Detection of BRCA1 mutations in women with early-onset ovarian cancer by use of the protein truncation test.

Abstract

Inherited mutations in the BRCA1 breast-ovarian cancer gene, isolated in 1994 (/), have been estimated to be responsible for approximately 5% of ovarian cancer diagnosed in women under the age of 50 years (2). Recently, the BRCA2 gene was identified (3), and it appears to confer a much lower risk of ovarian cancers (2,4). BRCA1 and BRCA2 are believed to be tumor suppressor genes, since both alleles appear to be inactivated during neoplastic development (5,6). Despite extensive BRCA1-screening efforts, only five somatic and 11 germline mutations have been identified among 267 sporadic (i.e., nonfamilial) ovarian cancers studied to date (7-11). As BRCA1 is a large gene with widely distributed alterations, mutation screening is labor intensive. Approximately 90% of the reported BRCA1 alterations are either frameshift or nonsense mutations, which result in the premature termination of protein synthesis (12-15). Therefore, we developed the protein truncation test as an efficient strategy to determine the role of BRCA1 mutations in patients with early-onset ovarian cancer. Epithelial ovarian tumor tissue and matched blood lymphocytes were obtained, with institutional review boardapproved, written informed consent, from 16 patients treated at the Division of Gynecologic Oncology at Duke University Medical Center. The mean age of the patients at disease onset was 48 years. Exon 11 (61% of coding sequence) was screened by the protein truncation test using primers and protocols that are available on-line in the Breast Cancer Information Core database (http://www. nchgr.nih.gov/intramural_research/lab_transfer/BIC/) (16). In addition, the entire BRCA1 -coding region of these tumor DNAs was analyzed by single-strand conformation analysis, as previously described (7,16). Analysis using the protein truncation test revealed alterations in BRCA1 protein size in patients EOO473 and EOO1906 (Fig. 1, A). Direct DNA sequencing identified an identical, previously unreported, single base-pair deletion, 2575delC, in exon 11, causing a frameshift and subsequent premature stop codon (Table 1). This deletion was present in the germline of both patients (data not shown) who were apparently unrelated. Patient EOO473 had no known family history of cancer, but patient EOO1906 reported a strong family history (Table 1). In patient EOO23, single-strand conformation analysis of tumor DNA revealed a mobility shift that was shown by DNA sequencing to result from a T—>G transversion 11 base pairs prior to the 3' splice site of intron 5 (Table 1). This previously described alteration (14,15) activates a cryptic splice site that includes 59 nucleotides of intron 5 in the BRCA1 messenger RNA (mRNA) of this patient. This mutation was also detectable by the protein truncation test using tumor complementary DNA (Fig. 1, B). Direct sequencing of lymphocyte DNA indicated that this was a germline alteration (data not shown), and this patient reported a sister with dual primary breast and ovarian cancers (Table 1). In this study, we have identified germline BRCA1 mutations in three of 16 women with early-onset ovarian cancer. Despite extensive searches by several groups (7-11), somatic mutations in BRCA1 appear to be quite rare. It is interesting that the sporadic ovarian cancer patients with germline mutations described previously (10,11) and in this brief communication (average age at disease onset = 45 years) developed cancer approximately 18 years earlier than patients with somatic mutations (average age at disease onset = 63 years) (8,9). This significant age difference at disease diagnosis (Student's t test, P<.003) is consistent with the need for two somatic events to inactivate BRCA1. Although mutation detection in BRCA1 remains technically challenging, the results of the current study and those obtained by other groups (17,18) suggest that the protein truncation test is an efficient screening tool. This test does have certain shortcomings, however. In this test, missense alterations are undetectable, and sensitivity is limited for mutations that yield either short or very large products. Adapting gel concentrations and electrophoresis times as well as substituting [H]leucine for [S]methionine may help to circumvent these problems (17,19). Screening the entire coding region of the gene by the protein truncation test requires the availability of cellular RNA, whereas only exon 11 can be screened using genomic DNA. Finally, in specific cases, the instability of mRNA-containing, premature stop codons may also reduce detection by this technique (20). In summary, although somatic BRCA1 mutations are responsible for only a