Abstract
Triple negative breast cancer is typically an aggressive and difficult to treat subtype. It is often associated with loss of function of the BRCA1 gene, either through mutation, loss of heterozygosity or methylation. This study aimed to measure methylation of the BRCA1 gene promoter at individual CpG sites in blood, tumour and normal breast tissue, to assess whether levels were correlated between different tissues, and with triple negative receptor status, histopathological scoring for BRCA-like features and BRCA1 protein expression. Blood DNA methylation levels were significantly correlated with tumour methylation at 9 of 11 CpG sites examined (p<0.0007). The levels of tumour DNA methylation were significantly higher in triple negative tumours, and in tumours with high BRCA-like histopathological scores (10 of 11 CpG sites; p<0.01 and p<0.007 respectively). Similar results were observed in blood DNA (6 of 11 CpG sites; p<0.03 and 7 of 11 CpG sites; p<0.02 respectively). This study provides insight into the pattern of CpG methylation across the BRCA1 promoter, and supports previous studies suggesting that tumours with BRCA1 promoter methylation have similar features to those with BRCA1 mutations, and therefore may be suitable for the same targeted therapies.
Introduction
The triple negative (TN) subtype of breast cancer accounts for 10–17% of all breast carcinomas [1–4]. Triple negative tumours are more likely to be of higher grade, to present with nodal or distant metastases, and there is a relative lack of effective therapies compared to other cancer subtypes, which all contribute to poor disease-free and overall survival [5]. By definition, these tumours are oestrogen receptor (ER) and progesterone receptor (PR) negative and negative for human epidermal growth factor receptor (HER2). Triple negative tumours are known to be a heterogeneous group with a significant proportion displaying the basal-like phenotype; with overexpression of cytokeratin 5/6(CK) and epidermal growth factor receptor (EGFR) proteins. However, all other molecular subtypes of breast cancer are also present in TN cohorts [6, 7]. Recent comprehensive RNA and DNA profiling analyses have identified at least four distinct subtypes of triple negative breast cancers that may have specific therapeutic targets based on their molecular signatures [8, 9].
The Breast Cancer susceptibility gene 1 (BRCA1) is the most commonly mutated gene in familial breast cancer cases and is strongly associated with both the TN subtype and basal-like breast tumours [10, 11]. Over 50% of BRCA1 mutation-associated tumours are TN [12], however, BRCA1 mutations are rarely found in sporadic breast cancer cases and less than 15% of TN tumours harbour BRCA1 mutations [13–15]. The BRCA1 gene is involved in homologous recombination DNA repair, which is the least error-prone mechanism for cells to repair double-stranded DNA breaks [16]. Cells that lack functional BRCA1, whether it is through mutation, loss of heterozygosity or epigenetic mechanisms, are deficient in homologous recombination repair. These cells utilise alternative DNA repair mechanisms that are more error prone, resulting in tumours with high levels of genomic instability [17, 18], a high frequency of TP53 mutations [19] and numerous copy number aberrations [20]. These characteristic patterns of gains and losses of genomic DNA associated with BRCA1 mutant tumours can be used to identify a larger group of sporadic cancers that are molecularly similar but lack BRCA1 mutations, known as BRCA1-like [20–22]. The term ‘BRCA-ness’ similarly refers to tumours in which no germline BRCA1 mutation has been identified but which share histopathological features frequently found in BRCA1 mutated tumours, including a high mitotic index, pushing borders, syncytial and circumscribed growth patterns [23, 24].
There is considerable evidence that epigenetic mechanisms, in particular hypermethylation of tumour suppressor gene promoters, represent an alternative method of gene silencing/ inactivation [7, 24–26]. Methylation of the BRCA1 promoter in breast tumours is associated with a poor overall survival and disease-free survival and has been suggested as a biomarker to guide prognosis and targeted therapies [27–30]. Severson et al found that germline mutation and BRCA1 promoter methylation overlap with BRCA1-like status (determined by copy number aberrations) in 70% and 79% of their samples respectively [20]. Triple negative tumours in young women with multiple BRCA1-like morphological features are associated with hypermethylation of the BRCA1 promoter in blood DNA [23]. However, there remains debate regarding whether blood and tumour data are concordant for gene specific methylation [31]. Tumour BRCA1 promoter methylation has been reported to predict response to platinum based chemotherapy agents and Poly(ADP-ribose) polymerase (PARP) inhibitors, therefore methylation status could potentially influence treatment decisions [32].
In order to examine the relationship between BRCA1 promoter methylation, BRCA1 protein expression, triple negativity and BRCA1 associated histopathological features we have analysed blood samples from 658 women with sporadic breast cancer and 170 matched tumour samples; 71 (11%) and 35 (21%) of these samples were classified as TN respectively.
Methods
Study population and data collection
The study population comprised women diagnosed with invasive breast cancer at Sheffield Teaching Hospitals NHS Foundation Trust, UK, recruited as part of the Sheffield Breast Cancer Study. The study was approved by the Sheffield Research Ethics Committee, and all women provided written informed consent. Women were recruited in two cohorts between 1998–2008 and 2009–2014. Women in the earlier cohort were recruited from surgical outpatient clinics, whilst women in the later cohort were newly diagnosed and recruited at pre-operative assessment. Women with known BRCA1/2 gene mutations were excluded. Data on tumour grade, receptor status, nodal status and age at diagnosis were obtained from clinical notes, and menopausal status and family history of breast cancer were obtained from the patient by questionnaire administered by a research nurse. Samples from women with triple negative tumours were preferentially selected for BRCA1 promoter methylation analysis.
