 binding factors, the viral oncoproteins the SV40 large T antigen, adenovirus E1B and the human papilloma virus E6 (Nguyen et al., 1994) have been shown to bind to p53. On the other hand, how p53 protein regulates the activity of the MDR1 gene remains to be determined. Several studies have recently demonstrated that acquired mutations in p53 are involved in the production of treatment-resistant tumors and associated with decreased sensitivity and increased resistance to DNA damaging agents in vivo and in vitro (Lee and Bernstein, 1993; Lowe et al., 1994). Moreover, in clinical studies mutations in p53 have also been correlated to the development of chemoresistance in lymphocytic leukemia and breast cancer (Fan et al., 1994; O'Connor et al., 1993; Rouby et al., 1993). Although there is no correlation between a particular point mutation in the p53 gene and the development of resistance in these studies, mutations in exon ®ve are among those that were correlated with a high degree of resistance to some drugs and u.v.-irradiation in cancer cells (Lowe et al., 1994; O'Connor et al., 1993; Rouby et al., 1993). In the light of these reports, the present study indicates that the deletion of codons 126 ± 133 in p53 might be involved in the development of the multidrug resistant phenotype and/or contribute to the degree of resistance in MCF-7/Adr breast cancer cells. The mechanism(s) by which the mutated p53 exerts its

e€ects in modulating P-gp-mediated drug resistance or on other drug resistance changes such as overexpression of GST-Pi and GST-Px in this cell line remain to be determined. However, there is evidence that alterations in p53 result in decreased susceptibility to apoptosis induced by DNA damaging agents in cancer cells (Haldar et al., 1994). The high degree of resistance to apoptotic cell death in MCF-7/Adr cells treated by doxorubucin, taxol, vincristine and VP-16 that we have reported in our previous study (Ogretmen and Safa, 1996) might be modulated by the mutated p53 in this cell line. Consequently, this study demonstrates that the 21 bp deletion beginning at exon 5 in the p53 gene alters the expression levels, stability and cellular localization of the protein in multidrug resistant MCF-7/Adr cells and suggests that the mutated p53 might be one of the several factors that are involved in the development of multidrug resistance in this cell line. Identi®cation of these factors and their mechanisms involving in the resistance to chemotherapeutic drugs in human breast cancer cells may help develop better strategies for the treatment of breast carcinomas. Materials and methods Cell lines and culture conditions Human breast cancer cell line MCF-7 and its multidrug resistant derivative MCF-7/Adr were obtained from American Type Culture Collection (Rockville, MD) and Dr Kenneth H Cowan of National Cancer Institute (Bethesda, MD), respectively. Cells were grown in RPMI-1640 medium containing 10% fetal calf serum and 100 ng/ml each of penicillin and streptomycin (Life Technologies; Grand Island, NY) at 378C in 5% CO2. MCF-7/Adr cells were maintained continuously in the presence of 8 mg/ml doxorubucin (Sigma; St. Louis, MO), and the drug was removed from the medium 7 days before each assay. Cell survival assay The degree of resistance in MCF-7 and MCF-7/Adr cells to doxorubucin was analysed by cell survival assay as described previously (Safa et al., 1994). MCF-7 and MCF-7/Adr cells were plated at 15 and 306103 cells/well, respectively, in 1 ml of growth medium in the absence or presence of increasing concentrations of doxorubucin into 24-well plates. Each treatment (including controls) was performed in triplicate wells at 378C in 5% CO2 for 96 h. After washing the cells with 1 ml of the medium twice, the attached viable cells were trypsinized and counted using a Coulter counter (Coulter Electronics; Hialeah, FL). Then, the concentration of doxorubucin that reduced cell numbers by 50% (IC50) was determined from cell survival plots. The ®nal concentration of DMSO (solvent for doxorubucin) in the growth medium was less than 0.1% (v/v) which had no e€ect on cell growth and survival in each experiment. Western blot analysis Total proteins were isolated using approximately 56106 cells lysed in lysing bu€er, containing 10 m M Tris-HCl, pH 7.4, 10 mM NaCl, 1.5 mM MgCl2 and 5 mg/ml each of phenylmethyl sulfonyl ¯uoride (PMSF), leupeptin and aprotinin, by quick freezing at 7808C and thawing at

