Association of the germline TP53R337H mutation with breast cancer in southern Brazil

The study group comprised 123 women with breast cancer that had been diagnosed and treated at the Hospital of the State University of Campinas (CAISM/UNICAMP), São Paulo, Brazil. The patient's cohort consisted of a subgroup of women who presented a familial history of breast cancer (n = 45) and another subgroup consisted of sporadic breast cancer cases (n = 78). The criteria for the selection of individuals with familial history were: more than three cases of breast cancer and more than one case of ovarian cancer in the family; more than two first-degree relatives with breast cancer; one case of male breast cancer [9, 10]. Thirty of the 45 patients with positive familial history were previously tested for BRCA1/2 mutations [11]. A control group of 223 age-matched female volunteers was selected among university personnel. Family cancer history and blood samples were obtained with informed consent under the guidelines and approval of the Ethical Research Committees of the Faculty of Medical Sciences (FCM) at UNICAMP and The Centro Infantil Boldrini, and the National Committee of Ethics in Research (CONEP). Aditional clinical information, slides and paraffin blocks were obtained for those patients who were shown to carry the R337H mutation.

Genomic DNA was isolated from peripheral blood by using a standard phenol-chloroform extraction method [12]. Paraffin-embedded tumor samples were treated with xylene-100% ethanol (1:1) and digested with proteinase K (14 mg/mL) in 50 mmol/L Tris-HCl (pH 8.0) before phenol-chloroform extraction.

TP53 exon 10 was amplified by using primers and PCR conditions described by Figueiredo et al. [6], with the exception of the forward primer used for amplification of exon 10 DNA from tumor biopsies: 5'-CCATCTTTTAACTCAGGTACTG-3'. TP53 intron 3 and exon 4 were amplified by using primers 5'-CCCTAGCAGAGACCTGTGGGAA-3' (foward) and 5'-AGGCATTGAAGTCTCATGGAA-3'(reverse). TP53 PCR products were directly sequenced by using the BigDye terminator cycle sequencing ready reaction kit (Applied Biosystems, Foster City, CA) in an ABI PRISM 3700 automatic sequencer (Applied Biosystems). The sequences were compared to the human TP53 reference sequence (National Center for Biotechnology Information, accession number U94788). All mutations were confirmed by a second, independent analysis. Tumor DNA was also screened for TP53 alterations in other exons and introns where technically feasible, essentially as previously described [13]. In cases heterozygous for codon 72 and ins16, haplotypes were determined by performing an allele-specific PCR assay as described by Osorio et al. [14]. The MDM2 promoter was amplified with primers described by Bond et al. [15] and PCR products were analyzed by restriction fragment length polymorphism as described by Walsh et al. [16].

Expression of p53 protein was examined by immunohistochemistry with the DO7 monoclonal antibody (Dako) at 1:100 dilution. The slides were incubated with Envision labeled polymer components (Dako, K1491) and developed with 3-3-diaminobenzidine (Sigma, D5637). Tissue from borderline ovarian tumors was included in each assay as a positive control for p53. Slides were interpreted at 40× magnification by a single pathologist. p53 protein expression was considered positive when more than 10% of the nuclei were stained; staining intensity was not considered.

Fisher's exact test (FREQ procedures in SAS v9.1.3, SAS Institute Inc., Cary, NC) was used to compare the frequency of the R337H mutation in the breast cancer and control groups. The R337H mutation rate in the breast cancer group was estimated and the 95% exact CI reported. This rate was compared to the estimated population frequency of the mutation (assumed to be the true prevalence rate for the population) in the region of interest in southern Brazil by using the one-sided exact test for binomial proportion.

The R337H mutation was detected in the germline DNA of three of 123 breast cancer patients in southern Brazil. None of the 223 control subjects carried the R337H mutation or any other mutation in TP53 exon 10. There was a significant association between the presence of the R337H mutation and breast cancer (p = 0.0442; Fisher's exact test, Table 1). The mutation rate of R337H in the cancer-positive group was 2.4%, with an exact Blyth-Still-Casella 95% CI of 0.7% to 6.6%. The overall frequency of the R337H in the region of interest in southern Brazil is assumed to be in the range 0.2%–0.3% [[17], B. Figueiredo, personal communication]. Therefore, the frequency of the R337H mutation was significantly greater among patients with breast cancer than in the general population of the region (p = 0.002 and 0.006 in comparison to a population prevalence rate of 0.2% and 0.3%, respectively; one-sided exact binomial test).

Table 2 summarizes clinical findings for the three carriers of the R337H mutation. All three developed breast cancer at an early age (19, 29, and 44 years). Patient 2 had a second cancer (brain tumor) at the age of 31 years. Patient 1 had no familial history of cancer. Patient 2 had two first-degree relatives with cancer; although the tumor type and age of onset were not known, this family might fulfill LFL-syndrome criteria. Patient 3's family cancer history met Birch's criteria [2] for LFL: a mother with bilateral breast cancer at age 53, three maternal aunts with post-menopausal breast-cancer, a sister with leukemia at age 48, and a grandson with adrenocortical carcinoma at age 18 months.

Tumors with a germline or somatic TP53 mutation usually experience loss of heterozygosity (LOH) in which the wild-type allele is deleted or mutated. Surprisingly, the mutant R337H allele was deleted in all three of the studied breast tumors. No somatic mutations were detected within sequences that could be determined (exons 2, 4, 5, 7, 8, 9, and 10) in any of the breast tumors. By contrast, the brain tumor of patient 2 was homozygous for the R337H TP53 mutation, it also had a second, acquired mutation (T125M), in the DNA binding domain.

R337H-mutant protein accumulates within the nuclei of ACT cells and is detectable by immunohistochemistry, as is other missense mutant p53 expressed in tumors [4]. Breast tumor tissue from patients 1 and 2 was negative for nuclear p53, consistent with selection against the mutant allele. However, breast tumor tissue from patient 3 expressed nuclear p53. Because this tumor had deleted the mutant allele (Table 2), this finding implies that a somatic TP53 mutation outside of exon 10 was acquired during tumorigenesis. However, insufficient tumor tissue hindered the identification of this genetic alteration. The brain tumor tissue from patient 2 also stained positive for nuclear p53, consistent with retention of the allele encoding 337H and acquisition of a second-site mutation (T125M) within the DNA binding domain (See additional file 1: IHC and LOH results of tumor samples from R337H patients).

To investigate whether TP53 polymorphisms were associated with the R337H mutation in the three breast tumor patients, we tested for two common TP53 variants (R72P and ins16) reported to affect breast cancer susceptibility [18‐21]. Patients 1 and 3 were heterozygous in the germline for both the R72P and ins16. Patient 2 was homozygous for R72 and negative for ins16 (Table 3). In the tumor tissue of patients 1 and 3, codon 72 encoded only arginine, indicating a selection against the proline allele.

Due to early age of tumor onset, patient 1 was also tested for BRCA1/2 (data not shown) and a polymophic change was detected at BRCA2 exon 10 (N372H). The other two R337H carriers were not screened for BRCA1/2 mutations.

We also investigated the presence of the SNP309-G within the first intron of the MDM2 gene, which is a key negative regulator of p53 protein [22]. SNP309-G has been reported to accelerate tumor formation in carriers of a TP53 mutation [15]. Patients 1 and 2 were heterozygous for SNP309-G (G/T) and patient 3 was homozygous for the T allele (T/T) (Table 3).