Introduction
The
TP53 tumor suppressor pathway is well-known to be crucial for maintaining genomic integrity and preventing cells from undergoing oncogenic transformation [
1,
2]. MDM2 plays a key role in regulating the
TP53 pathway by binding directly to the p53 protein, inhibiting its activity and mediating degradation via the ubiquitination system [
3]. p53 also positively regulates MDM2 expression, thereby creating a negative feedback loop [
3]. Overexpression of MDM2 is observed both in epithelial cells of transgenic mice with induced mammary carcinomas [
4] and in multiple human tumors, including breast cancer [
5‐
7].
The
TP53 codon 72 Arg>Pro (C
GC to C
CC) polymorphism of exon 4 [
8] (National Center for Biotechnology Information single-nucleotide polymorphism (SNP) identification number rs1042522) has been suggested to play a role in several different cancer types. These two variant protein forms may behave differently, as the Arg/Arg genotype has been reported to induce apoptosis more effectively than the Pro/Pro genotype [
9,
10], which may be due to enhanced mitochondrial localization of p53 protein in cells with the Arg/Arg genotype [
9]. In contrast, the Pro/Pro genotype appears to induce a higher level of G1 arrest than the Arg/Arg genotype [
11,
12]. Patients with the Pro/Pro genotype of
TP53 in breast cancers have poorer survival than those with other genotypes [
13]. Furthermore, retention of the Arg allele of
TP53 in tumor tissue of Arg/Pro heterozygous breast cancer patients has been associated with reduced disease-free and overall survival [
14]. Taiwanese lung cancer patients and Israeli colorectal cancer patients with the Pro/Pro genotype of
TP53 also showed poorer survival [
15,
16]. A recent study showed that breast cancer patients with the Pro/Pro genotype were less sensitive to chemotherapy than those with Arg/Arg or Arg/Pro genotypes [
17]. Similar results were reported in head and neck carcinoma [
11]. On the other hand, estrogen receptor (ER) positive patients possessing the Pro allele had better distant recurrence-free survival when randomized to tamoxifen compared to those who did not receive tamoxifen, while homozygous Arg/Arg patients did not [
18]. After the initial report of a statistically significantly increased risk of breast cancer in women homozygous for the Pro allele [
19], numerous studies examined a possible role of this
TP53 polymorphism in breast cancer risk. Meta-analysis of nine studies has recently shown that this
TP53 polymorphism is not associated with breast cancer risk [
20].
A SNP in the promoter of the
MDM2 gene, referred to as SNP309 (a T→G change) (rs2279744), has been implicated in earlier age of onset of Li-Fraumeni syndrome and sporadic cancers [
21]. The
MDM2 SNP309 G/G homozygous genotype elevates MDM2 protein expression [
21]. A recent study showed that cells that harbor this genotype had a compromised
TP53 response pathway and formed transcriptionally inactive p53-MDM2 complexes in response to stress [
22]. The G/G genotype was also associated with increased incidence of esophageal squamous cell carcinoma [
23]. Colorectal cancer patients who had both the SNP309 G allele and wild-type
TP53 were diagnosed at a younger age than those with the T/T genotype and wild-type
TP53 [
24]. On the other hand, no association was found between SNP309 status and breast cancer incidence [
25‐
27]. However, a recent study showed that, in women whose breast cancers expressed high levels of ER, those having the
MDM2 SNP309 G/G genotype showed earlier age of onset than those with the T/T genotype [
28]. Another study showed that among patients with the T/T genotype, mutant status of
TP53 and aberrant p53 protein expression were associated with poor survival, suggesting an interaction between
MDM2 SNP309 and tumor
TP53 status [
25].
Here, we have examined germ-line DNA samples from 557 consecutive primary breast cancer patients to investigate the association between these SNPs and breast cancer prognosis.
Discussion
We demonstrate here that SNPs of TP53 codon 72 may have an important role in breast cancer. We show that the Pro/Pro genotype at codon 72 of TP53 was associated with poor DFS, particularly in patients who received adjuvant chemotherapy while MDM2 SNP309 was not associated with prognosis.
TP53 codon 72 encodes two distinct functional allelic forms: arginine (Arg) or proline (Pro) [
8]. Polymorphism at this codon has been suggested to modulate
TP53-dependent apoptosis and modify sensitivity to chemotherapeutic agents [
9,
11,
12,
17]. Recent studies reported that the Pro/Pro genotype of
TP53 codon 72 was associated with poorer survival in Finnish breast cancer patients [
13], and suggested that
MDM2 SNP309 status is associated with p53 protein function [
21,
22]. These findings inspired us to investigate the association of breast cancer prognosis with SNPs of
TP53 codon 72 and
MDM2 SNP309. We used the TaqMan SNP genotyping assay, which is amenable to high-throughput genotyping and avoids many problems of traditional genotyping assays, such as PCR-restriction fragment length polymorphism [
31]. Although the TaqMan assay is convenient and reliable, it is less accurate than direct sequencing. Therefore, we analyzed our data carefully. Our results show that the Pro/Pro genotype of
TP53 was associated with poorer DFS in Japanese breast cancers patients, thus supporting the Finnish study mentioned above [
13]. However, the authors of that study did not address the question of why the Pro/Pro genotype of
TP53 adversely affected the prognosis of breast cancer patients. Analysis of our entire set of 557 patients showed a relationship between the
TP53 Pro/Pro genotype and DFS, but with a
P value (0.047) at the borderline of significance, leaving doubt as to whether the Pro/Pro genotype is an independent risk factor for poor DFS.