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

378C. After protein concentrations were determined by Bradford assay using the BioRad Protein Assay kit (BioRad; Richmond, CA), protein samples were separated on 5 ± 15% or 10 ± 15% SDS-polyacrylamide gradient gels (SDS ± PAGE) containing 4.5 M urea (Safa et al., 1994). Protein samples were then transferred to an ImmobilonNC membrane (Millipore Corp.; Bedford, MA) in transfer bu€er (25 mM Tris base, pH 8.3, 200 mM glycine, 0.005% SDS and 20% methanol) using a semi-dry blotter (Pharmacia; Piscataway, NJ) as described by manufacturer. The membranes were then blocked in phosphatebu€ered saline (PBS) containing 0.1% Tween-20 and 10% dry milk at room temperature for 1 h, and Western blot analysis was performed using the primary mouse monoclonal antibodies DO-1 (Santa Cruz Biotechnology; Santa Cruz, CA) speci®c for human p53, and C219 (Safa et al., 1994) which recognizes P-gp, respectively. Proteins were visualized using a peroxidase-conjugated anti-mouse secondary antibody and the ECL detection kit (Amersham; Arlington Heights, IL) as described by manufacturer. Western blotting was performed using 1 mg/ml of primary antibody and 1 : 2500 (v/v) dilution of secondary antibody in 5 ± 15 ml PBS containing 0.1% Tween-20 and 10% dry milk. Equal loading was con®rmed by Coomassie blue staining of SDS ± PAGE strips cut from the gels containing 100 mg protein/lane from both cell lines prior to Western blotting. The levels of p53 proteins were quantitated by densitometry of the autoradiograms after Western blotting. Analysis of p53 protein stability in MCF-7 and MCF-7/Adr cells Seventy to eighty percent con¯uent cells were treated with 80 mg/ml cycloheximide (CHX) (stock solution prepared as 10 mg/ml in dH2O) for 0.5 ± 24 h and then p53 protein levels were measured by Western blotting as described above. It is not known whether MCF-7/Adr cells display resistance to CHX or whether CHX is e‚uxed from these cells. Therefore, in order to determine the half-life of the wild-type and mutated p53 proteins, in MCF-7 and MCF7/Adr cells, respectively, we used an alternative approach by pulse-chase labelling with [35S]methionine followed by immunoprecipitation of the p53 protein. For pulse-chase labelling 70 ± 80% con¯uent MCF-7 or MCF-7/Adr cultures (about 36106 cells) were incubated in 5 ml of the methionine-free RPMI-1640 medium containing 15 mCi/ml [35S]methionine (Amersham, Arlington Heights, IL) for 2 h at 378C in 5% CO2. The medium was then removed and the cells were washed three times with RPMI1640 medium containing L-methionine and 80 mg/ml CHX which was used to prevent further protein synthesis, then further incubated in this medium for 0 ± 24 h as described above. The labelled cells were washed with PBS and dissolved in 500 ml of 0.56RIPA bu€er (50 mM Tris-base, pH 7.5, 1 mM EDTA, 150 mM NaCl, 0.1% SDS, 1% Triton X-100, 1 mM PMSF, 2 mg/ml aprotinin and 1% sodium deoxycholate) and then incubated with the DO-1 mouse monoclonal antibody (1 mg/ml) which recognizes human p53 at 48C for 1 h. The immune complexes were removed using 250 ml Sepharose A at 48C for 1 h. After the precipitates were washed twice with TBS-EDTA-P-SDS (50 mM Tris-base, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS and 1 mM PMSF) and TBS (50 mM Tris-base, pH 7.5 and 150 mM NaCl), they were dissolved in 100 ml sample bu€er and separated on a 10% SDS-polyacrylamide gel. After staining and destaining, the gels were ¯uorographed, dried and exposed to Kodak-XAR-5 ®lm. The p53 levels at each time point were measured by densitometry of X-ray ®lms. The level of p53 protein in controls (0 h) was considered 100%, and the half-life of the protein was determined as

the time point where it degraded to 50% of the controls. Each experiment was performed in two independent trials that gave similar results. Isolation of total RNA and RT ± PCR Total RNA was isolated using a modi®ed phenol-SDS method and the p53 mRNA levels were detected using hotstart RT ± PCR as we recently described (Ogretmen and Safa, 1995). The p53 mRNA levels were normalized against rRNA levels in both cell lines as measured by densitometry of the ethidium bromide stained bands on photographs of agarose gels. The primers and pro®les of the ampli®cation reactions were as follows: p53 A (forward) 5'-TTGCCGTCCCAAGCAATGGAT -3' p53 B (reverse)

5'--AGTCACAGACTTGGCTGTCCCAGA -3'

rRNA (forward) 5'-TTACCAAAAGTGGCCACTA -3' rRNA (reverse)

5'-GAAAGATGGTGAACTATGCC -3'