Sample collection and DNA isolation
Genomic DNA was extracted from peripheral blood mononuclear cells isolated from 6ml whole blood or 2ml ‘buffy coat’ samples according to the manufacturer’s protocol (Flexigene DNA extraction kit, Qiagen). The concentration of extracted DNA was quantified using a NanoDrop spectrophotometer (Thermo Scientific, ND-1000 software). DNA samples were stored at -80°C until required. Tumour or normal tissue DNA was isolated following macrodissection from five 10 micron paraffin sections per FFPE block (to ensure greater than 80% tumour cells), and extracted according to the manufacturer’s protocol (QIAamp DNA FFPE kit, Qiagen). Tumour DNA was eluted in a final volume of 70μl buffer and then quantified using a NanoDrop Spectrophotometer.
Receptor status
For the 1998–2008 cohort, tumour receptor status for ER, PR and HER2 were determined by immunohistochemistry of triplicate tumour cores in tissue microarrays, and scored by SSC. Antibodies were as follows; ER: Vector 6F11/2 (1:50), PR: Vector 1A6 (1:40), HER2: Dako HercepTest Kit (pre-diluted). For the 2009–2014 cohort, ER and HER2 status determined according to UK guidelines were obtained from NHS histopathology records.
Morphological scoring
Haematoxylin and Eosin (H&E) staining was performed on one slide per tumour and the slides were systematically reviewed by a consultant histopathologist (SSC), who scored them for the presence of nine BRCA1-associated morphological features; high mitotic index, malignant nuclear grade, little or no (<10%) tubule formation, trabecular growth pattern, syncytial growth pattern, pushing margins (>50%), circumscribed growth pattern, necrosis, moderate or intense lymphocytic infiltrate [23, 33].
BRCA1 protein expression
BRCA1 immunohistochemistry was carried out using the anti-BRCA1 (Ab-1) mouse antibody (MS110 OP92 Calbiochem) on 5micron FFPE sections at 1:400 dilution as described previously [36]. MCF-7 cell line cytospins were used to provide positive and negative (no primary antibody) controls. Slides were scored for BRCA1 nuclear staining by SSC and OW. The percentage of positive nuclei were scored between 0 and 5, the intensity of nuclear staining was scored between 0 and 3, then these were added to form the combined score (Allred quick score). Tumours with a score equal or less than 4 were deemed to be negative for BRCA1 expression.
Methylation analysis
Sodium-bisulphite modification of blood and tumour DNA was performed using the CpGenome DNA modification kit (EMD Millipore, USA) according to the manufacturer’s protocol, to convert unmethylated cytosine residues to uracil. Sodium bisulphite-treated DNA was then analysed by pyrosequencing as described previously [34]. Two sets of pyrosequencing primers were designed for a 313 base pair (bp) region of the BRCA1 promoter using Pyromark assay design software (version 2.0) and are detailed in Table 1. The pyrosequencing targets contained 11 CpG sites, which included all of those studied by Wong et al plus two additional adjacent sites (Fig 1) [23, 35]. CpG sites are referred to by their base pair position relative to the BRCA1 transcription start site (position zero). The oligonucleotides (Sigma Aldrich, Ebersberg, Germany) were reconstituted with deionised water at a stock concentration of 100pmol/μl.
Table 1. Details the primer sequences used for pyrosequencing.
| 5’3’ sequence | Strand | ||
|---|---|---|---|
| Target region 1 | F | (Biotin)TGATTTAGTATTTTGAGAGGTTGTTGTT | Sense |
| R | CAATTATCTAAAAAACCCCACAACCTA | Reverse | |
| S | CCCACAACCTATCCC | Reverse | |
| Target region 2 | F | GTATTTTTGAGAGGTTGTTGTT | Sense |
| R | (Biotin)AAACCCCACAACCTATCC | Reverse | |
| S | TTTGAGAGGTTGTTGTTTA | Sense |
Primer orientation: Forward (F), Reverse (R), Sequencing (S).
Fig 1. Diagram of the 11 CpG sites analysed by pyrosequencing.
Individual CpG sites are underlined and numbering is relative to the BRCA1 Transcription Start Site (TSS). The start and end of the two overlapping target regions are detailed. No CpGs were present in the overlapping section.
PCR was performed using Hotstart taq DNA polymerase (Hotstart PCR kit, Qiagen) under the following PCR conditions; denaturation at 95°C for 15 minutes followed by 50 cycles of the following profile; 95°C for 20 seconds, 61°C for 20 seconds, 72°C for 20 seconds followed by a final 5 minutes extension at 72°C. The PCR products were analysed by gel electrophoresis on a 1.5% Agarose gel stained with ethidium bromide and visualised by UV trans-illumination prior to pyrosequencing. The 11 CpG sites were analysed by pyrosequencing using PyroMark Q96 MD and Pyromark Gold reagents (Qiagen AG, Basel, Switzerland). Bisulphite-modified universally methylated DNA (Chemicon International, NY) and distilled water were included in each run as positive and negative controls.
Statistical Analysis
Correlation of methylation levels at individual CpG sites between matched tumour and blood DNA, and between matched pairs of tumour samples was based on Spearman’s rank correlation coefficient. Methylation levels between matched pairs of samples were compared using the Wilcoxon matched-pairs signed-ranks test. Unmatched groups were compared using the Mann-Whitney test. Associations between different pathological features were assessed using contingency tables. The data set used for these analyses is provided in S1 Table. All analyses were implemented in Stata V12.1 and all statistical tests were two-sided.
