BRCA1 and BRCA2 mutations in central and southern Italian patients.

Protein truncation test (PTT) and single-strand conformation polymorphism (SSCP) assay were used to scan the BRCA1 and BRCA2 genes in 136 unrelated Italian breast/ovarian cancer patients. In the sample tested, BRCA1 and BRCA2 equally contributed to site-specific breast cancer patients who reported one to two breast cancer-affected first-/ second-degree relative(s) or who were diagnosed before age 40 years in the absence of a family history of breast/ovarian cancer. BRCA1 and BRCA2 mutations were mostly found in patients with disease diagnosis before and after age 50 years, respectively. Moreover, in cases with familial clustering of site-specific breast cancer, BRCA1 mostly accounted for tumours diagnosed before age 40 years and BRCA2 for tumours diagnosed after age 50 years. The BRCA1 and BRCA2 mutation spectrum was consistent with a lack of significant founder effects in the sample of patients studied.

pancreatic or uterine cancers [synchronous/metachronous or in first-degree relative(s), n = 9]; breast or ovarian cancer with family history of breast-ovarian/ovarian cancer (at least 1 first-/ second-degree relative, n = 10); and ovarian cancer with no breast/ovarian cancer family history (n = 5). Males with breast cancer were recruited regardless of age and family history. BRCA1 exon 11 and BRCA2 exons 10 and 11 were screened by PTT. Coding BRCA1 exons 2, 3, 5-10 and 12-24 and BRCA2 exons 2-9 and 12-27 were screened by SSCP. Primers are listed in Table 1. In 27 cases, analyzed by PTT along the entire BRCA1 coding sequence, BRCA1 SSCP analysis was limited to exons 2, 5, 20 and 24. Mutations were verified by sequence analysis on two independent blood samples. Results: Deleterious germline BRCA1/BRCA2 mutations were detected in 11 out of 136 cases (8%). Only three BRCA2 mutations were novel. One BRCA2 mutation recurred in two unrelated probands. Table 2 shows the mutations and data concerning carriers and their families. Table 3 shows correlations between BRCA1/BRCA2 mutations and sex, age at disease diagnosis and familial clustering of breast/ovarian cancer in the total patient population. Table 4 shows the proportions of BRCA1 and BRCA2 mutations in females with site-specific breast and breast-ovarian/ovarian cancer. Table 5 shows the frequency of BRCA1/BRCA2 mutations in males. BRCA1 and BRCA2 mutations, respectively, accounted for four out of 68 (6%) and one out of 68 (1%) cases diagnosed before age 50 years, and for one out of 68 (1%) and five out of 68 (7%) cases diagnosed after age 50 years. BRCA1 mutations were found in five out of 117 females (4%) and in none of 19 males (0%), and BRCA2 mutations were found in four out of 117 females (3%) and in two out of 19 males (10%). The proportions of BRCA1 and BRCA2 mutations coincided in site-specific female breast cancers (four out of 102; ie 4% each). BRCA1 and BRCA2 equally contributed to female breast cancers, with no familial clustering in those diagnosed before age 40 years (one out of 28; 4% each), and to female breast cancers, all ages, with familial clustering in one to two relatives (three out of 55; ie 5% each). In the latter subset of cases, BRCA1 mostly accounted for tumours diagnosed before age 40 years (two out of eight; 25%), and BRCA2 for tumours diagnosed after age 50 years (three out of 34; 9%). Regardless of family history, the respective contributions of BRCA1 and BRCA2 to site-specific female breast cancers diagnosed before age 40 years were 8% (three out of 36) and 3% (one out of 36). One BRCA1 mutation was detected among the 15 female probands from breast-ovarian/ovarian cancer families (7%). Among male breast cancers, BRCA2 mutations were identified in one out of five (20%) cases with family history and in one out of 14 (7%) apparently sporadic cases. No BRCA1 or BRCA2 mutations were found in female probands with nonfamilial bilateral breast cancer (10 cases) or in those with breast cancer associated with gastrointestinal, pancreatic or uterine cancers, synchronous/metachronous or in first-degree relative(s) (nine cases). These cases were all diagnosed after age 40 years. Discussion: Our results indicate a lack of relevant founder effects for BRCA1-and BRCA2-related disease in the sample of patients studied, which is consistent with other Italian studies and with ethnical and historical data. Overall, the Sense primers used for PTT contain a T7 promoter and a eukaryotic translation initiation sequence: 5′-GGATCCTAATACGACTCACTATAGGGA-GACCACCATG-3′. A, antisense; S, sense.  