This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Concin, N. Articles by Zeillinger, R. Search for Related Content PubMed PubMed Citation Articles by Concin, N. Articles by Zeillinger, R.

Clinical Relevance of Dominant-Negative p73 Isoforms for Responsiveness to Chemotherapy and Survival in Ovarian Cancer: Evidence for a Crucial p53-p73 Cross-talk In vivo

Authors' Affiliations:¹ Department of Obstetrics and Gynecology, Innsbruck Medical University, Innsbruck, Austria; ² Molecular Oncology Group, Department of Obstetrics and Gynecology, Medical University of Vienna, Vienna, Austria; and ³ Department of Pathology, State University of New York at Stony Brook, Stony Brook, New York

Requests for reprints: Nicole Concin, Department of Obstetrics and Gynecology, Innsbruck Medical University, A-6020 Innsbruck, Austria. Phone: 43-512-504-81433; E-mail: nicole.concin{at}uibk.ac.at.

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Purpose: We aimed to determine the clinical role of the p53 family members p53 and p73 in the responsiveness to platinum-based chemotherapy and survival in ovarian cancer, considering their cross-talk and the p53 polymorphism at codon 72.

Experimental Design: A detailed analysis of p53 and p73 in a series of 122 ovarian cancers was done. We used a functional yeast-based assay to determine the p53 mutational status. Red yeast colonies, indicating mutant p53, were subsequently sequenced to determine the specific p53 alteration. p53 mutations were divided into two groups according to their previous characterization in the literature: those that efficiently inhibit transcriptionally active TAp73 function and those that do not. A p53 polymorphism at codon 72 was determined in corresponding normal tissue or blood of ovarian cancer patients. Isoform-specific p73 expression analysis using real-time reverse transcription-PCR has previously been done in the majority of ovarian cancers included in this study. In a retrospective chart review, responsiveness to chemotherapy was assessed, and survival data with long follow-up times were collected.

Results: Eighty of 122 (65.6%) of ovarian cancers harbored p53 mutations. p53 mutational status was an important determinant of responsiveness to platinum-based chemotherapy in all patients with a residual tumor of <2 cm in diameter after initial surgery (wild-type versus mutant, P = 0.029). In addition, p53 mutational status was a strong prognosticator for recurrence-free and overall survival (P < 0.0001 and P = 0.003, respectively) in univariate analyses. High expression levels of dominant-negative p73 isoforms (Np73 and N'p73) significantly correlated with chemotherapeutic failure (P = 0.048) and with worse recurrence-free and overall survival in patients with p53 mutant cancers (P = 0.048 and P = 0.005, respectively). Eight p53 mutations, present in 19 cases, were found that efficiently inhibit TAp73 (i.e., 175H, 220C, 245S, 245D, 248W, 248Q, 266E, and 273H). Patients with p53 mutations that efficiently inhibit TAp73 function had a significantly shorter overall survival than patients with p53 mutations of unknown effect on TAp73 (P = 0.044). The p53 polymorphism at codon 72 had no influence on responsiveness to chemotherapy or survival.

Conclusion: We provide the first clinical evidence that dominant-negative p73 isoforms contribute to drug resistance in vivo, underscoring the importance of a p53-p73 cross-talk. NH₂-terminally truncated p73 isoforms were of significant clinical effect by providing an additional unfavorable factor for response to platinum-based chemotherapy and survival in p53 mutant ovarian cancers.

There is increasing recognition that intact p53 family signaling is crucial for the cellular response to chemotherapeutic agents. p53 is implicated in mediating cytotoxicity of anticancer agents (3). Two recent studies showed that the p53 family member p73 also functions in the transduction of drug-induced DNA damage to the apoptotic machinery of the cell. Bergamaschi et al. and Irwin et al. showed that transcriptionally active TAp73 is induced by a wide range of clinically used agents, such as cisplatin, anthracyclines, camptothecin, taxanes, and gemcitabine (4, 5). Irwin et al. showed that specific small interfering RNA–dependent down-regulation of TAp73 results in a decrease in drug-induced apoptosis (5). Moreover, in vitro experiments showed that the interaction between p53 family members is important in modulating cytotoxicity of anticancer agents. NH₂-terminally truncated isoforms of p73 (Np73 and N'p73), which have been found to be overexpressed in many cancers (6, 7), act as dominant-negative inhibitors of TAp73 and wild-type p53 and are capable of inhibiting drug-induced apoptosis (4, 7). Furthermore, certain human tumor-derived p53 mutants are able to bind and inhibit TAp73 function (4, 8–11). It has been shown that these p53 mutants, at least partly, confer cellular resistance through abrogation of TAp73 function (5). A single nucleotide polymorphism at codon 72 of p53 has been found to influence the property of p53 mutants in neutralizing p73. p53 mutants harboring an arginine at codon 72 (72R) are able to more efficiently associate with TAp73 and inhibit TAp73-dependent, drug-induced apoptosis compared with p53 mutants with a proline at codon 72 (72P; ref. 4).