Results
Baseline characteristics of cases and controls
Blood methylation analysis was successfully carried out on 658 cases, of whom 170 had sufficient tumour tissue available for methylation analysis; 71 (11%) and 35 (21%) of these were classified as TN respectively. Normal breast tissue was available for 26 cases, and for 20 cases two tumour FFPE blocks were available from the same tumour. There were no significant differences in baseline demographics (age, menopausal status and history of first degree relative affected with breast cancer) between those from whom tumours were available and those where only blood was available (Table 2). However, the cases where tumours were available were of higher grade (p<0.001), more likely lymph node positive (p = 0.03) and more likely to be TN (p<0.001) (Table 2), reflecting the fact that cases with TN disease were preferentially selected for tumour analysis, and that tumours with sufficient tissue available for DNA extraction tend to be of higher grade and node positive since these features are associated with larger tumours.
Table 2. Study population demographics according to blood and tumour tissue availability.
| All cases with blood | Cases with blood only | Cases with blood and tumour | Blood plus tumour versus blood only | Cases with two tumour FFPE blocks | Cases with normal tissue | |
|---|---|---|---|---|---|---|
| Number, n | 658 | 488 | 170 | 20 | 26 | |
| Median age at diag (range) | 60 (23–92) | 60 (23–92) | 58 (24–85) | p = 0.52 | 62.5 (24–84) | 58.5 (39–84) |
| First degree relative pos | 119 (18.1%) | 93 (19.1%) | 26 (15.3%) | 3 (15.0%) | 4 (15.4%) | |
| First degree relative neg | 539 (81.9%) | 395 (81%) | 144 (84.7%) | 17 (85.0%) | 22 (84.6%) | |
| Total | 658 | 488 | 170 | p = 0.27 | 20 | 26 |
| Pre/peri-menopausal | 177 (28.8%) | 130 (28.6%) | 47 (29.2%) | 6 (31.6%) | 8 (32%) | |
| Post-menopausal | 438 (71.2%) | 324 (71.4%) | 114 (70.8%) | 13 (68.4%) | 17 (68%) | |
| Total | 615 | 454 | 161 | p = 0.89 | 19 | 25 |
| Tumour grade 1 | 135 (21.8%) | 112 (25%) | 23 (14.6%) | 0 (0%) | 3 (12.0%) | |
| Tumour grade 2 | 290 (46.8%) | 224 (48.6%) | 66 (41.8%) | 12 (63.2%) | 12 (48.0%) | |
| Tumour grade 3 | 194 (31.3%) | 125 (27.1%) | 69 (43.7%) | 7 (36.8%) | 10 (40.0%) | |
| Total | 619 | 461 | 158 | p<0.001 | 19 | 25 |
| Lymph node negative | 385 (63.6%) | 298 (66.1%) | 87 (56.5%) | 5 (27.8%) | 14 (53.9%) | |
| Lymph node positive | 220 (36.4%) | 153 (33.9%) | 67 (43.5%) | 13 (72.2%) | 12 (46.1%) | |
| Total | 605 | 451 | 154 | p = 0.03 | 18 | 26 |
| Non Triple Negative (NTN) | 585 (89.2%) | 452 (92.6%) | 133 (79.2%) | 19 (95.0%) | 22 (95.7%) | |
| Triple Negative (TN) | 71 (10.8%) | 36 (7.4%) | 35 (20.8%) | 1 (5.0%) | 1 (4.3%) | |
| Total | 656 | 488 | 168 | p<0.001 | 20 | 23 |
Numbers for each sample group are given in each column. Cases with blood only are compared to cases with blood and tumour available.
Methylation levels vary between CpG sites, and are higher in tumour DNA compared to blood DNA
The levels of BRCA1 promoter methylation at 9 of the 11 CpG sites in blood DNA were found to correlate with methylation at the corresponding sites in matched tumour DNA in the 170 cases for which both were available. Specifically, methylation levels at all sites apart from +27 and +44 were significantly correlated at p<0.0007 (Fig 2A and S2 Table). The levels of methylation were significantly higher in tumour DNA compared to matched blood DNA at all sites apart from +16 (p<0.0025; Fig 2A and S2 Table). Blood DNA methylation levels for subjects with matched tumour (n = 170) were representative of the larger set of blood DNA methylation data (n = 658; Fig 2A). Furthermore, in both blood and tumour DNA there was a distinctive pattern of methylation around the transcription start site, with higher levels of methylation at -37 and -29 compared to the other sites (Fig 2A). DNA was available from normal breast tissue for 26 cases. The BRCA1 promoter methylation levels were generally higher in tumour DNA compared to normal breast tissue DNA, most significantly at -51, -21, -19 and +19 (p<0.03; Fig 2B and S2 Table). For 20 cases, two tumour FFPE blocks from the same tumour were available. The methylation levels were significantly correlated between the pairs of blocks for all CpG sites except -51 and +44 (p<0.04; S3 Table).
Fig 2. Methylation plots comparing blood, tumour and normal breast tissue.
Mean (+/-SD) methylation levels plotted against CpG site position along the chromosome in relation to the BRCA1 transcription start site (position zero). (A) Blood DNA methylation level is shown in red and tumour DNA methylation level in blue for the matched samples (n = 170) and blood DNA methylation level for the whole sample set is shown in orange (n = 658). (B) Normal breast tissue DNA methylation level is shown in green (n = 26) and tumour DNA methylation level in blue (n = 170).
