BRCA1 promoter hypermethylation is not associated with germline variants in Polish breast cancer patients

TL;DR

This study found that a specific genetic variant in the BRCA1 gene is not linked to BRCA1 methylation or breast cancer risk in Polish patients, but may be associated with certain cancer types and metastases.

Contribution

The study identifies a new genetic variant in the BRCA1 promoter and explores its potential associations with breast cancer subtypes and metastases in a Polish population.

Findings

The BRCA1 c.20 + 101 C/G variant (rs799905) was not significantly associated with BRCA1 methylation or breast cancer risk.

A borderline association was found between rs799905 and triple-negative breast cancer and lymph node metastases.

Among BRCA1 pathogenic variant carriers, the GG genotype was more frequent compared to non-carriers.

Abstract

Methylation of BRCA1 has been associated with an increased risk of breast cancer and specific clinical characteristics of the disease. In the British population, the genetic alteration c.-107 A/T has been shown to cause allelic methylation, leading to familial breast and ovarian cancer. However, this variant has not been detected in Polish population. Nonetheless, other genetic variants may still be associated with BRCA1 methylation, highlighting the need for further research.

This study aimed to analyze BRCA1 promoter region to identify germline alterations associated with BRCA1 methylation in peripheral blood DNA. Additionally, the correlation between the detected variants and breast cancer incidence, as well as clinical characteristics, was assessed.

One hundred breast cancer patients with BRCA1 methylation were analyzed using pyrosequencing to quantify methylation levels. Immunohistochemistry (IHC) was performed on formalin-fixed paraffin-embedded (FFPE) tumor tissue samples from these patients to assess BRCA1 expression. In 47 patients with the highest BRCA1 methylation and decreased BRCA1 expression, Sanger sequencing was performed encompassing 643 bp of BRCA1 promoter to identify potential variants associated with methylation and/or breast cancer. A variant was identified and genotyped in multiple patient groups, including 336 BRCA1 methylation-positive women, 1898 unselected breast cancer cases, 2234 healthy controls, and 309 BRCA1 pathogenic variant (PV) carriers. Statistical analyses were performed using Fisher’s exact test and Chi-square test in Stata/IC version 16.1.

The study identified BRCA1 c.20 + 101 C/G (rs799905) in the promoter region of the gene. However, no significant association was found between rs799905 and BRCA1 methylation or breast cancer risk. Nevertheless, a borderline association was observed between rs799905 and the occurrence of triple-negative breast cancer (TNBC) (p = 0.048) as well as lymph node (LN) metastases (p = 0.046). Among BRCA1 PV carriers, the GG genotype was significantly more frequent, while the CC genotype was 10 times less common compared to non-carriers (p = 0.003).

Rs799905 variant is not associated with BRCA1 methylation breast cancer risk. However, it may be linked to the occurrence of TNBC and LN metastases. Additionally, rs799905 may be associated to some extent with BRCA1 PVs, though further research is needed to clarify the nature of this potential correlation.

Full text

Background

DNA methylation is an epigenetic mechanism that targets cytosines adjacent to guanines (CpG dinucleotides), which are often found in CpG islands within gene promoters and repetitive DNA regions [1]. The overall methylation pattern (methylome) is established after fertilization and maintained throughout development [2]. In healthy cells, repetitive DNA regions are methylated to preserve genomic stability, whereas gene promoters typically remain unmethylated to enable gene expression [1]. This process is essential for cell differentiation and tissue specificity, leading to tissue-specific methylation patterns based on functional requirements [3]. DNA methylation has been implicated in carcinogenesis [1]. Somatic inactivation of tumor suppressor genes, such as BRCA1 and MLH1, through promoter methylation leads to gene silencing and contributes to cancer development [4, 5]. Additionally, promoter methylation detected in peripheral blood DNA has been associated with an increased cancer risk, similar to germline PVs [6]. It has been suggested that aberrant promoter methylation may result from germline DNA sequence variants acting in cis [7]. These genetic alterations could help explain the missing heritable component of cancer in patients who lack detectable germline PVs in gene’s coding sequence [8]. This phenomenon has been described for MLH1 [9, 10] and MSH2 [11], where specific germline variants in 5’UTR were linked to promoter methylation, leading to loss of expression and causing hereditary colorectal cancer. A cis-acting germline alteration located in the 5’ UTR of BRCA1 gene (c.-107A > T, rs2154580237), has been previously identified in 2018 in two British families and was found to segregate with BRCA1 promoter hypermethylation, predisposing carriers to breast and ovarian cancer [12]. However, this variant was not detected in 193 Dutch [13] and 3297 German [14] patients. It was also not found in 942 Polish breast cancer patients and 500 controls [unpublished data]. Nonetheless, it remains possible that other germline variants in the BRCA1 5’ UTR region are associated with promoter hypermethylation and may be connected with cancer risk in different populations.

