Cancer-specific methylation in the BRCA1 promoter in sporadic breast tumours

Abstract

To the EditorSilencing of the BRCA1 gene via promoter hyperme-thylation has been proposed as one mechanism for itsinactivation and found to occur in 10–31% of all sporadicbreast tumours [1, 2]. The impact of promoter hyperme-thylation on the BRCA1 gene is similar to that of a germ-line BRCA1 mutation in terms of gene expression patterns[3, 4]. Breast cancer tumours, with extensive hyperme-thylation in the BRCA1 promoter, have consistently beencorrelated with reduced BRCA1 expression [2, 5].Previous studies have shown a region of 267 bp (-224to ?43; GenBank for BRCA1, U37574) within the BRCA1promoter to be functionally important for its optimaltranscriptional regulation. In particular, deletion of a 222bp (-220 to ?20) fragment within this region, termed thepositive regulatory region (PRR), has been shown to resultin complete loss of the BRCA1 promoter activity [6, 7].Alterations by DNA methylation are thought to be animportant event in tumour initiation; hence, the identifi-cation of tumour-specific methylation patterns in BRCA1promoter may serve as targets for breast cancer detection.The aim of this study was to investigate the presence ofcancer-specific methylation within the PRR of BRCA1 insporadic breast tumours. Histologically proven archivalmatched breast cancer tissues, as well as normal breasttissues from the affected breast, from 23 sporadic breastcancer cases were selected, and their use had the approvalof the institutional ethical committee, National UniversitySingapore-Institutional Review Board (NUS-IRB). Mat-ched tumour and corresponding normal tissues of the samepatient were used to compensate for any age-relatedmethylation [8, 9].BRCA1 promoter hypermethylation was detected in34.8% (8/23) of Asian sporadic breast tumours using MS-PCR-based method (Fig. 1a, b). MS-PCR was carried outat least twice for all samples to confirm the percentage ofcases with BRCA1 promoter hypermethylation. Differencein BRCA1 methylation levels between each matchedtumour and normal tissue was initially determined by cal-culating the mean, standard deviation and 95% CI of the 23matched samples. Tumour samples with difference inBRCA1 methylation levels greater than the upper value of95% CI were considered to be hypermethylated. In addi-tion, methylation in the BRCA1 promoter was alsoobserved in normal breast tissues from breast cancerpatients, an observation similarly reported in a previousstudy [10].Our data also showed that breast tumours with BRCA1hypermethylation exhibited a significant loss of BRCA1transcripts (Fig. 1c). BRCA1 hypermethylated tumourswere then screened for BRCA1 mutation where the entirecoding sequence, including splice junctions and neigh-bouring intronic regions of BRCA1 were analysed usingpreviously described methods [11]. This was carried out toexclude somatic mutation as the cause for loss of BRCA1expression in these samples. No mutations were found thussuggesting BRCA1 transcriptional silencing by promoterhypermethylation in these tumours. MS-PCR products