Background

Breast cancer is the most frequent malignancy among women [1], and the presence of a family history is one of the most fundamental risk factors for the disease [2]. Currently known susceptibility genes including BRCA1, BRCA2, ATM, CHEK2, PALB2, RAD51C and BRIP1 explain less than 25% of familial breast cancer. The rest of the cases could be explained by mutations in mainly moderate and low penetrance cancer susceptibility genes together with environmental factors. Many of the genes already associated with breast cancer susceptibility encode proteins that operate together with BRCA1 and BRCA2 in the DNA damage response pathway (DDR) [3–7]. Other genes with similar functions thus represent good candidates for being new susceptibility genes.

Recent evidence indicates ubiquitin chain formation, recognitionand breakdown at the site of DNA double-strand breaks (DSB) as an essential component of the DDR [8]. RNF8 is a RING-finger ubiquitin ligase (E3), which is recruited to the sites of DNA damage after ATM/ATR-dependent phoshorylation of the H2AX histone variant [9–11]. Together with its ubiquitin-conjugating enzyme (E2) partner UBC13 it mediates K63-linked polyubiquitin conjugation to histones H2A and H2AX. The RNF8/UBC13-dependent histone ubiquitylation is then amplified by the RNF168 E3-ligase acting in concert with UBC13 [12]. Ubiquitynated histones are recognized by RAP80 through its ubiquitin interaction motifs (UIMs), which provide an ubiquitin recognition element to target the BRCA1 E3 ligase, Abraxas, MERIT40, BRCC45 and a K63-ubiquitin specific deubiquitinating enzyme BRCC36 to DSBs. Each of these activities is required for appropriate checkpoint and repair responses to ionizing radiation [13–15]. Depletion of RNF8 or UBC13 in vitro leads to inhibition of the recruitment of 53BP1, BRCA1, RAP80 and Abraxas to DSB sites [9–11, 16]. It has also been demonstrated that the depletion of RNF8 leads to increased ionizing radiation sensitivity and defective G2/M checkpoint [9–11]. In addition, Rnf8 -/- mice display increased genomic instability and higher risk for tumorigenesis, proposing that RNF8 is a novel tumor suppressor [17].

Besides RNF8, the ubiquitin E2 variant (UEV) MMS2 seems to be essential for certain functions of UBC13. MMS2 forms a complex with UBC13 [18], and this heterodimer formation has been demonstrated to be essential for the DNA damage repair function of UBC13 [19]. Suppression of UBC13 or MMS2 has been shown to increase the sensitivity to DNA damaging agents [19], although the exact role of MMS2 in DDR is still unclear [20].

We have previously reported a potentially deleterious germline variant of RAP80 (del81E) that abrogated ubiquitin binding and DNA damage response in breast cancer cases [21]. Additionally, recent findings of an extensive genome-wide linkage consortium study suggested an association between the rare allele of single nucleotide polymorphism (SNP) rs8170 in MERIT40 and an increased propensity for hormone receptor-negative breast cancer, both in the general population and in BRCA1 mutation carriers [22]. Based on the evidence that several players in the ubiquitin-mediated BRCA1-dependent DDR seem to contribute to breast cancer predisposition (summarized in Table 1), we decided to examine the role of RNF8, UBC13 and MMS2 in familial breast cancer by performing a comprehensive mutation screening of these genes in 123 Northern Finnish breast cancer families.

Table 1 Key operators in the BRCA1-dependent ubiquitin -mediated DNA damage recognition pathway and their currently known role in breast cancer predisposition
Full size table

Methods

Study population

Affected index cases of 123 breast cancer families from Northern Finland were screened for germline mutations in the RNF8, UBC13 and MMS2 genes. From the studied families, 77 were classified as high-risk ones, defined as follows: 1) three or more cases of breast and/or ovarian cancer in first or second-degree relatives (median age 49 years, variation 37-80 years), or 2) two cases of breast cancer in first- or second-degree relatives, of which at least one with early disease onset (age ≤ 35 years), bilateral disease or multiple primary tumors. Most of the high-risk families presented three or more cancer cases. The remaining 46 families with moderate disease susceptibility indicated two cases of breast cancer in first- or second-degree relatives. Fourteen of the studied index cases had previously been tested positive for known breast cancer-associated germline mutations in BRCA1 or BRCA2 (eleven) and PALB2 (three). DNA samples from anonymous cancer-free female individuals obtained from Finnish Red-Cross blood donors (N = 104-299, depending on the tested mutation), originating from the same geographical region as the studied cancer cases, were used as controls. All patients had given their informed consent for acquisition of pedigree data and blood specimens for the study of cancer susceptibility gene mutations. Approval to perform the study was obtained from the Ethical Board of the Northern Ostrobothnia Health Care District and the Finnish Ministry of Social Affairs and Health.