Because in vitro data argue for a central role of the p53-p73 interaction in mediating cytotoxicity to anticancer agents, analysis of their clinical significance as a putative biomarker becomes important. To address this, detailed studies correlating clinical responsiveness with p53 status and p73 expression are highly warranted. We therefore determined the p53 mutational status in a series of 122 ovarian cancers and grouped the distinct p53 mutations according to their ability to inhibit TAp73 function. We also determined the p53 polymorphism at codon 72 for these cases. Corresponding isoform-specific p73 expression analysis, discriminating between TAp73 and the dominant-negative p73 isoforms, were done by real-time reverse transcription-PCR and have previously been reported for the majority of cancers included in this study (6). Data were correlated with updated clinical data on chemoresponsiveness and survival. Patients with p53 mutations that efficiently inhibit transcriptionally active TAp73 function had a significantly shorter overall survival than patients with non-TAp73-inhibiting p53 mutations. Moreover, high expression levels of dominant-negative p73 isoforms significantly correlated with failure to respond to platinum-based chemotherapy, worse recurrence-free and overall survival in patients with p53 mutant cancers.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Tissue. Ovarian cancer tissues from 122 patients were collected during surgery at the Departments of Obstetrics and Gynecology of the Innsbruck Medical University, Austria and the Medical University of Vienna, Austria in compliance with and approved by the Institutional Review Board. Tissues were immediately snap-frozen in liquid nitrogen, pulverized in the frozen state, and stored at –80°C until used. Ninety-four of the ovarian cancer tissues used in this study were identical to the cohort analyzed in our previous study (6).

Corresponding normal tissues or blood of the cancer patients were available in 112 of 122 cases. These normal tissues were collected during the same or a subsequent surgical procedure, and tissues were embedded in paraffin. The origin of normal tissues used in this study is given in Table 1.

PCR from matched normal tissues. Primers p53Co72-SE (5'-CAGGTCCAGATGAAGCTC-3') and p53Co72-AS (5'-biotin-GGGACAGAAGATGACAGGG-3'; Genbank accession no. K03199), were used to amplify a 120-bp fragment of genomic p53 DNA. The antisense primer was biotinylated. PCR was carried out in a volume of 25 µL, including 25 ng template, 5 pmol each of sense and antisense primers, and a puReTaq Ready-To-Go PCR Bead (Amersham Biosciences, Buckinghamshire, United Kingdom), which contains 2.5 units of puReTaq DNA polymerase; 10 mmol/l Tris-HCl (pH 9.0 at room temperature); 50 mmol/L KCl; 1.5 mmol/L MgCl₂; 200 µmol/L dATP, dCTP, dGTP, and dTTP; and stabilizers, including bovine serum albumin. The reaction was done on a Perkin-Elmer (Norwalk, CT) GeneAmp PCR system 9700 with 40 cycles at 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds. The reaction was preceded by a primary denaturation step at 94°C for 1 minute.

Detection of polymorphisms by pyrosequencing in matched normal tissues. Polymorphism at codon 72 of p53 was detected using 20 µL PCR products and the PSQ 96 Single Nucleotide Polymorphism Reagent kit and analyzed on a Pyrosequencer PSQ 96 (Uppsala, Sweden) according to the manufacturer's instructions (14). Twenty-five picomoles of the sequencing primer p53Co72-SEQ (5'-GAGGCTGCTCCCC-3') were used to detect the polymorphism.

RNA isolation and cDNA preparation from frozen ovarian cancer tissue. RNA isolation and cDNA preparation have been described in our previous study (6).

Yeast-based functional assay for p53 and sequence analysis. To detect inactivating p53 mutations in the ovarian cancer tissues, the functional yeast-based assay was used as described previously (15, 16). Briefly, functional yeast-based assay is based on transcriptional activity of wild-type versus mutant p53 alleles in a yeast reporter system. p53 mRNA was reverse transcribed, amplified by PCR, and cotransformed into Saccharomyces cerevisiae with a linearized yeast homologous recombination expression vector carrying the 5' and 3' ends of the p53 open reading frame. Wild-type p53, which activates transcription of the yeast ADE2 gene that encodes the phosphoribosyl-aminoimidazole carboxylase, results in white colonies, whereas mutant alleles lack transcriptional activity and result in smaller, red colonies. DNA from at least two red colonies was sequenced to characterize the p53 mutations.

Statistical analysis. Fisher's exact test or ² test was used to examine the relationships of clinical data, p53 mutational status, and p73 isoform expression with responsiveness to chemotherapy. (Fisher's exact test was used when the number of cases in subgroup analyses was <30 in each group. This was the case in all analyses on dominant-negative p73 isoforms and chemoresponsiveness, as well as for all analyses on p53 in different groups of residual tumor and chemoresponsiveness). As previously reported, distribution of p73 isoform expression levels in the cohort was non-Gaussian (6). The dominant-negative p73 isoforms, Np73 and N'p73, produce the same polypeptide; therefore, values were pooled, and cancers were divided by the median cancer level into two approximately same-sized groups [i.e., a high expressing group (above the 50th percentile) and a low expressing group (below the 50th percentile) of Np73 and N'p73 expression]. For correlations with patients' age, a young age group and an old age group were defined above or below the median age of all patients.

Survival probabilities were calculated by the product limit method of Kaplan and Meier. Differences between groups were tested using the log-rank test. The results were analyzed for the end points of recurrence-free and overall survival. Recurrence-free survival was calculated from the day of diagnosis until the date when progressive disease, relapse, or death was reported, whichever occurred first. Patients who had not experienced any unfavorable event were censored at the last date they were known to have been alive. Overall survival was defined as the time between diagnosis and death. Patients who had not died were censored at the last date they were known to have been alive. The Cox proportional hazard model was used for multivariate analysis to assess the independence of different prognostic factors.

Although several subgroup analyses regarding the clinical effect of the examined p53 family members were done, no corrections for multiple comparisons were applied due to the explorative character of these analyses. Ps < 0.05 were considered statistically significant.