BRCA1-like morphological features are associated with triple negativity, loss of BRCA1 protein expression and higher grade
We were able to score 147 tumours for the presence of 9 BRCA1-like morphological features (Fig 3), and tumours were then grouped according to whether they exhibited five or more features [33]. BRCA1 protein expression was assessed by immunohistochemistry in 119 cases (Fig 3). Higher levels of BRCA1 protein expression were associated with fewer BRCA1-like morphological features (p = 0.019; Table 3). The number of morphological features was strongly associated with triple negativity, larger tumour size and higher grade (p<0.0001, p = 0.034, p<0.0001 respectively; Table 3).
Fig 3. Morphological features and BRCA1 expression.
H&Es demonstrating BRCA1 associated morphological features scores. A: Low score demonstrating good tubule formation, little nuclear pleomorphism, no lymphocytes and no mitoses. B: High score demonstrating syncytial islands, marked nuclear pleomorphism and heavy lymphocytic infiltrate. C, D&E: IHC performed using BRCA1 specific antibody Ab-1. C: Breast tumour demonstrating nuclear staining for BRCA1. D: Breast tumour demonstrating lack of nuclear or cytoplasmic staining for BRCA1. E: Normal breast tissue demonstrating normal nuclear staining for BRCA1. F: Histogram of the number of BRCA1-like morphological features from zero to 9 (n = 147).
Table 3. Morphological feature scoring.
| Number of morphological features (n) | Mean age in years | Median grade | Mean size (mm) | Number TN (%) | BRCA1 expression (%positive) |
|---|---|---|---|---|---|
| ≥5 (53) | 57.0 | 2.8 | 25.4 | 26 (76.5) | 9 (23.7) |
| <5 (94) | 58.8 | 2.1 | 20.7 | 8 (23.5) | 29 (76.3) |
| p = 0.41 | p<0.0001 | p = 0.034 | p<0.0001 | p = 0.019 |
Comparison of number of BRCA1-like morphological features with clinicopathological features; age, grade, size, TN receptor status and BRCA1 expression.
Levels of both blood and tumour DNA methylation are higher in tumours with high BRCA1-like features scores
The levels of tumour methylation were significantly higher in cases with tumours having more than or equal to 5 BRCA1-like features compared to those with less than 5, at all CpG sites except +44 (p = 0.007 to p<0.0001; Fig 4A and 4B and S4 Table). This pattern was also seen at the majority of CpG sites in blood DNA, although the differences were less statistically significant. Interestingly, the +27 CpG site was an exception, where the levels of blood methylation were not significantly different in the group with over 5 BRCA1-like features compared to those with fewer than 5 (p = 0.08 in blood, p<0.0001 in tumour; S4 Table)
Fig 4. Methylation plots comparing blood and tumour for BRCA1-like features, Triple negativity and BRCA1 protein expression.
Mean (+/-SD) methylation levels plotted against CpG site position along the chromosome in relation to the BRCA1 transcription start site (position zero). Blood methylation level is shown on the left and tumour methylation level on the right. A&B: Tumours with <5 BRCA1-like features are shown in orange and those with ≥5 BRCA1-like features are shown in green. C&D: Triple negative tumours are shown in maroon and non-triple negative are shown in teal. E&F: Tumours with high levels of BRCA1 protein expression are shown in brown and those with low levels are shown in cyan.
Levels of both blood and tumour DNA methylation are higher in triple negative tumours
Tumour DNA methylation was available for 35 triple-negative cases and 133 non-triple negative cases, and blood DNA methylation was available for 71 triple-negative and 585 non-triple negative cases. The levels of both blood and tumour methylation were generally higher in triple negative tumours compared to non-triple negative, at the majority of CpG sites apart from -51, -29, +19 and +44 (p value range <0.0001 to 0.03 for blood DNA, and 0.0001 to 0.01 for tumour DNA; Fig 4C and 4D and S5 Table). Again the +27 CpG site was the exception, with lower methylation levels in blood DNA of TN cases compared to non-triple negative cases, in contrast to the higher levels in tumour DNA from TN cases (p = 0.0008 in blood and p = 0.0003 in tumour; S5 Table).
Levels of tumour and blood methylation compared to BRCA1 protein expression levels
There were 38 cases with high BRCA1 protein expression and 81 with low expression for whom blood and tumour DNA methylation were available. There was no difference in blood methylation levels at any CpG site between those with high or low levels of BRCA1 expression. The levels of tumour DNA methylation were generally higher in the samples with lower levels of BRCA1 protein although the differences were not statistically significant (Fig 4E and 4F; S6 Table).
Discussion
BRCA1 promoter methylation may be used to guide therapy
Tumours arising in women with hereditary BRCA1 mutations tend to be TN and basal-like, features that are associated with a poor prognosis. However, in recent years it has become possible to exploit the DNA-repair defects in tumours carrying BRCA1 or BRCA2 gene mutations using PARP inhibitors [37, 38]. The response of BRCA1 mutation-associated cancers to both PARP inhibitors and cisplatin-based chemotherapeutic agents has driven the interest in identifying tumours with a similar DNA-repair deficient phenotype, so that these difficult to treat patients might also benefit from targeted therapies. DNA methylation of the BRCA1 promoter is a moderately frequent event in sporadic breast tumours and an alternative mechanism of BRCA1 inactivation. In vitro studies suggest that cells with BRCA1 CpG island methylation are also sensitive to PARP1 inhibitors and tumour BRCA1 promoter methylation predicts response to platinum based chemotherapy agents and PARP inhibitors [32, 39, 40].