contribution of BRCA1 and BRCA2 to breast/ovarian cancer in Italian patients appears to be less significant than in patients from communities with founder mutations. The present study is in agreement with direct estimates on other outbred populations, indicating that 7-10% of all female breast cancers that occur in patients aged under 40 years are due to BRCA1/BRCA2. We found that BRCA1 and BRCA2 equally contributed to sitespecific breast cancers who had one/two breast cancer-affected first-/second-degree relative(s) or who were diagnosed within age 40 years in the absence of family history. This is consistent with recent data that indicated that the respective frequencies of BRCA1 and BRCA2 mutations are comparable in early onset breast cancer. Considering the total population of patients analyzed here, however, BRCA1 and BRCA2 mutations were mostly found in cases with disease diagnosis before and after age 50 years, respectively. Moreover, in cases with familial clustering of site-specific Table 4 Proportions of BRCA1 and BRCA2 mutations in 117 female breast/ovarian cancer probands, classified by number of cancer patients in the family, age at disease diagnosis, site-specific breast or breast-ovarian/ovarian cancer Age of proband at disease diagnosis 40 years or less 41-50 years 51 years or more All ages Breast One 4 (1/28) 4 (1/28) 0 (0/9)* 0 (0/9)* 0 (0/10)* 0 (0/10)* 2 (1/47)* 2 (1/47)* Two or three 25 (2/8) (1/15) 0 (0/15) † Including proband; numbers of BRCA1-and BRCA2-positive cases over number of cases in each age subset are given in parentheses. *The 19 breast cancer patients with no familial clustering of breast/ovarian cancer diagnosed above age 40 years included 10 cases with bilateral breast cancer and nine patients with breast cancer associated with gastrointestinal, pancreatic or endometrial cancer (synchronous/metachronous, four cases; in a first-degree relative, five cases). breast cancer, BRCA1 mostly accounted for tumours diagnosed before age 40 years, and BRCA2 for tumours diagnosed after age 50 years. This is in agreement with a trend, which has been observed in other populations, for the proportion of cases with BRCA2 mutations to increase, and the proportion with mutations in BRCA1 to decrease, as the age at cancer onset increases. As in other studies, the frequency of BRCA1/BRCA2 mutations taken together was lower than the estimated frequencies at comparable ages for all susceptibility alleles derived from the Contraceptive and Steroid Hormones (CASH) study. The discrepancy between direct data deriving from BRCA1/BRCA2 mutational analysis and CASH estimates could be due to several factors, including contribution of gene(s) other than BRCA1/BRCA2, differences between populations and relative insensitivity of mutational screening.
Only BRCA1 mutations were found in breast/ovarian and sitespecific ovarian cancer families. BRCA2, but not BRCA1 mutations were found in the male breast cancers. The overall proportion of males with BRCA2 mutations was high when compared with data from other studies on outbred populations, but was low compared with data from populations with founder effects.
The present results should be regarded as an approximation,

Introduction
The proportion of breast cancers that are attributable to autosomal-dominant susceptibility genes is estimated to be approximately 7% in the general population [1][2][3][4]. Germline mutations of the BRCA1 and BRCA2 genes are estimated to contribute to the majority of the breast cancers that have very early disease onset, strong family history and/or association with ovarian cancer [4][5][6]. BRCA1 and BRCA2 also account for a proportion of common breast/ovarian cancer patients that typically do not meet stringent criteria for highly penetrant autosomaldominant cancer predisposition, but rather report one or two disease-affected relatives and/or manifest an early disease onset [4,6,7]. Knowledge of the contribution of BRCA1 and BRCA2 to breast cancer in these patients is still incomplete [4][5][6][7][8]. A better understanding of the frequencies of BRCA1 and BRCA2 mutations in such moderate risk patients is fundamental to our appreciation of the importance of these genes as a cause of disease in the general population. Furthermore, BRCA2 and, to a lesser extent, BRCA1 also appear to be responsible for an important, but still debated proportion of male breast cancers [4,[9][10][11][12][13][14].