Clinical data. In this continuative study, 77% (94 of 122) cases of ovarian cancer patients are identical to the patients analyzed in our previous study (6). For the present study, clinical data of these 94 patients were completed, their survival was updated, and an additional 28 new ovarian cancer patients were recruited. In addition, the patients' responsiveness to chemotherapy was assessed in a retrospective chart review.

Patients were treated at the Department of Obstetrics and Gynecology, Innsbruck Medical University, Austria, between May 1990 and June 2001 and at the Department of Obstetrics and Gynecology, Medical University of Vienna, Austria, between February 1997 and November 1998. All but one patient underwent surgery; this patient received carboplatin chemotherapy alone due to impaired general health. Surgical treatment included total hysterectomy and bilateral salpingo-oophorectomy in all but three patients; 101 of 121 patients underwent additional omentectomy, and partial bowel resection was necessary in 17 patients. Lymph node evaluation involved pelvic lymphadenectomy in 52 of 121 patients.

The median age at the time of diagnosis was 61 years (range, 24-88 years). Patients presented with International Federation of Gynecologists and Obstetricians (FIGO) stage I (n = 30), stage II (n = 7), stage III (n = 67), and stage IV (n = 16) at the time of diagnosis. In two cases, the FIGO stage was unknown. To classify tumor grade, the widely used system based on architectural and nuclear grade and mitotic activity was applied. Ovarian cancers consisted of well (n = 8), moderately (n = 61), and poorly differentiated tumors (n = 46), whereas in seven cases the tumor grade was not available. The histopathology of tumors included adenocarcinomas of the serous (n = 45), mucinous (n = 46), endometrioid (n = 23), and clear cell subtype (n = 4) and four undifferentiated carcinomas. Distant metastases were seen in 17 patients at the time of diagnosis [lung (n = 2), pleura (n = 6), liver (n = 5), skin (n = 1), others (n = 6)]. With the exception of 11 patients with stage I carcinoma, all patients received a platinum-based chemotherapy. This involved four to six courses of carboplatin monotherapy (n = 23) or a combination therapy of cisplatin and cyclophosphamide (n = 38); cisplatin, carboplatin, and cyclophosphamide (n = 8); carboplatin and paclitaxel (n = 41); and cisplatin, paclitaxel, and cyclophosphamide (n = 2). During follow-up, 65 of 119 patients suffered recurrent ovarian cancer (three patients could not be followed for recurrence but were known to be alive), and 60 of 122 patients died. The median time of follow-up was 55 months (range, 3-235 months); the mean time of follow-up was 62 ± 46 months. Data on responsiveness to chemotherapy were available in 104 of 122 patients (11 patients with FIGO stage I disease did not receive chemotherapy; in seven patients, responsiveness to chemotherapy could not be clearly determined in the chart review). Chemotherapeutic failure occurred in 34 of 104 patients.

	Results

Top Abstract Materials and Methods Results Discussion References

 
p53 mutational status in ovarian cancers. Using the highly sensitive functional yeast-based assay, 65.6% (80 of 122) of ovarian cancers were found to harbor p53 mutations, whereas 34.4% (42 of 122) were wild-type p53. In one case, two different p53 mutations were detected. The distribution of p53 missense mutations is illustrated in Fig. 1. Table 2 lists the individual p53 alterations.

View larger version (14K): [in this window] [in a new window]   Fig. 1. Distribution of p53 missense mutations (n = 65) over 34 codons in 122 ovarian cancers. Functional regions loop 1 (L1), loop 2 (L2), loop 3 (L3), and the loop-sheet-helix (LSH) motif, as well as the conserved regions (II-V) of the p53 gene are indicated.

p53 mutations were scattered between exons 4 and 9 with 92.6% (75 of 81) of mutations lying within exons 5 to 8 (codons 130-286). The highest mutational rates were found in exons 7, 6, and 5 with 29.6%, 25.9%, and 22.2%, respectively. p53 missense mutations occurred in 34 different codons, spanning a total of 214 codons; 23 of 65 (35.4%) of mutations clustered in five codons: six in codon 220, five in codon 273, and four cases each in codons 195, 234, and 237.

In the examined ovarian cancers, nucleotide transitions (48 of 71, 67.6%) were found more frequently than transversions (23 of 71, 32.4%). The most common nucleotide changes were G-to-A transitions in 21 of 71 (29.6%) and A-to-G transitions in 13 of 71 (18.3%) cases; 34 of 65 (52.3%) of the observed missense mutations occurred in highly conserved regions of the p53 gene (areas II-V); 17 of 65 (26.2%) were located in the loop 3, 5 of 65 (7.8%) in loop 2, one mutation in loop 1, and one mutation in the loop-sheet-helix motif. Nine patients showed p53 mutations in three of the seven codons that are important for DNA binding (one in codon 241, three in codon 248, and five in codon 273). In three cases, p53 mutations occurred in one of the four amino acids that directly bind the zinc atom (one case in codon 176 and two cases in codon 238).

Grouping of p53 mutations based on their ability to inhibit TAp73. For further clinical correlations, we aimed to not solely consider p53 mutational status (wild type versus mutant) but also to take the interaction between mutant p53 and TAp73 into account. Thus, we grouped the observed p53 mutations according to known functional p73 interactions reported in the literature into p53 mutations that efficiently inhibit transcriptionally active TAp73 function and those with unknown effect on TAp73.