A recent meta-analysis of BRCA1 promoter methylation studies reported an association between BRCA1 methylation and BRCA-like clinico-pathological features such as lymph node metastasis, histological grade 3, ER and PR negativity, triple-negative phenotype and decreased BRCA1 protein expression [41]. However the majority of studies included in the meta-analysis used methylation-specific PCR or other methods that do not distinguish individual CpG sites, limiting mechanistic interpretation. Even a recent study of BRCA1 promoter methylation using pyrosequencing analysed the results by averaging the methylation levels across all sites, thus not utilizing the CpG site-specific results generated by pyrosequencing [42]. The considerable heterogeneity between studies highlights the difficulties in drawing meaningful conclusions when different CpG sites have been studied, methylation detection methods used, populations studied and tissues examined [40]. In this study we have used pyrosequencing to distinguish methylation levels at individual CpG sites in the BRCA1 promoter, and analysed CpG sites that had been studied in at least three previous studies [23, 43, 44]. We have focused on obtaining a comprehensive dataset consisting of methylation levels for blood and tumour DNA, BRCA1 protein expression, hormone receptor status, and morphological and clinico-pathological features.
Blood and tumour methylation levels are related to BRCA-like phenotypes
We found that there was a strong correlation between methylation levels in blood and tumour DNA at all CpG sites apart from +27 and +44, with levels in the tumour being consistently higher than those in the blood (or normal breast tissue), at all CpG sites. Whether the strong correlation between these different tissues reflects independent events, constitutional changes or global methylation changes secondary to carcinogenesis is beyond the scope of this study due to its retrospective nature; prospective analyses are required to determine whether blood methylation could be used as a predictor of tumour methylation [45, 46]. The effect of DNA methylation on gene expression is complex with both hypo- and hypermethylation at specific gene regions differentially affecting gene expression [47], however there is limited mechanistic work on individual CpG sites.
Consistent with previous observations, we found that BRCA1-like morphological features are correlated with triple negativity, loss of BRCA1 protein expression and higher grade. The levels of both blood and tumour DNA methylation at most CpG sites were higher in tumours with high BRCA1-like features scores and were also higher in triple negative tumours, as was shown by Wong et al [23]. The overall picture was that the associations in blood DNA were weaker than those in tumour DNA but in the same direction. The +27 CpG site was an exception to this general pattern where the associations tended to be in the opposite direction in blood and tumour DNA. We had limited power to detect associations with BRCA1 protein expression, and although mean tumour DNA methylation (at sites +8 to +44 in particular) were higher in tumours with lower levels of BRCA1 protein, these effects did not reach statistical significance. Previous studies have shown a relationship between DNA methylation and BRCA1 protein [24, 48, 49].
Other methylation analysis techniques more commonly used, including MSP (Methylation Specific PCR), MS-HRM (methylation-sensitive high-resolution melting) Methyl-light and MS-MLPA (methylation-specific multiplex ligation-dependent probe amplification) do not give as much detailed information about the methylation status and often quintiles or arbitrary cut off points are chosen to define ‘methylated’ or ‘unmethylated’ promoters. The meta-analysis by Zhang et al found that over half of the studies included used MSP as their predominant method for analysing methylation, whilst only one study used pyrosequencing [41, 50]. Methylation analysis using pyrosequencing is becoming widely used in diagnostics laboratories, which may drive further translational research [51]. Future studies may need to focus on appropriate methods of methylation analysis to detect levels in biopsy specimens, particularly in TN tumours, because of the increasing use of neo-adjuvant chemotherapy.
The distribution of histopathological features, including receptor status, age at diagnosis and grade, can be used to predict women more likely to harbour germline BRCA1 or BRCA2 mutations [52]. Scoring breast tumours for morphological features associated with ‘BRCAness’, as has been performed in our study, can be used to help identify which tumours may have higher levels of promoter methylation [23, 33]. This information could be used in future studies alongside receptor status, age at diagnosis and histological grade to select a subgroup of patients for epigenetic and genetic testing and subsequent targeted therapies. The use of tumour histopathology is gaining acceptance as a way to target costly and time-consuming genetic testing to ‘at-risk’ individuals based on their tumour characteristics [53].
This study highlights the variability in methylation level at different CpG sites close to the BRCA1 transcription start site. Methylation levels in tumour are generally greater than those in blood, and methylation at most sites (apart from +27) increases in triple negative tumours and those with a high BRCA1-like features scores. Analysis of BRCA1 promoter methylation may contribute to strategies for the identification of women who may benefit from PARP inhibition or other targeted therapies, as has occurred in BRCA associated ovarian cancer [54].
Supporting Information
(XLSX)
The paired sign-rank tests have been used.
(XLSX)
(XLSX)
(XLSX)
(XLSX)
(XLSX)
Acknowledgments
We would particularly like to thank all the women who participated in the study. We would also like to thank the Sheffield Diagnostic Genetics Service and the Leeds Institute of Cancer and Pathology for use of their pyrosequencing facilities.
Data Availability
All relevant data are within the paper and its Supporting Information files.