We therefore attempted to identify subgroups in which the effect of the
TP53 codon 72 was more significant. A recent study showed that breast cancer patients with the Pro/Pro genotype demonstrated less sensitivity to a neoadjuvant chemotherapy regimen that included 5-FU, cyclophosphamide, and the anthracycline derivative pirarubicin [
17]. This suggested that
TP53 codon 72 polymorphism might be a strong predictive marker for chemotherapy response in breast cancer patients. Our data show that the Pro/Pro genotype was associated with poor DFS in node-positive but not in node-negative patients, and 76% of node-positive patients in our series had received adjuvant chemotherapy while only 26% of node-negative patients had.
Although lymph node status is an important factor in classical staging of breast cancer based on histological/anatomical markers, there is little evidence that breast cancers with lymph node involvement are biologically different from those without it [
32‐
34]. We therefore considered that the apparent effect of node status on the relationship between
TP53 genotype and node status was due to a correlation between codon 72 polymorphism and effect of adjuvant chemotherapy, since 76% of node-positive patients in our series had received such therapy while only 26% of node-negative patients had. We thus analyzed breast cancer survival with respect to this
TP53 polymorphism and the type of adjuvant therapy administered. Our data show that among patients who had received adjuvant chemotherapy, those with the Pro/Pro genotype of
TP53 exhibited poorer DFS. Our finding is also consistent with a previous study of head and neck carcinoma showing that among patients who had received chemo-radiotherapy, those with the Pro/Pro genotype of
TP53 showed poorer survival compared to patients with other genotypes [
11].
Moreover, in ovarian cancer patients who received adjuvant cisplatinum/paclitaxel chemotherapy, the
TP53 Pro allele was associated with a poorer prognosis [
35]. An
in vitro study showed that anticancer agents such as doxorubicin, 5-FU, and cisplatin induced a higher level of apoptosis in human H1299 cells expressing the Arg/Arg genotype of
TP53 codon 72 than in those expressing the Pro/Pro genotype [
11]. In addition, in a colony-survival assay, doxorubicin and cisplatin were more cytotoxic to cells expressing the Arg variant than to those expressing the Pro variant [
11]. Our data are consistent with this
in vitro study. The results among patients receiving both chemotherapy and hormonal therapy were similar to those among patients who had received adjuvant chemotherapy alone (data not shown), but no correlation was found between
TP53 polymorphism and DFS among the patients receiving adjuvant hormonal therapy alone. Positive correlation between
TP53 polymorphism and DFS was observed among patients receiving any chemotherapeutic agents. Since these compounds are cytotoxic while hormonal therapeutic agents are cytostatic, our results may reflect differential effects of
TP53 polymorphism on different apoptotic pathways. Recently, ER positive patients possessing the Pro allele of
TP53 codon 72 have been reported to show better distant recurrence-free survival when randomized to tamoxifen compared to those not receiving tamoxifen [
18]. That report suggested that the Pro allele of
TP53 codon 72 might be a predictor of tamoxifen response [
18], although the authors did not present a Kaplan-Meier analysis of DFS by
TP53 codon 72 genotype among patients receiving tamoxifen. We did not find any correlation between the genotype of
TP53 codon 72 and prognosis in patients receiving tamoxifen alone (data not shown).
Neither did we find any statistically significant association between
MDM2 SNP309 and survival in any subgroups or in the total population, although we did observe a non-significant tendency toward better DFS with the T/T genotype in patients who received adjuvant tamoxifen with or without LH-RH analog. Bond
et al. [
28] proposed a model in which an estrogen-signaling pathway allows the G-allele of
MDM2 SNP309 to accelerate breast cancer formation. This allele might also alter the efficacy of tamoxifen, although the mechanism is unclear. A recent report showed that the G/G genotype of
MDM2 SNP309 was associated with poor prognosis, as well as
TP53 mutations and p53 protein immunopositivity, in gastric carcinoma [
36].
TP53 alteration is correlated with shortened survival of patients with gastric carcinoma [
37,
38]. This suggests that the G/G genotype of
MDM2 SNP309 might, therefore, be predictive of poor survival in gastric carcinoma patients. Many reports demonstrate that
TP53 mutations confer a worse overall and disease-free survival in breast cancer patients [
39,
40], while the prognostic value of p53 protein accumulation has not been consistently demonstrated [
41,
42]. Although
TP53 mutation status was not available in our series, we analyzed the correlation between
MDM2 SNP309 and p53-immunopositivity, but did not find any association between them.
In previous reports, the frequency of the Pro/Pro genotype of
TP53 codon 72 was about 7% to 30% [
13,
15,
17,
18,
23,
43,
44], although 39.1% of patients with signet ring cell gastric carcinomas and 37.7% of female patients with lung carcinomas had the Pro/Pro genotype in subgroup analyses [
15,
44]. The reported frequency of the G/G genotype of
MDM2 SNP309 is about 10% to 30% [
23‐
28,
36,
43]. We found that 38% of patients had the Pro/Pro genotype of
TP53 codon 72 and 33% had the G/G genotype of
MDM2 SNP309. Therefore, the allelic discrimination data from TaqMan SNP genotyping assays were confirmed by direct sequencing of representative PCR products.
One of the goals of translational cancer research is to identify molecular predictors of chemotherapy treatment. Molecular genetic determinants of treatment outcome are important to facilitate identification of patients most likely to benefit from chemotherapy. The present results suggest that the SNP of TP53 codon 72 is a potentially useful marker. However, this is a retrospective pilot study. Further work will be needed to verify the effect of this SNP in breast cancer.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
TT conceived the design, carried out genotyping assays and drafted the manuscript. ZZ carried out genotyping assays and DNA sequencing. MN carried out genotyping assays. MH performed the statistical analysis. NK collected the patients' data. H Iwase and ST provided study material. H Iwata participated in the study design. HY participated in the study design. YF supported the study financially. All authors read and approved the final manuscript.