To the best of our knowledge, a total of 80 different p53 mutations have thus far been analyzed and reported for their ability to functionally interfere with p73 (4, 8–11). In our cohort of ovarian cancers, we found eight p53 mutations leading to altered p53 proteins that have been well characterized by several laboratories to strongly bind to p73 protein and efficiently inhibit its transactivation function in vitro. These comprise the p53 mutants 175H, 220C, 245S, 245D, 248W, 248Q, 266E, and 273H (4, 8–11). Moreover, Bergamaschi et al. identified mutants 175H, 220C, 245S, 245D, and 248W to strongly inhibit cisplatin-induced, p73-dependent apoptosis (4). The eight identified p53 mutations occurred in 19 cases of ovarian cancers: 175H was found in three cases, 220C in six cases, 273H in four cases, 248W in two cases, and 248Q, 245S, 245D, and 266E in one case each.

The remaining p53 mutations found in our cohort of ovarian cancers (n = 61) have not been analyzed for their interaction with p73. Thus, we cannot exclude that p53 mutations that are able to neutralize p73 are still contained in this group.

Significance of the p53 mutational status for chemoresponsiveness in ovarian cancer patients debulked to a residual tumor <2 cm. We wished to analyze whether failure to respond to platinum-based chemotherapy in vivo is influenced by the p53 mutational status, the property of p53 mutations to inhibit TAp73 function, the expression levels of dominant-negative p73 isoforms (Np73 and N'p73), and by established clinicopathologic variables (FIGO stage, tumor grade, histologic subtype, the presence of distant metastases at the time of diagnosis, and patients age).

In ovarian cancer, progressive disease during first-line chemotherapy and early recurrence within 6 months after termination of chemotherapy are defined as failure to respond to platinum-based chemotherapy. These patients are referred to have platinum-refractory and platinum-resistant disease, respectively (18). In our cohort, progressive disease during first-line chemotherapy occurred in 16.3% (17 of 104) patients. Early recurrence was seen in another 16.3% (17 of 104) patients. Thus, a total of 32.6% (34 of 104) patients failed to respond to first-line, platinum-based chemotherapy. Of note, all but one case of chemotherapeutic failure occurred in patients with FIGO stage III [40% (25 of 62) of chemotherapeutic failure] or stage IV [50% (8 of 16) of chemotherapeutic failure]. This clearly shows that failure to chemotherapy is mainly a problem of advanced ovarian cancer (FIGO stages and responsiveness to chemotherapy: ² = 13.7, P = 0.008; Table 3).

View larger version (16K): [in this window] [in a new window]   Fig. 2. Factors that determine responsiveness to first-line platinum-based chemotherapy. Chemotherapeutic failure is defined as primary progression during chemotherapy or early recurrence within 6 months after termination of chemotherapy and is compared with a recurrence-free survival of >12 months. A, residual tumor after initial surgery is a crucial determinant of responsiveness to chemotherapy. B, in the subgroup of patients with residual tumors <2 cm, responsiveness to chemotherapy significantly correlates with p53 mutational status. C, expression level of dominant-negative p73 isoforms (Np73 and N'p73) in p53 mutant cancers.

High levels of dominant-negative p73 isoforms in p53 mutant cancers correlate with chemotherapeutic failure in vivo. Furthermore, overexpression of dominant-negative p73 isoforms (Np73 and N'p73) was significantly correlated with responsiveness to chemotherapy in p53 mutant cancers (Fig. 2C; Table 3). Although a substantial percentage (52%, 13 of 25) of p53 mutant cancers that highly express dominant-negative p73 isoforms failed to respond to chemotherapy, only 23.3% (7 of 30) of p53 mutant cancers with low expression levels of dominant-negative p73 isoforms showed chemotherapeutic failure (Fisher's exact test, P = 0.048). Within p53 wild-type cancers, no robust correlation between expression levels of dominant-negative p73 isoforms and response to chemotherapy could be drawn due to small case numbers in these subgroups.

The property of p53 mutations to inhibit TAp73 function did not influence responsiveness to chemotherapy (Table 3). Of note, patients with residual tumors 2 cm showed the highest rate of TAp73-neutralizing p53 mutations (39.3%, 11 of 28) followed by 21.1% (4 of 19) in patients with residual tumors <2 cm and 12% (3 of 25) in patients with residual tumor 0 (² = 5.5, P = 0.065).

p53, p73, and their cross-talk have a significant effect for survival in ovarian cancer. For survival analysis, all FIGO stages were included. In a first step, we determined the relevance of established clinicopathologic variables for recurrence-free and overall survival to validate the reliability of the survival data in our cohort of ovarian cancer patients. As expected, in univariate analysis established variables proved to be significant prognosticators for survival, as summarized in Table 4. The residual tumor after initial surgery was by far the strongest determinant of recurrence-free and overall survival (P < 0.0001 and P < 0.0001, respectively).

View larger version (31K): [in this window] [in a new window]   Fig. 3. Kaplan-Meier survival curves in ovarian cancer patients stratified by (A) p53 mutational status (n = 122), (B) the property of p53 mutations to inhibit TAp73 (n = 79), (C) expression levels of dominant-negative p73 isoforms in the subgroup of p53 mutant cancers (n = 61), and (D) the distinct type of nucleotide change (n = 70).

For multivariate evaluation of p53 family members as independent prognostic markers, we considered the three strongest clinicopathologic prognosticators from the univariate analysis (residual tumor, FIGO stage, and tumor grading for recurrence-free survival; residual tumor, patients age, and tumor grading for overall survival). Of note, the significance of dominant-negative p73 isoforms in p53 mutant cancers for recurrence-free and overall survival remained in multivariate analysis, even in the presence of other known prognostic markers (P = 0.048 and P = 0.042, respectively). This was not the case for the p53 mutational status (P = 0.149 and P = 0.087, respectively), the property of p53 mutations to inhibit TAp73 function, or transitions/transversions.