Funding Statement
Cancer Research UK (CRUK) provided funding for this project in the form of a Research Bursary: Career and Development Fund ref C47655/A16817 www.cruk.org SLD IWB LW AC. The National Institute for Health Research (NIHR) provided salary contribution for SLD via an Academic Clinical Fellowship www.nihr.ac.uk. The Sheffield Experimental Cancer Medicine Centre provided a research grant www.sheffield.ac.uk to SLD LW AC. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Badve S, Dabbs DJ, Schnitt SJ, Baehner FL, Decker T, Eusebi V, et al. Basal-like and triple-negative breast cancers: a critical review with an emphasis on the implications for pathologists and oncologists. Mod Pathol. 2011;24(2):157–67. 10.1038/modpathol.2010.200 [DOI] [PubMed] [Google Scholar]
- 2.Reis-Filho JS, Lakhani SR. Breast cancer special types: why bother? J Pathol. 2008;216(4):394–8. 10.1002/path.2419 [DOI] [PubMed] [Google Scholar]
- 3.Carey LA, Dees EC, Sawyer L, Gatti L, Moore DT, Collichio F, et al. The triple negative paradox: primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res. 2007;13(8):2329–34. [DOI] [PubMed] [Google Scholar]
- 4.Dent R, Trudeau M, Pritchard KI, Hanna WM, Kahn HK, Sawka CA, et al. Triple-negative breast cancer: clinical features and patterns of recurrence. Clin Cancer Res. 2007;13(15 Pt 1):4429–34. [DOI] [PubMed] [Google Scholar]
- 5.Hudis CA, Gianni L. Triple-negative breast cancer: an unmet medical need. Oncologist. 2011;16 Suppl 1:1–11. 10.1634/theoncologist.2011-S1-01 [DOI] [PubMed] [Google Scholar]
- 6.Perou CM. Molecular stratification of triple-negative breast cancers. Oncologist. 2011;16 Suppl 1:61–70. 10.1634/theoncologist.2011-S1-61 [DOI] [PubMed] [Google Scholar]
- 7.Foulkes WD, Stefansson IM, Chappuis PO, Begin LR, Goffin JR, Wong N, et al. Germline BRCA1 mutations and a basal epithelial phenotype in breast cancer. J Natl Cancer Inst. 2003;95(19):1482–5. [DOI] [PubMed] [Google Scholar]
- 8.Burstein MD, Tsimelzon A, Poage GM, Covington KR, Contreras A, Fuqua SA, et al. Comprehensive genomic analysis identifies novel subtypes and targets of triple-negative breast cancer. Clin Cancer Res. 2015;21(7):1688–98. 10.1158/1078-0432.CCR-14-0432 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr Y, et al. Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011;121(7):2750–67. 10.1172/JCI45014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Turner NC, Reis-Filho JS, Russell AM, Springall RJ, Ryder K, Steele D, et al. BRCA1 dysfunction in sporadic basal-like breast cancer. Oncogene. 2007;26(14):2126–32. [DOI] [PubMed] [Google Scholar]
- 11.Thike AA, Cheok PY, Jara-Lazaro AR, Tan B, Tan P, Tan PH. Triple-negative breast cancer: clinicopathological characteristics and relationship with basal-like breast cancer. Mod Pathol. 2010;23(1):123–33. 10.1038/modpathol.2009.145 [DOI] [PubMed] [Google Scholar]
- 12.Rummel S, Varner E, Shriver CD, Ellsworth RE. Evaluation of BRCA1 mutations in an unselected patient population with triple-negative breast cancer. Breast Cancer Res Treat. 2013;137(1):119–25. 10.1007/s10549-012-2348-2 [DOI] [PubMed] [Google Scholar]
- 13.Evans DG, Howell A, Ward D, Lalloo F, Jones JL, Eccles DM. Prevalence of BRCA1 and BRCA2 mutations in triple negative breast cancer. J Med Genet. 2011;48(8):520–2. 10.1136/jmedgenet-2011-100006 [DOI] [PubMed] [Google Scholar]
- 14.Futreal PA LQ. BRCA1 mutations in primary breast and ovarian carcinomas. Science. 1994;266:120–2. [DOI] [PubMed] [Google Scholar]
- 15.Arver B DQ, Chen J, Luo L and Lindblom A. Hereditary breast cancer: a review. Seminars in Cancer Biology. 2000;10:271–88. [DOI] [PubMed] [Google Scholar]
- 16.Bouwman P, Jonkers J. The effects of deregulated DNA damage signalling on cancer chemotherapy response and resistance. Nat Rev Cancer. 2012;12(9):587–98. 10.1038/nrc3342 [DOI] [PubMed] [Google Scholar]
- 17.Turner N, Tutt A, Ashworth A. Hallmarks of 'BRCAness' in sporadic cancers. Nat Rev Cancer. 2004;4(10):814–9. [DOI] [PubMed] [Google Scholar]
- 18.Wang H, Zeng ZC, Bui TA, DiBiase SJ, Qin W, Xia F, et al. Nonhomologous end-joining of ionizing radiation-induced DNA double-stranded breaks in human tumor cells deficient in BRCA1 or BRCA2. Cancer Res. 2001;61(1):270–7. [PubMed] [Google Scholar]
- 19.Manie E, Vincent-Salomon A, Lehmann-Che J, Pierron G, Turpin E, Warcoin M, et al. High frequency of TP53 mutation in BRCA1 and sporadic basal-like carcinomas but not in BRCA1 luminal breast tumors. Cancer Res. 2009;69(2):663–71. 10.1158/0008-5472.CAN-08-1560 [DOI] [PubMed] [Google Scholar]