(958C for 1 min, 588C for 1 min and 728C for 3 min for 30 cycles). The ampli®cation cycles were preceded by a denaturation step (958C for 5 min) and followed by an elongation step (728C for 10 min). After ampli®cation, PCR products were analysed on 2% agarose gels by ethidium bromide staining. All the primers were synthesized by Life Technologies (Grand Island, NY). Cellular localization of p53 protein by indirect immunofluorescence Cellular localization of p53 was determined using indirect immuno¯uorescence (Martin et al., 1987) as follows. MCF7 and MCF-7/Adr cells grown on glass coverslips were washed with PBS and ®xed with 3.6% formaldehyde in 90% ethanol for 15 min on ice. The cells were then reacted with the primary antibody DO-1 (1 mg/ml) speci®c for human p53 for 30 min at room temperature, washed in PBS for 30 min, and then reacted with a ¯uoresceinlabelled secondary antibody (1 : 2500 dilution) for 30 min at room temperature. After staining, the cells were washed in PBS and mounted on slides using a 67% glycerol/33% PBS with 5% n-propyl gallate and then visualized by phase-contrast microscopy. MCF-7 and MCF-7/Adr cells reacted with only the ¯uorescein-labelled secondary antibody (1 : 2500 dilution), omitting the primary antibody, were used as controls in which no signals were detected. In addition, no signals were observed after staining the two cell lines for normal mouse immunoglobulin. Detection of codon 72 polymorphism of the p53 gene Genomic DNA and total RNA samples were isolated from the two cell lines and polymorphism of the p53 gene on codon 72 was detected by PCR and RT ± PCR as described elsewhere (Takahashi, 1991). PCR and RT ± PCR products from MCF-7 and MCF-7/Adr cells were generated using primers p53 A and B and separated on 2% agarose gels as described above. Then 265 bp PCR products from both cell lines were puri®ed from the gels using the Gene Clean kit (Bio 101 Inc.; Vista, CA) and digested by ThaI restriction enzyme (Life Technologies) at 608C for 2 h. Restriction digested PCR products were then run on 2% agarose gels and visualized by ethidium bromide staining. Sequencing of full-length p53 cDNA One mg total RNA from MCF-7 and MCF-7/Adr cells was used in RT ± PCR for generating cDNA as described above, and then the following primers were used for amplifying DNA fragments which overlap full-length p53,

505

Mutated p53 and MDR in human breast cancer cells B Ogretmen and AR Safa

506

consisting of codons 1 ± 148 (5'-ATGGAGGAGCCGCAGTCA-3' and 5'-ATCAACCCACAGCTGCACAGGG3'), codons 118 ± 353 (5'GGGACAGCCAAGTCTGTGACT-3' and 5'-CCTGGGCATCCTTGAGTT-3') and codons 253 ± 393 (5'-ACCATCATCACACTGGAAGAC-TCC-3' and 5'-ATGTCAGTCTGAGTCAGG-3'). The PCR products were separated on 2% agarose gels, puri®ed using the Wizard PCR product isolation kit (Promega; Madison, WI) and then sequenced using both DNA strands separately by dye terminator cycle sequencing (Applied Biosystems; Foster City, CA).

Acknowledgements We would like to thank Dr Mary D McCauley for her editorial assistance. The immuno¯uorescence and cDNA sequencing were performed by The University of Chicago, Cancer Research Center and The University of Pennsylvania, DNA Sequencing Facility, respectively. This work was supported by research grants to ARS from the National Cancer Institute (CA-56078) and the American Cancer Society (DHP-100 and DHP-100A).