Correlation between clinicopathologic variables and p53 alterations. p53 mutational status correlated significantly with tumor grade (² = 27.3, P < 0.001). Eighty percent (32 of 40) of p53 wild-type cancers showed moderate differentiation (grade 2), and the majority (57.3%, 43 of 75) of p53 mutant cancers were poorly differentiated (grade 3). A trend was seen for a higher rate of p53 mutations in advanced FIGO stages (² = 6.9, P = 0.08). Regarding the histologic subtype, 43.2% (32 of 74) p53 mutations were mainly found in serous cancers, whereas 34.8% (25 of 74) occurred in mucinous and 23% (17 of 74) in endometrioid cancers. In contrast, the majority of p53 wild-type cancers were of the mucinous subtype (52.5%, 21 of 40; ² = 3.8, P = 0.15). Among the p53 mutations, it was predominantly the missense mutations in highly conserved regions that correlated with the serous histologic subtype (42.4%, 14 of 33), whereas the majority of p53 mutations occurring outside the highly conserved regions were seen in the mucinous subtype (51.7%, 15 of 29; ² = 6.8, P = 0.03). When focusing on p53 mutations in functional areas (loop 1, loop 2, loop 3, and loop-sheet-helix motif), this correlation with the serous histologic subtype became even more pronounced (² = 9.4, P = 0.009). Furthermore, transitions in the p53 gene significantly correlated with older age (² = 5.3, P = 0.022). Although 59.6% (28 of 47) transitions were found in patients older than the median age of 61 years, the majority of transversions (69.6%, 16 of 23) were seen in patients 61 years. Particularly, the C>T transition was found in considerably higher frequency in older patients (77.8%, 7 of 9), whereas the G>T and the T>A transversions were predominantly found in younger patients [66.7% (6 of 9) and 75% (3 of 4), respectively].

The codon 72 polymorphism in p53 mutants does not influence clinical response to chemotherapy in ovarian cancer. In vitro data from several laboratories have shown that p53 mutants with an arginine at position 72 (72R) show worse response to chemotherapeutic agents, including cisplatin, than mutants with the 72 proline (72P) variants (4). Thus, we aimed to determine whether this polymorphism at codon 72 also plays a role in response to chemotherapy in vivo in ovarian cancer patients; 53.6% (60 of 112) cases were homozygous for arginine, 36.6% (41 of 112) were heterozygous, and 9.8% (11 of 112) were homozygous for proline. All p53 mutations detected were analyzed for their codon 72 polymorphism; 67.1% (53 of 79) of mutations were on the arginine allele (72R), 30.4% (24 of 79) were on the proline allele (72P), and in two cases (2.5%, 2 of 79), a histidine was found at codon 72. In p53 mutant cancers arising in codon 72 R/P individuals (n = 31), mutations were detected with similar frequency on the arginine (45.2%, 14 of 31, 72R) and on the proline allele (51.6%, 16 of 31, 72P). Of note, no significant difference was found in the rate of chemotherapeutic failure between 72P p53 mutations (8 of 21, 38.1%) and 72R mutations (19 of 49, 38.8%). The two p53 mutant cancers with a histidine at position 72 showed no progression during chemotherapy or early recurrence. In addition, no significant difference in the rate of 72R and 72P mutations was found between TAp73-neutralizing and non-TAp73-neutralizing p53 mutations. Furthermore, there was no significant difference in recurrence-free or overall survival between patients with mutant 72R compared with mutant 72P.

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
Prompted by a series of detailed in vitro studies on the p53-p73 network and its importance in mediating cytotoxicity of chemotherapeutic drugs, this study examines the clinical relevance of these findings. We herein present the first clinical study combining a complete p53 mutational analysis, including codon 72 polymorphism, quantitative isoform-specific p73 expression, and clinical data on chemoresponsiveness and survival. This allows us to address several crucial clinical questions that await resolution.

First, we wanted to determine whether NH₂-terminally truncated, dominant-negative p73 isoforms contribute to chemotherapeutic failure in vivo. In vitro experiments clearly showed that NH₂-terminally truncated p73 isoforms act as dominant-negative inhibitors of TAp73, thereby inhibiting drug-induced, TAp73-depending apoptosis (4, 5). To date, definitive evidence for a clinical role of NH₂-terminally truncated p73 isoforms in mediating drug resistance is lacking. Of note, our data indicate that dominant-negative p73 isoforms (Np73 and N'p73) are important determinants for failure to platinum-based chemotherapy in p53 mutant ovarian cancers. In view of the in vitro experiments, these data are consistent with the notion that dominant-negative p73 isoforms promote chemotherapeutic failure through inhibition of the alternative TAp73-mediated apoptotic pathway in p53 mutant ovarian cancer. This hypothesis would further implicate a critical in vivo role of TAp73 in the response to chemotherapy. Besides inhibition of TAp73 function, other currently unknown mechanisms might be active through which dominant-negative p73 isoforms promote failure to chemotherapy.