- 20.Severson TM, Peeters J, Majewski I, Michaut M, Bosma A, Schouten PC, et al. BRCA1-like signature in triple negative breast cancer: Molecular and clinical characterization reveals subgroups with therapeutic potential. Mol Oncol. 2015;9(8):1528–38. 10.1016/j.molonc.2015.04.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lips EH, Laddach N, Savola SP, Vollebergh MA, Oonk AM, Imholz AL, et al. Quantitative copy number analysis by Multiplex Ligation-dependent Probe Amplification (MLPA) of BRCA1-associated breast cancer regions identifies BRCAness. Breast Cancer Res. 2011;13(5):R107 10.1186/bcr3049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schouten PC, van Dyk E, Braaf LM, Mulder L, Lips EH, de Ronde JJ, et al. Platform comparisons for identification of breast cancers with a BRCA-like copy number profile. Breast Cancer Res Treat. 2013;139(2):317–27. 10.1007/s10549-013-2558-2 [DOI] [PubMed] [Google Scholar]
- 23.Wong EM, Southey MC, Fox SB, Brown MA, Dowty JG, Jenkins MA, et al. Constitutional methylation of the BRCA1 promoter is specifically associated with BRCA1 mutation-associated pathology in early-onset breast cancer. Cancer Prev Res (Phila). 2011;4(1):23–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bal A, Verma S, Joshi K, Singla A, Thakur R, Arora S, et al. BRCA1-methylated sporadic breast cancers are BRCA-like in showing a basal phenotype and absence of ER expression. Virchows Arch. 2012;461(3):305–12. 10.1007/s00428-012-1286-z [DOI] [PubMed] [Google Scholar]
- 25.Esteller M, Silva JM, Dominguez G, Bonilla F, Matias-Guiu X, Lerma E, et al. Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J Natl Cancer Inst. 2000;92(7):564–9. [DOI] [PubMed] [Google Scholar]
- 26.Pang D, Zhao Y, Xue W, Shan M, Chen Y, Zhang Y, et al. Methylation profiles of the BRCA1 promoter in hereditary and sporadic breast cancer among Han Chinese. Med Oncol. 2012;29(3):1561–8. 10.1007/s12032-011-0100-0 [DOI] [PubMed] [Google Scholar]
- 27.Zhu X, Shan L, Wang F, Wang J, Wang F, Shen G, et al. Hypermethylation of BRCA1 gene: implication for prognostic biomarker and therapeutic target in sporadic primary triple-negative breast cancer. Breast Cancer Res Treat. 2015;150(3):479–86. 10.1007/s10549-015-3338-y [DOI] [PubMed] [Google Scholar]
- 28.Wu L, Wang F, Xu R, Zhang S, Peng X, Feng Y, et al. Promoter methylation of BRCA1 in the prognosis of breast cancer: a meta-analysis. Breast Cancer Res Treat. 2013;142(3):619–27. 10.1007/s10549-013-2774-9 [DOI] [PubMed] [Google Scholar]
- 29.Guo T, Ren Y, Wang B, Huang Y, Jia S, Tang W, et al. Promoter methylation of is associated with estrogen, progesterone and human epidermal growth factor receptor-negative tumors and the prognosis of breast cancer: A meta-analysis. Mol Clin Oncol. 2015;3(6):1353–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Xu X, Gammon MD, Zhang Y, Bestor TH, Zeisel SH, Wetmur JG, et al. BRCA1 promoter methylation is associated with increased mortality among women with breast cancer. Breast Cancer Res Treat. 2009;115(2):397–404. 10.1007/s10549-008-0075-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cho YH, McCullough LE, Gammon MD, Wu HC, Zhang YJ, Wang Q, et al. Promoter Hypermethylation in White Blood Cell DNA and Breast Cancer Risk. J Cancer. 2015;6(9):819–24. 10.7150/jca.12174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Veeck J, Ropero S, Setien F, Gonzalez-Suarez E, Osorio A, Benitez J, et al. BRCA1 CpG island hypermethylation predicts sensitivity to poly(adenosine diphosphate)-ribose polymerase inhibitors. J Clin Oncol. 2010;28(29):e563–4; author reply e5-6. 10.1200/JCO.2010.30.1010 [DOI] [PubMed] [Google Scholar]
- 33.Southey MC, Ramus SJ, Dowty JG, Smith LD, Tesoriero AA, Wong EE, et al. Morphological predictors of BRCA1 germline mutations in young women with breast cancer. Br J Cancer. 2011;104(6):903–9. 10.1038/bjc.2011.41 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Tost J, Gut IG. DNA methylation analysis by pyrosequencing. Nat Protoc. 2007;2(9):2265–75. [DOI] [PubMed] [Google Scholar]
- 35.Wojdacz TK, Hansen LL, Dobrovic A. A new approach to primer design for the control of PCR bias in methylation studies. BMC Res Notes. 2008;1:54 10.1186/1756-0500-1-54 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wilson CA, Ramos L, Villasenor MR, Anders KH, Press MF, Clarke K, et al. Localization of human BRCA1 and its loss in high-grade, non-inherited breast carcinomas. Nat Genet. 1999;21(2):236–40. [DOI] [PubMed] [Google Scholar]
- 37.Bryant HE, Schultz N, Thomas HD, Parker KM, Flower D, Lopez E, et al. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. Nature. 2005;434(7035):913–7. [DOI] [PubMed] [Google Scholar]