We analyzed the entire BRCA1 and BRCA2 coding regions in 136 unrelated probands: 117 female breast/ ovarian cancer patients and 19 male breast cancer patients. This sample, selected from patients attending breast cancer clinics in Rome and Naples, was primarily drawn from moderate-risk families originating from central and southern Italy, a geographic region that is known to be ethnically heterogeneous [15]. We used a screening strategy based on a combination of PTT and SSCP analyses, techniques that are mutually complementary in sensitivity and that may identify more than 80% of mutations [16].

Patients
A total of 136 probands, subclassified as listed below according to selection criteria, were consecutively recruited depending on informed consent from among breast/ovarian cancer patients attending the breast cancer clinics participating in the study in Rome and Naples. The patients originated from the regions of Latium and Abruzzo (central Italy) and Campania and Molise (southern Italy). They included 55 patients with female breast cancer, any age, with breast cancer in one or two first-/second-degree relative(s); 28 patients with female breast cancer diagnosed before age 40 years, who reported no family history of breast/ovarian cancer; 10 patients with female breast/ovarian cancer, any age, with a family history of ovarian/breast-ovarian cancer in at least one first-/second-degree relative; 19 patients with male breast cancer, selected regardless of age and family history; five patients with ovarian cancer, any age, who reported no familial history of breast/ovarian cancer; 10 patients with bilateral breast cancer, selected regardless of age, who reported no family history of breast/ovarian cancer; and nine patients with breast cancer associated with gastrointestinal, pancreatic or uterine cancers, synchronous/metachronous, or in first-degree relative(s), selected regardless of age. Analysis of genomic DNA, RNA and cDNA preparations from peripheral blood lymphocytes were performed following standard procedures. The research protocol was approved by the ethical committee of the University 'Gabriele D'Annunzio'.

BRCA1 and BRCA2 mutational analysis
All patients were analyzed for constitutional mutations throughout the entire BRCA1 and BRCA2 coding regions. BRCA1 exon 11 and BRCA2 exons 10 and 11 were screened by PTT from genomic DNA using the primers listed in Table 1. Coding BRCA1 exons 2, 3, 5-10 and 12-24, and BRCA2 exons 2-9 and 12-27 were screened by SSCP analysis using previously reported primers [11,17]. Primers for BRCA2 exons 14 and 18 and for a 281 bp BRCA2 fragment that encompassed the intron 10/exon 11 boundary are reported in Table 1. In 27 cases, who were analyzed by PTT along the entire BRCA1 coding sequence as described by Friedman et al [17], BRCA1 SSCP analysis was limited to exons 2, 5, 20 and 24, to identify missense mutations that are reportedly frequent in these exons and to allow a better investigation of the 5′ and 3′ ends of the coding sequence. PTT and polymerase chain reaction (PCR)-SSCP were performed as described previously [11,17,18]. When truncated peptides or variant SSCP conformers were identified, genomic DNA was reamplified and directly sequenced with the PCR product sequencing kit (Sequenase Version 2.0, USB-Amersham, Cleveland, OH, USA). Mutations were verified on two independent blood samples.

Results
Deleterious germline BRCA1/BRCA2 mutations were detected in 11 out of 136 cases (8%). Table 2 shows the mutations and data concerning carriers and their families. Table 3 shows correlations between BRCA1/BRCA2 mutations and sex, age at disease diagnosis and familial clustering of breast/ovarian cancer in the total patient population. Table 4 shows the proportions of BRCA1 and BRCA2 mutations in females with site-specific breast and breast-ovarian/ovarian cancer. Table 5 shows the frequency of BRCA1/BRCA2 mutations in males.