Our study could not prove an influence of dominant-negative p73 isoforms on clinical chemoresponsiveness in p53 wild-type cancers. In vitro data clearly implicate dominant-negative p73 isoforms as potent inhibitors of the proapoptotic TAp73 and wild-type p53 functions (4, 7). This suggests that high expression levels of dominant-negative p73 isoforms in p53 wild-type cancers are unfavorable for responsiveness to chemotherapy. In addition, we previously found dominant-negative p73 isoforms to be expressed at a significantly higher level in p53 wild-type compared with p53 mutant ovarian cancers (6). However, from our clinical study, the possibility arises that dominant-negative p73 isoforms might have a minor in vivo relevance regarding chemoresponsiveness in p53 wild-type cancers. Caution has to be applied when interpreting our findings in p53 wild-type tumors because the small case numbers do not allow us to draw reliable conclusions. Further studies, including a large number of p53 wild-type cancers, are needed to clarify the in vivo role of dominant-negative p73 isoforms in responsiveness to chemotherapy.

In view of the in vitro data, we also determined whether the property of the specific p53 mutation to inhibit TAp73 function is of clinical importance in ovarian cancers. However, in our cohort of ovarian cancers, the capacity of p53 mutations to neutralize TAp73 had no influence on responsiveness to chemotherapy. This finding is in contrast to the study of Bergamaschi et al., who examined a series of 70 inoperable head and neck cancers treated with chemoradiotherapy and reported a significant higher rate of chemoresistance in p53 mutant cancers that neutralize TAp73 compared with cancers with non-TAp73-neutralizing p53 mutations (4). Of note, several p53 mutations found in our cohort of ovarian cancers have been analyzed for their interaction with p73. Thus, we cannot exclude that p73-neutralizing p53 mutations are hidden in this group and bias our results. Nevertheless, from our data, we conclude that in p53 mutant ovarian cancers, the expression of dominant-negative isoforms is of stronger influence on responsiveness to chemotherapy than the property of the specific p53 mutation to inhibit TAp73 function.

Furthermore, we addressed the question whether dominant-negative p73 is a prognostic marker for survival. Here, our data on chemotherapy response are reflected in the survival analysis. With the updated and completed survival data, we now show that high expression levels of dominant-negative p73 isoforms are an unfavorable prognostic marker for recurrence-free and overall survival in p53 mutant cancers. Previously, based on a subset of the current ovarian cancer cohort with shorter follow-up, we found a strong but not significant trend towards worse survival in patients with high expression levels of dominant-negative p73 isoforms (6). Studies in human cancers, discriminating between TAp73 and its NH₂-terminally truncated isoforms with opposing function, remain uncommon and even when analyzed often disregard the functional interactions with p53/p73 (19, 20). In neuroblastomas and lung cancer, expression of dominant-negative p73 was correlated with poor survival, although both studies did not analyze p53 mutational status. Further clinical studies combining a full analysis of p53 and p73 isoforms are therefore necessary to better understand the in vivo relevance of p73 and its interaction with p53. Indeed, the strongest endorsement for a tumor-promoting role of any candidate human oncogene, such as NH₂-terminally truncated p73 isoforms, is a direct effect on survival. Interestingly, although not relevant for responsiveness to chemotherapy and therefore for recurrence-free survival, the property of p53 mutations to inhibit TAp73 function was of significant influence on overall survival.

In p53 mutant cancers, the proapoptotic TAp73 can be inactivated by specific p53 mutants and Np73 proteins. In addition, a remaining wild-type p53 protein from a second allele can also be inactivated by mutant p53 and dominant-negative p73. Thus, the question arises whether Np73 protein is functionally redundant in p53 mutant cancers. Our data on responsiveness to chemotherapy and survival strongly argue that dominant-negative p73 is not functionally redundant in p53 mutant cancers but provides an additional unfavorable factor.

In the examined cohort of ovarian cancers, the p53 mutational status itself (wild type versus mutant) was a strong predictive determinant for response to chemotherapy in satisfactorily debulked patients (<2 cm). Of interest, in patients who were left with residual tumor 2 cm after surgery, p53 mutational status was not crucial for responsiveness to chemotherapy. We hypothesize that these tumors might differ in their inherent biology from the ones that could be debulked satisfactorily. Debulking efficiency might not solely depend on the skills of the surgeon but might be influenced by the intrinsic aggressiveness of the tumor (21). This hypothesis is in line with the finding that the smallest incidence of p53 wild-type tumors (only 7 of 33 cases) and the highest incidence of TAp73-inhibiting p53 mutations were found in patients with residual tumor 2 cm. The independence of chemoresponsiveness from p53 mutational status in patients with residual tumor 0 can probably be explained by the high number of FIGO stage I patients in this group. In these early stages, chemotherapy is less important compared with advanced disease. Although the whole tumor can surgically be removed in patients with FIGO stage I, patients with advanced disease present with peritoneal carcinomatosis where surgery alone is insufficient.

p53 was also a prognostic marker for recurrence-free and overall survival in univariate analyses. Contradictory data in the literature on the relevance of p53 mutational status in ovarian cancer might be explained in part by less sensitive methods used to determine p53 mutations, such as immunohistochemistry or direct sequencing from normal tissue-contaminated samples. This point is underscored by the higher percentage of p53 mutations in our ovarian cancer cohort compared with other studies that report rates between 46% and 56% (22–25). Of note, in addition to a large number of advanced ovarian cancers, our study also included 37 cases of FIGO stage I and II cancers.