In this study, we aimed to analyze the BRCA1 promoter region to identify germline alterations associated with BRCA1 methylation in peripheral blood DNA. Additionally, detected variants were assessed and correlated with breast cancer incidence and other clinical characteristics.

Materials and methods

The study was conducted in 2 related analyses:

1. Identification of germline alterations in 5’UTR/promoter region of BRCA1 gene associated with methylation detected in peripheral blood DNA.
2. Association analysis and clinical characteristics of identified germline variants in Polish breast cancer patients.

Study population

1. The initial study group consisted of 100 breast cancer (BC) patients with previously detected BRCA1 promoter methylation in peripheral blood and paired tumor tissue DNA with the MS-HRM technique [15]. Among them 47 BC patients selected on the results of pyrosequencing and BRCA1 expression in tumor tissue that subsequently were analyzed by Sanger sequencing for germline alterations associated with BRCA1 promoter methylation in peripheral blood DNA. All patients were negative for the common Polish germline BRCA1/2 PVs [16, 17].
2. Association analysis of germline variant identified by sequencing was performed in 3 independent groups:

• Group I: 336 women with detected BRCA1 promoter methylation in peripheral blood, consisted of 189 BC patients and 147 healthy controls [15]; 109 healthy women were from an ongoing prospective study.
• Group II: 1898 unselected BC patients and 2234 healthy controls.
• Group III: 309 BRCA1 PV carriers consisting of 41 BC patients and 268 healthy women.

In Group II and Group III status of BRCA1 promoter methylation was unknown.

In breast cancer patients from Group II clinical and histopathological data including age of diagnosis, age of death, cancer type, hormone receptor status, tumor size, metastases, bilaterality, family history and BRCA1 PV status were collected.

All individuals for this study were selected from the registry of International Hereditary Cancer Center, Department of Genetics and Pathology Pomeranian Medical University in Szczecin. Each patient has signed an informed consent for study participation.

The study was approved by Ethics Committee of the Pomeranian Medical University in Szczecin.

Genomic DNA isolation

The DNA isolation was performed using the detergent method as previously described [18]. After isolation the DNA samples were stored at 4°C prior to analysis.

Pyrosequencing of BRCA1 promoter region

Quantitative DNA methylation analysis by pyrosequencing was performed on peripheral blood DNA samples of 100 breast cancer patients (with detected BRCA1 methylation, as indicated in the Study population), which was previously bisulfite converted with the use of EZ DNA Methylation Kit (Zymo Research) according to the manufacturer’s protocol. Pyrosequencing was performed on the PyroMark MD Instrument (Biotage) according to manufacturer’s procedure. The 6 CpG sites in BRCA1 promoter region were examined (+ 8, + 14, +16, + 19, +27, + 44). Primer sequences used for pyrosequencing are presented in Table 1. The methylation levels of the individual CpG sites varied between 4% and 25% across all samples. The average methylation level of the samples ranged from 5 to 21%.

Table 1. Sequences of primers used in pyrosequencing (5’→3’)Primer FGGGTAGATTGGGTGGTTAATTPrimer R^a^TTATCTAAAAAACCCCACAACCTATCPrimer SGGGAATTATAGATAAATTAAAATTG^a^biotinylated

BRCA1 expression

Expression of BRCA1 was performed by immunohistochemistry (IHC) and encompassed 100 BC tissue samples (with previously detected BRCA1 gene methylation, as indicated in Study population). Tissue microarrays of BC patients were constructed as described previously [19]. The slides were immunostained using a Dako EnVision FLEX + visualization system according to the manufacturer’s instructions. Mouse monoclonal BRCA1 antibody (clone MS110, Calbiochem; dilution 1: 50; incubation time 30 min) was used. The expression of BRCA1 protein was classified as negative (< 10% positive tumor nuclei) or positive (≥ 10% tumor cells showing nuclear immunoreactivity).