- 38.Farmer H, McCabe N, Lord CJ, Tutt AN, Johnson DA, Richardson TB, et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature. 2005;434(7035):917–21. [DOI] [PubMed] [Google Scholar]
- 39.Cai F, Ge I, Wang M, Biskup E, Lin X, Zhong X. Pyrosequencing analysis of BRCA1 methylation level in breast cancer cells. Tumour Biol. 2014;35(4):3839–44. 10.1007/s13277-013-1508-2 [DOI] [PubMed] [Google Scholar]
- 40.Ibragimova I, Cairns P. Assays for hypermethylation of the BRCA1 gene promoter in tumor cells to predict sensitivity to PARP-inhibitor therapy. Methods Mol Biol. 2011;780:277–91. 10.1007/978-1-61779-270-0_17 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhang L, Long X. Association of BRCA1 promoter methylation with sporadic breast cancers: Evidence from 40 studies. Sci Rep. 2015;5:17869 10.1038/srep17869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Cai FF, Chen S, Wang MH, Lin XY, Zhang L, Zhang JX, et al. Pyrosequencing quantified methylation level of BRCA1 promoter as prognostic factor for survival in breast cancer patient. Oncotarget. 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Rice JC, Ozcelik H, Maxeiner P, Andrulis I, Futscher BW. Methylation of the BRCA1 promoter is associated with decreased BRCA1 mRNA levels in clinical breast cancer specimens. Carcinogenesis. 2000;21(9):1761–5. [DOI] [PubMed] [Google Scholar]
- 44.Snell C, Krypuy M, Wong EM, kConFab i, Loughrey MB, Dobrovic A. BRCA1 promoter methylation in peripheral blood DNA of mutation negative familial breast cancer patients with a BRCA1 tumour phenotype. Breast Cancer Res. 2008;10(1):R12 10.1186/bcr1858 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Terry MB, Delgado-Cruzata L, Vin-Raviv N, Wu HC, Santella RM. DNA methylation in white blood cells: association with risk factors in epidemiologic studies. Epigenetics. 2011;6(7):828–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Brennan K, Garcia-Closas M, Orr N, Fletcher O, Jones M, Ashworth A, et al. Intragenic ATM methylation in peripheral blood DNA as a biomarker of breast cancer risk. Cancer Res. 2012;72(9):2304–13. 10.1158/0008-5472.CAN-11-3157 [DOI] [PubMed] [Google Scholar]
- 47.Gyorffy B, Bottai G, Fleischer T, Munkacsy G, Budczies J, Paladini L, et al. Aberrant DNA methylation impacts gene expression and prognosis in breast cancer subtypes. Int J Cancer. 2016;138(1):87–97. 10.1002/ijc.29684 [DOI] [PubMed] [Google Scholar]
- 48.Ignatov T, Poehlmann A, Ignatov A, Schinlauer A, Costa SD, Roessner A, et al. BRCA1 promoter methylation is a marker of better response to anthracycline-based therapy in sporadic TNBC. Breast Cancer Res Treat. 2013;141(2):205–12. 10.1007/s10549-013-2693-9 [DOI] [PubMed] [Google Scholar]
- 49.Stefansson OA, Jonasson JG, Olafsdottir K, Hilmarsdottir H, Olafsdottir G, Esteller M, et al. CpG island hypermethylation of BRCA1 and loss of pRb as co-occurring events in basal/triple-negative breast cancer. Epigenetics. 2011;6(5):638–49. 10.4161/epi.6.5.15667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Sturgeon SR, Balasubramanian R, Schairer C, Muss HB, Ziegler RG, Arcaro KF. Detection of promoter methylation of tumor suppressor genes in serum DNA of breast cancer cases and benign breast disease controls. Epigenetics. 2012;7(11):1258–67. 10.4161/epi.22220 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Newton K, Jorgensen NM, Wallace AJ, Buchanan DD, Lalloo F, McMahon RF, et al. Tumour MLH1 promoter region methylation testing is an effective prescreen for Lynch Syndrome (HNPCC). J Med Genet. 2014;51(12):789–96. 10.1136/jmedgenet-2014-102552 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Spurdle AB, Couch FJ, Parsons MT, McGuffog L, Barrowdale D, Bolla MK, et al. Refined histopathological predictors of BRCA1 and BRCA2 mutation status: a large-scale analysis of breast cancer characteristics from the BCAC, CIMBA, and ENIGMA consortia. Breast Cancer Res. 2014;16(6):3419 10.1186/s13058-014-0474-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Hopper JL, Jenkins MA, Dowty JG, Dite GS, Apicella C, Keogh L, et al. Using tumour pathology to identify people at high genetic risk of breast and colorectal cancers. Pathology. 2012;44(2):89–98. 10.1097/PAT.0b013e32834e8e5b [DOI] [PubMed] [Google Scholar]
- 54.De Picciotto N, Cacheux W, Roth A, Chappuis PO, Labidi-Galy SI. Ovarian cancer: Status of homologous recombination pathway as a predictor of drug response. Crit Rev Oncol Hematol. 2016;101:50–9. 10.1016/j.critrevonc.2016.02.014 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(XLSX)
The paired sign-rank tests have been used.
(XLSX)
(XLSX)
(XLSX)
(XLSX)
(XLSX)
Data Availability Statement
All relevant data are within the paper and its Supporting Information files.