An interesting aspect of our data was the crucial influence of the distinct type of a p53 alteration (i.e., transversion or transition) on survival. From a multivariate analysis, we can conclude that this influence is probably due the strong correlation of transitions/transversions with patient age, a well-established prognostic marker in ovarian cancer. In concordance with our data, a recent study of Wang et al. also showed a clear correlation of the distinct type of p53 alteration with patient age in ovarian cancer (26). However, the question remains why transitions rather than transversions preferentially occur in elderly patients. Environmental and genetic factors have been linked to distinct nucleotide changes in the p53 gene, such as tobacco smoke and alcohol in the case of G>T transversions (27) or in breast cancer patients carrying the BRCA1/2 germ line variant to A>T transitions (28). Thus, we speculate that the different distribution of transition versus transversion might reflect variable genetic or environmental factors influencing carcinogenesis in the two age groups.

An arginine/proline polymorphism at codon 72 of p53 has been described to influence the ability of specific p53 missense mutants in the DNA-binding domain to inhibit the TAp73 function (29). In a recent study, Bergamaschi et al. examined a large panel of tumor-derived p53 mutants in vitro and found that the 72R variants of specific p53 mutants more efficiently inhibited cisplatin-induced, p73-dependent apoptosis compared with the corresponding 72P variants (4). In a first clinical study in head and neck tumors, these authors found their in vitro data reflected in vivo, showing a significant better response to chemoradiotherapy in patients with 72P p53 mutants rather than with 72R mutants. However, we could not confirm a clinical relevance regarding responsiveness to chemotherapy or survival for 72P p53 mutants versus 72R mutants in ovarian cancers.

When interpreting our results, the limitations of this study have to be considered. Several subgroup analyses regarding the clinical effect of the examined p53 family members were done, and no corrections for multiple comparisons were applied due to the explorative character of these analyses. Some of our findings arose from subgroup analyses with relatively small case numbers. From analyses in p53 wild-type cancers, no reliable conclusions can be drawn due to small case numbers. Another drawback of this study is that in addition to the p53 mutations that are known to efficiently inhibit TAp73 function, our cohort harbored a large amount of p53 mutations of unknown interaction with TAp73. This can clearly bias our results. Indeed, confirmatory clinical studies are needed.

Taken together, we provide the first evidence that dominant-negative p73 isoforms contribute to drug resistance in vivo. NH₂-terminally truncated p73 isoforms were of significant clinical effect by providing an additional unfavorable factor for response to chemotherapy and survival in p53 mutant ovarian cancers.

	Acknowledgments

 
We thank Andrea Wolf for technical assistance.

	Footnotes

 
Grant support: NIH grant R01 (U.M. Moll) and Fond zur Foerderung Wissenschaftlicher Forschung (N. Concin).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: N. Concin and G. Hofstetter contributed equally to the article.

Received 4/25/05; revised 8/27/05; accepted 9/ 8/05.

	References

Top Abstract Materials and Methods Results Discussion References

 