DNA extraction and mutation analysis

Genomic DNA was extracted from blood lymphocytes using either the standard phenol-chloroform method or the Puregene D-50K purification kit (Gentra, Minneapolis, MN, USA). Mutation screening of the coding regions and exon-intron boundaries of the RNF8, UBC13 and MMS2 genes was carried out by conformation sensitive gel electrophoresis (CSGE) [23], high resolution melting (HRM) analysis [24], or by direct sequencing. Samples with band shifts in CSGE or deviant melt curves in HRM were reamplified and sequenced with Li-Cor IR2 4200-S DNA Analysis system (Li-Cor, Inc., Lincoln, NE, USA) or with capillary sequencing using ABI3130xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). For Li-Cor the Sequi Therm EXEL II DNA Sequencing Kit-LC (Epicentre Technologies, Madison, WI, USA) and for ABI the Big dye terminator kit v1.1 (Applied Biosystems, Foster City, CA, USA) were used. Chromatograms were interpreted using CodonCode Aligner v. 3.5.4 (Codon Code Corporation, Dedham, MA, USA) and with MEGA4 [25]. Oligonucleotides for CSGE, HRM and sequencing (Table 2) were designed using Primer3 software [26], based on sequence information obtained from public databases (NC_000006.11, NC_000012.11 and NC_000008.10).

Table 2 Primers used for the screening of RNF8, UBC13 and MMS2
Full size table

Statistical and bioinformatical analysis

Carrier frequencies between patients and healthy controls were compared by using Pearson Chi-Square or Fisher's exact test in PASW Statistics (version 18 for Windows, SPSS Inc., Chicago, IL, USA), which was also used for the generation of odds ratios and confidence intervals. All alterations were checked with NNSplice software for potential splicing effects http://www.fruitfly.org/seq_tools/splice.html.

Results

The mutation analysis of RNF8 revealed two exonic, two intronic and three 5'UTR variants. Only one of these variants was novel (not reported the NCBI SNP database, http://www.ncbi.nlm.nih.gov/SNP/). Both of the observed exonic variants of RNF8 were synonymous. In the UBC13 gene, one unreported and one known intronic variant were observed, whereas no sequence alterations were observed in MMS2. All observed alterations in RNF8 and UBC13 were assessed for possible effects on consensus splice sites, but none of them had a predicted effect on splicing. In order to evaluate possible pathogenicity of the observed changes, their frequencies were compared between cases and healthy control individuals. None of the found sequence changes, however, appeared to associate with breast cancer susceptibility (Table 3).

Table 3 Observed alterations in the RNF8,UBC13 and MMS2 genes in Finnish breast cancer families
Full size table

Discussion

RNF8, UBC13 and MMS2 have important roles in the maintenance of genomic integrity and cell-cycle checkpoint control [9–11, 19]. Based on their involvement in DDR and importance in BRCA1-mediated DNA damage recognition it was considered possible that mutations in these genes might contribute to hereditary predisposition to breast cancer.

In the current study, the whole coding region of the RNF8, UBC13 and MMS2 genes was systematically screened for mutations in 123 breast cancer families. No deleterious sequence alterations were observed in any of the genes. Previous studies have suggested that the RNF8 gene could be novel tumor suppressor [17], but it seems that germline mutations predisposing to breast cancer in this gene do not exist or, at least, are very rare. It is of interest that another E3 ligase, RNF168, which acts together with UBC13 to amplify the RNF8-dependent histone ubiquitylation has been shown to be defected in RIDDLE syndrome, which is an immunodeficiency and radiosensitivity disorder. However, it is still unclear whether RIDDLE syndrome is associated with genome instability or increased tumor incidence [12].

Conclusion

The present data suggest that mutations predisposing to breast cancer are either very rare or absent in the coding region of the RNF8, UBC13 and MMS2 genes, which could possibly point to the essentiality of their protein products in the DNA damage response and other functions maintaining genomic integrity. Although a small study like this cannot exclude the possibility of some other rare mutations in RNF8, UBC13 and MMS2 might predispose to breast cancer, based on our findings they unlikely make any sizeable contribution to cancer predisposition. To our knowledge, this is the first study reporting the mutation screening of the RNF8, UBC13 and MMS2 genes in familial breast cancer cases.