The five deleterious BRCA1 mutations (

Full article
and one missense mutation (BRCA1 300TtoG). These mutations were already reported in the literature [18][19][20][21][22] or in the Breast Cancer Information Core electronic database (http://nchgr.nih.gov/Intramural_research/Lab_transfer/Bic/). In addition to the above described deleterious mutations, the neutral BRCA1 coding variants Glu1038Gly and, in homozygosity, Ser1613Gly were found in two patients [20]. A novel C to G transversion, affecting the conserved C/T residues of the consensus sequence for the 3′-splice site of BRCA1 intron 22, was also found in a 60-year-old woman with synchronous breast and gastric cancers, but no family history of cancer. Sequence analysis from cDNA and genomic DNA revealed normal exon 23 and exon 24 transcripts. Allele expression analysis was not feasible, because the patient was homozygous at multiple BRCA1 polymorphisms.
Five deleterious BRCA2 mutations, all localized in exon 11 and including three nonsense and two frameshift mutations, were identified in six patients ( Table 2). BRCA2 4109TtoA, BRCA2 4339CtoT and BRCA2 5117CtoG are novel, whereas BRCA2 5950delTC and BRCA2 6696delTC were previously reported [23,24]. One mutation (BRCA2 6696delTC) recurred in two unrelated probands. The haplotype of the mutation-bearing chromosomes could not be reconstructed, because DNA samples from relatives of the two patients were not available.
BRCA1 and BRCA2, respectively, accounted for four out of 68 (6%) and one out of 68 (1%) patients diagnosed before age 50 years, and for one out of 68 (1%) and five out of 68 (7%) cases diagnosed after age 50 years. BRCA1 mutations were found in five out of 117 females (4%) and in none of 19 males (0%), and BRCA2 mutations were found in four out of 117 females (3%) and in two out of 19 males (10%). BRCA1 and BRCA2 mutations accounted for the same number of cases among patients with family history of breast/ovarian cancer (four out of 66; ie 4% each), whereas BRCA1 and BRCA2 mutations, respectively, accounted for one out of 66 (1%) and for two out of 64 (3%) of the cases without a family history (Table 3).
In site-specific female breast cancers, the proportions of BRCA1 and BRCA2 mutations coincided (four out of 102, ie 4% each). BRCA1 and BRCA2 equally contributed to female breast cancers with no familial clustering diagnosed before age 40 years (one out 28; 4% each) and to female breast cancers, all ages, with familial clustering in one or two relatives (three out of 55; 5% each). In the latter subset of cases, BRCA1 mostly accounted for tumours diagnosed before age 40 years (two out of eight; 25%) and BRCA2 for tumours diagnosed after age 50 years (three out of 34; 9%). Regardless of family history, the respective contributions of BRCA1 and BRCA2 to site-specific female breast cancers diagnosed before age 40 years were 8% (three out of 36) and 3% (one out of 36). One BRCA1 mutation was detected among the 15 female probands from breastovarian/ovarian cancer families (7%). This patient was among the 10 who had a family history of breast/ovarian cancer (one out of 10, 10%; Table 4).
Among male breast cancers, BRCA2 mutations were identified in one out of five (20%) patients with a family history and in one out of 14 (7%) apparently sporadic cases. No BRCA1 or BRCA2 mutations were found in female probands with nonfamilial bilateral breast cancer (10 cases) and with breast cancer associated with gastrointestinal, pancreatic or uterine cancers, synchronous/metachronous or in first-degree relative(s) (nine cases). These cases were all diagnosed after age 40 years (Table 5).

Discussion
We screened the coding sequences of the BRCA1 and BRCA2 genes in 136 breast/ovarian cancer patients, including 102 females with breast cancer who were mostly at moderate risk for mutation-carrier status, 15 females with breast-ovarian/ovarian cancer and 19 males with breast cancer. The sensitivity of the combined PTT/SSCP screening assays is reportedly high [16]. The present results should be regarded as an approximation, because the following types of mutation are predicted to escape detection: mutations in noncoding regions, which are estimated to account for a minimum of 10% of pathogenetic BRCA1 and BRCA2 mutations [5]; missense mutations in functionally important regions within BRCA1 and BRCA2 exons 11; large deletions undetectable by PCR-based assays; and mutations within the first and last 180 nucleotides of the amplicons analyzed by PTT. In spite of these limitations, the present study contributes evidence that is useful for assessing the importance of BRCA1 and BRCA2 mutations in patients who are not in high-risk families from outbred populations.