1. Perez RP, Hamilton TC, Ozols RF, et al. Mechanisms and modulation of resistance to chemotherapy in ovarian cancer. Cancer 1993;15:1571–80.
2. Stuart GE. First-line treatment regimens and the role of consolidation therapy in advanced ovarian cancer. Gynecol Oncol 2003;90:S8–15.
3. Lowe SW, Ruley HE, Jacks T, et al. p53-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 1993;74:957–67.[CrossRef][Medline]
4. Bergamaschi D, Gasco M, Hiller L, et al. p53 polymorphism influences response in cancer chemotherapy via modulation of p73-dependent apoptosis. Cancer Cell 2003;3:387–402.[CrossRef][Medline]
5. Irwin MS, Kondo K, Marin MC, et al. Chemosensitivity linked to p73 function. Cancer Cell 2003;3:403–10.[CrossRef][Medline]
6. Concin N, Becker K, Slade N, et al. Transdominant deltaTAp73 isoforms are frequently up-regulated in ovarian cancer. Role for their evidence as epigenetic p53 inhibitors in vivo. Cancer Res 2004;64:2449–60.[Abstract/Free Full Text]
7. Zaika AI, Slade N, Erster SH, et al. Np73, a dominant-negative inhibitor of wild-type p53 and TAp73 is up-regulated in human tumors. J Exp Med 2002;196:765–80.[Abstract/Free Full Text]
8. Monti P, Campomenosi P, Ciribilli Y, et al. Characterization of the p53 mutants ability to inhibit p73ß transactivation using a yeast-based functional assay. Oncogene 2003;23:5252–60.
9. Gaiddon C, Lokshin M, Ahn J, et al. A subset of tumor-derived mutant forms of p53 down-regulate p63 and p73 through a direct interaction with the p53 core domain. Mol Cell Biol 2001;21:1874–87.[Abstract/Free Full Text]
10. Di Como CJ, Gaiddon C, Prives C. p73 function is inhibited by tumor-derived p53 mutants in mammalian cells. Mol Cell Biol 1999;19:1438–49.[Abstract/Free Full Text]
11. Strano S, Munarriz E, Rossi M, et al. Physical and functional interaction between p53 mutants and different isoforms of p73. J Biol Chem 2000;275:29503–12.[Abstract/Free Full Text]
12. Speiser P, Gharehbaghi-Schnell E, Eder S, et al. A constitutional de novo mutation in exon 8 of the p53 gene in a patient with multiple primary malignancies. Br J Cancer 1996;74:269–73.[Medline]
13. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988;16:1215.[Free Full Text]
14. Nyren P, Karamohamed S, Ronaghi M. Detection of single-base changes using a bioluminometric primer extension assay. Anal Biochem 1997;244:367–73.[CrossRef][Medline]
15. Deissler H, Kafka A, Schuster E, et al. Spectrum of p53 mutations in biopsies from breast cancer patients selected for preoperative chemotherapy analysed by the functional yeast assay to predict therapeutic response. Oncol Rep 2004;11:1281–6.[Medline]
16. Flaman JM, Frebourg T, Moreau V, et al. A simple p53 functional assay for screening cell lines, blood, and tumors. Proc Natl Acad Sci U S A 1995;92:3963–7.[Abstract/Free Full Text]
17. Carbone D, Chiba I, Mitsudomi T. Polymorphism at codon 213 within the p53 gene. Oncogene 1991;6:1691–2.[Medline]
18. Cannistra SA. Cancer of the ovary. NEJM 2005;351:2519–29.
19. Casciano, I, Mazzocco K, Boni L, et al. Expression of Np73 is a molecular marker for adverse outcome in neuroblastoma patients. Cell Death Differ 2002;9:246–51.[CrossRef][Medline]
20. Uramoto H, Sugio K, Oyama T, et al. Expression of deltaNp73 predicts poor prognosis in lung cancer. Clin Cancer Res 2004;10:6905–11.[Abstract/Free Full Text]
21. Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis. JCO 2002;20:1248–59.[Abstract/Free Full Text]
22. Rose SL, Robertson AD, Goodheart MJ, et al. The impact of p53 protein core domain structural alteration on ovarian cancer survival. Clin Cancer Res 2003;9:4139–44.[Abstract/Free Full Text]
23. Reles A, Wen WH, Schmider A, et al. Correlation of p53 mutations with resistance to platinum-based chemotherapy and shortened survival in ovarian cancer. Clin Cancer Res 2001;7:2984–97.[Abstract/Free Full Text]
24. Petty R, Evans A, Duncan I, et al. Drug resistance in ovarian cancer: the role of p53. Pathol Oncol Res 1998;4:97–102.[Medline]
25. Ferrandina G, Fagotti A, Salerno MG, et al. P53 overexpression is associated with cytoreduction and response to chemotherapy in ovarian cancer. Br J Cancer 1999;81:733–40.[CrossRef][Medline]
26. Wang Y, Kringen P, Kristensen GB, et al. Effect of the codon 72 polymorphism (c.215G<C, p.Arg72Pro) in combination with somatic sequence variants in the TP53 gene on survival in patients with advanced ovarian carcinoma. Hum Mutat 2004;24:21–34.[CrossRef][Medline]
27. Pfeifer GP, Denissenko MF, Olivier M, et al. Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking-associated cancers. Oncogene 2002;21:7435–51.[CrossRef][Medline]
28. Olivier M, Hainaut P. TP53 mutation patterns in breast cancers: searching for clues of environmental carcinogenesis. Semin Cancer Biol 2001;11:353–60.[CrossRef][Medline]
29. Marin MC, Jost CA, Brooks LA, et al. A common polymorphism acts as an intragenic modifier of mutant p53 behaviour. Nat Genet 2000;25:47–54.[CrossRef][Medline]

This article has been cited by other articles:

				S. Marchini, M. Marabese, E. Marrazzo, P. Mariani, D. Cattaneo, R. Fossati, A. Compagnoni, R. Fruscio, A. A. Lissoni, and M. Broggini {Delta}Np63 expression is associated with poor survival in ovarian cancer Ann. Onc., March 1, 2008; 19(3): 501 - 507. [Abstract] [Full Text] [PDF]

				R. Malaguarnera, V. Vella, G. Pandini, M. Sanfilippo, V. Pezzino, R. Vigneri, and F. Frasca TAp73{alpha} Increases p53 Tumor Suppressor Activity in Thyroid Cancer Cells via the Inhibition of Mdm2-Mediated Degradation Mol. Cancer Res., January 1, 2008; 6(1): 64 - 77. [Abstract] [Full Text] [PDF]

				A. Tannapfel, K. John, N. Mise, A. Schmidt, S. Buhlmann, S. M. Ibrahim, and B. M. Putzer Autonomous growth and hepatocarcinogenesis in transgenic mice expressing the p53 family inhibitor DNp73 Carcinogenesis, January 1, 2008; 29(1): 211 - 218. [Abstract] [Full Text] [PDF]

				P. Lonning, S Knappskog, V Staalesen, R Chrisanthar, and J. Lillehaug Breast cancer prognostication and prediction in the postgenomic era Ann. Onc., August 1, 2007; 18(8): 1293 - 1306. [Abstract] [Full Text] [PDF]

				R Malaguarnera, V Vella, R Vigneri, and F Frasca p53 family proteins in thyroid cancer Endocr. Relat. Cancer, March 1, 2007; 14(1): 43 - 60. [Abstract] [Full Text] [PDF]

				M. Bagnoli, E. Balladore, E. Luison, P. Alberti, F. Raspagliesi, B. Marcomini, S. Canevari, and D. Mezzanzanica Sensitization of p53-mutated epithelial ovarian cancer to CD95-mediated apoptosis is synergistically induced by cisplatin pretreatment Mol. Cancer Ther., February 1, 2007; 6(2): 762 - 772. [Abstract] [Full Text] [PDF]

				I. R. Watson, A. Blanch, D. C. C. Lin, M. Ohh, and M. S. Irwin Mdm2-mediated NEDD8 Modification of TAp73 Regulates Its Transactivation Function J. Biol. Chem., November 10, 2006; 281(45): 34096 - 34103. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Concin, N. Articles by Zeillinger, R. Search for Related Content PubMed PubMed Citation Articles by Concin, N. Articles by Zeillinger, R.