References Ara S, Lee PSY, Hansen MF and Saya H. (1990). Nucl. Acids Res., 18, 4961. Buchman VL, Chumakov PM, Ninkina NN, Samarina OP and Georgiev GP. (1988). Gene, 70, 245 ± 252. Casey G, Lo-Hsueh M, Lopez ME, Vogelstein B and Stanbridge EJ. (1991). Oncogene, 6, 1791 ± 1797. Chen P, Chen Y, Bookstein R and Lee W. (1990). Science, 250, 1576 ± 1580. Fairchild CR, Moscow JE, O'Brien EE and Cowan KH. (1990). Mol. Pharmacol., 37, 801 ± 809. Fairchild CR, Ivy SP, Kao-Shan C, Whang-Peng J, Rosen N, Israel M, Melera PW, Cowan KH and Goldsmith ME. (1987). Cancer Res., 47, 5141 ± 5148. Fan S, El-Deiry WS, Bae I, Freeman J, Jondle D, Bhatia K, Fornace AJ, Magrath I, Kohn KW and O'Connor PM. (1994). Cancer Res., 54, 5824 ± 5830. Fan S, Smith ML, Rivet II DJ, Duba D, Zhan Q, Kohn KW, Fornace AJ and O'Connor PM. (1995). Cancer Res., 55, 1649 ± 1654. Fuchs B, O'Connor D, Fallis L, Scheidtmann KH and Lu X. (1995). Oncogene, 10, 789 ± 793. Gant TW, Silverman JA and Thorgeirsson SS. (1992). Nucleic Acids Res., 20, 2841 ± 2846. Goldsmith ME, Gudas JM, Schneider E and Cowan KH. (1995). J. Biol. Chem., 270, 1894 ± 1898. Halder S, Negrini M, Monne M, Sabbioni S and Croce CM. (1994). Cancer Res., 54, 2095 ± 2097. Harris N, Brill E, Shohat O, Prokocimer M, Wolf D, Arai N and Rotter V. (1986). Mol. Cell Biol., 6, 4650 ± 4656. Krajewski S, Tanaka S, Takayama S, Schibler MJ, Fenton W and Reed JC. (1993). Cancer Res., 53, 4701 ± 4714. Lamb P and Crawford L. (1986). Mol. Cell. Biol., 6, 1379 ± 1385. Landesman Y, Bringold F and Kimchi A. (1994). Oncogene, 9, 1241 ± 1245. Lee JM and Bernstein A. (1993). Proc. Natl. Acad. Sci. U.S.A., 90, 5742 ± 5746. Lowe SW, Bodis S, McClatchey A, Remington L, Ruley HE, Fisher DE, Housman DE and Jacks T. (1994). Science, 266, 807 ± 810. Martin TE, Barghusen SC, Leser GP and Spear PG. (1987). J. Cell Biol., 105, 2069 ± 2082. Maxwell SA and Roth JA. (1994). Crit. Rev. Oncogen., 5, 23 ± 57.

Moll UM, LaQuaglia M, Benard J and Riou G. (1995). Proc. Natl. Acad. Sci. U.S.A., 92, 4407 ± 4411. Negrini M, Sabbioni S, Haldar S, Possati L, Castagnoli A, Corrallini A, Barbanti-Brodano G and Croce CM. (1994). Cancer Res., 54, 1818 ± 1824. Nguyen KT, Liu B, Ueda K, Gottesman MM, Pastan I and Chin K. (1994). Oncol. Res., 6, 71 ± 77. O'Connor PM, Jackman J, Jondle D, Bhatia K, Magrath I and Kohn KW. (1993). Cancer Res., 53, 4776 ± 4780. Ogretmen B and Safa AR. (1996). Int. J. Cancer, 67, 1 ± 7. Ogretmen B and Safa AR. (1995). BioTechniques, 19, 374 ± 376. Reich NC, Oren M and Levine AJ. (1983). Mol. Cell. Biol., 3, 2143 ± 2150. Rouby SE, Thomas A, Costin D, Rosenberg CR, Potmesil M, Silber R and Newcomb EW. (1993). Blood, 82, 3452 ± 3459. Runnebaum IB, Nagarajan M, Bowman M, Soto D and Sukumar S. (1991). Proc. Natl. Acad. Sci. U.S.A., 88, 10657 ± 10661. Ryan JJ, Prochownik E, Gottlieb CA, Apel IJ, Merino R, Nunez G and Clarke MF. (1994). Proc. Natl. Acad. Sci. U.S.A., 91, 5878 ± 5882. Safa AR, Agresti M, Bryk D and Tamai I. (1993). Biochemistry, 33, 256 ± 265. Segawa K, Hokuto I, Minowa A, Ohyama K and Takano T. (1993). FEBS, 329, 283 ± 286. Shay JW, Werbin H, Funk WD and Wright WD. (1992). Mutation Res., 277, 163 ± 171. Takahashi K, Sumimoto H, Suzuki K and Ono T. (1993). Mol. Carcinogen., 8, 58 ± 66. Takahashi T. (1991). BioTechniques, 11, 166 ± 170. Ullrich SJ, Anderson CW, Mercer WE and Appella E. (1992). J. Biol. Chem., 267, 15259 ± 15262. Wells WW, Rocque PA, Xu D, Meyer EB, Charamella LJ and Dimitrov NV. (1995). Free Rad. Biol. Med., 18, 699 ± 708. Wosikowski K, Regis JT, Robey RW, Alvarez M, Buters JTM, Gudas JM and Bates SE. (1995). Cell Growth Di€., 6, 1395 ± 1403. Zastawny RL, Salvino R, Chen J, Benchimol S and Ling V. (1995). Oncogene, 8, 1529 ± 1535.