In contrast to studies on North and East European populations [10,14,[25][26][27], the present results indicate a lack of relevant founder effects for BRCA1-and BRCA2-related disease in the sample of patients analyzed, which is in agreement with other Italian studies [24,[28][29][30] and with ethnical and historical data [15]. The BRCA1 mutations detected in the present study were previously reported in families with high cancer incidence of different ethnic or geographic origin, but not in other Italian surveys [24,[28][29][30]. Interestingly, three of the six BRCA2 mutations were nonsense and three were novel. BRCA2 6696delTC, reported in another series of Italian breast cancer patients [24] but only twice in the Breast Cancer Information Core database, was the only mutation detected more than once. This mutation may represent a candidate frequent mutation in the Italian population. A possible common origin of the mutation detected in two unrelated patients could not be verified, however, because the haplotypes of the mutation-bearing chromosomes could not be reconstructed.
The contributions of BRCA1 and BRCA2 to breast/ ovarian cancer in Italian patients appear to be less significant than in patients from communities with founder mutations [4,10,31,32]. The present results are in agreement with recent direct estimates on other outbred populations, which indicate that 7-10% of all female breast cancers that occur before age 40 years are due to BRCA1/ BRCA2 [4,8,32,33]. Moreover, the 8% BRCA1 mutation rate found in women with site-specific breast cancer diagnosed before age 40 years, regardless of family history, is consistent with an indirect estimate that 5.3% of all female breast cancers among those who are under 40 years old may be due to mutations in BRCA1 [34].
Overall, BRCA1 and BRCA2 equally contributed to sitespecific breast cancer in patients who reported one or two breast cancer-affected first-/second-degree relative(s) or who were diagnosed before age 40 years in the absence of a family history. This is consistent with recent data [4,7] that indicated that the respective frequencies of BRCA1 and BRCA2 mutations are comparable in early onset breast cancer. Considering the total population of patients analyzed here, however, BRCA1 and BRCA2 mutations were mostly found in cases with disease diagnosis before and after age 50 years, respectively. Moreover, in cases with familial clustering of site-specific breast cancer, BRCA1 mostly accounted for tumours diagnosed before age 40 years, and BRCA2 for tumours diagnosed after age 50 years. This is in agreement with a trend, observed in other populations, for the proportion of cases with BRCA2 mutations to increase, and the proportion with mutations in BRCA1 to decrease as the age at cancer onset increases [4,31,35,36]. As in other studies [4], the frequency of BRCA1/BRCA2 mutations taken together was lower than the estimated frequencies at comparable ages for all susceptibility alleles derived from the CASH study [3]. The discrepancy between direct data derived from BRCA1IBRCA2 mutational analysis and CASH estimates could be due to several factors, including contribution(s) of gene(s) other than BRCA1/BRCA2, differences between populations and relative insensitivity of mutational screening.
In the present study, BRCA1 mutations were detected in only one out of 10 cases from breast/ovarian and site-specific ovarian cancer families. This is a low proportion compared with studies that suggested that BRCA1 and BRCA2 are responsible for the large majority of breast/ovarian cancer families, with the greater proportion due to BRCA1 [2][3][4][5]7,19,31]. In this respect, the limits of mutation detection techniques and the small number of breast/ovarian and site-specific ovarian cancer cases tested should be taken into account, together with the fact that most of the probands examined here were from families with only one case of ovarian cancer.
As expected, BRCA2 mutations were detected in male breast cancer patients. BRCA2 mutations were found in 20% of the males reporting familial clustering of breast cancer. In the males with no family history of breast/ovarian cancer, the proportion of carriers of BRCA2 mutations (7%) was comparable to that obtained for BRCA1 and BRCA2 combined in site-specific female breast cancer patients (8%). The overall proportion of cancer-affected males with BRCA2 mutations (10%) was high compared with data from other outbred populations, but was lower than that reported for populations with founder effects [4,[11][12][13][14]37].