Sanger sequencing of BRCA1 5'UTR

Sequencing of 5’UTR region of BRCA1 gene, encompassing 643 nucleotides, has been performed on samples initially analyzed by pyrosequencing (quantitative DNA methylation analysis) and immunohistochemical analysis of tissue microarrays (BRCA1 expression). Out of the 100 patients tested, 47 were selected for Sanger sequencing. These patients exhibited the highest average levels of BRCA1 methylation (> 14%) and negative BRCA1 protein expression (< 10% of cells showing immunoreactivity) in paired tumor tissues. The sequencing was performed using the BigDye Terminator v3.1 Cycle Sequencing kit (Life Technologies, Carlsbad, California) according to commonly used protocols. The sequencing products were analyzed using the ABI Prism 3500XL Genetic Analyzer (Life Technologies). Sequences of primers are presented in Table 2.

Table 2. Sequences of primers used in Sanger sequencing (5’→3’)Primer FACTGTGATGCAATAAGCCGCPrimer RGCATATTCCAGTTCCTATCACG

Genotyping

Germline variant identified by sequencing has been genotyped in 3 study groups by Real-time PCR using TaqMan probes on LightCycler 480 Instrument (Roche Diagnostics) according to standard protocol. The reaction mix for each sample consisted of GoTaq^®^ Probe qPCR Master Mix (Promega), TaqMan Genotyping Assay × 40 (Predisigned, C_3178681_10; Applied Biosystems), and deionized water (Promega). Samples were analyzed on 384well plates. Each plate included a negative control with water and positive controls for each genotype.

Statistical analysis

The frequency of each genotype for the detected variant was calculated. Then, for each genotype, odds ratios (OR) with 95% confidence intervals (CI) were determined by comparing it to the reference (most frequent) genotype using Fisher’s exact test. The correlation of detected variants with clinico-pathological features of breast cancers has been analysed using the Chi-square test for the categorical variables. P-values < 0.05 were considered statistically significant. All statistical analyses have been performed with the use of Stata/IC version 16.1.

Results

Identification of germline variant in promoter region of BRCA1 gene

Sequencing of the promoter region revealed the presence of a single variant in BRCA1 c.20 + 101 C/G (rs799905), localized in first intron between alternative exons 1α and 1β. Among 47 women rs799905 was detected in 18 patients (38.3%), including 3 homozygotes.

Association analyses of BRCA1 c.-20 + 101 C/G (rs799905)

The genotype frequency of rs799905 variant was similar in Groups I and II, with the genotype GG being the most common in BC patients (44.4% and 46%) and controls (46.9% and 44.5%), followed closely by genotype GC (BC cases 41.3% and 40.5%, controls 40.1% and 43.8%); the genotype CC was detected in 13.2% and 10.4% of BC patients, and 8.8% and 10% controls from Group I and II, respectively. In contrast, the frequencies of BRCA1 c.-20 + 101 C/G in Group III (consisted of BRCA1 PV carriers) differed significantly from those in Groups I and II. The most common GG genotype was detected in 69.5% BC patients and 61.9% healthy controls. Genotype GC was less frequent, appearing in 29% of cases and 37% of controls. The largest difference between the tested groups was observed in the frequency of genotype CC, which was detected in only 1.2% BC cases and 0.9% of controls in Group III. Association analyses of rs799905 did not show any statistically significant correlation with breast cancer in the 3 studied groups. Genotyping results and calculated odds ratios are presented in Table 3.

Table 3. Association of rs799905 with breast cancer in 3 tested groupsGroup I Genotype

Cases Met (+) ** n ** = 189 (%) ^a^

Controls Met (+) ** n ** = 147 (%) ^a^

OR (95%CI)

p -value GG84 (44.4)69 (46.9)Ref.-GC78 (41.3)59 (40.1)1.09 (0.66–1.78)0.81CC25 (13.2)13 (8.8)1.58 (0.71–3.62)0.27Group II Genotype Cases,n** = 1898 (%)^b^Controls,n = 2234 (%)**^b^ OR (95%CI)

p -value GG874 (46)995 (44.5)Ref.-GC768 (40.5)978 (43.8)0.89 (0.78–1.02)0.09CC198 (10.4)223 (10)1.01 (0.81–1.26)0.96Group III Genotype Cases,n** = 82 (%)^c^Controls,n = 268 (%)**^c^ OR (95%CI)

p -value GG57 (69.5)166 (61.9)Ref.-GC24 (29.3)100 (37.3)0.70 (0.39–1.23)0.23CC1 (1.2)2 (0.7)1.46 (0.02–28.4)1.00^a^2 (1%) cases and 6 (4%) controls failed to genotype^b^58 (3,1%) cases and 38 (1,7%) controls failed to genotype^c^comprise 41 patients from unselected breast cancer cases

Association of BRCA1 c.-20 + 101 C/G (rs799905) with clinical characteristics

The clinical characteristics were analyzed among unselected breast cancer cases from Group II. In this analysis the correlation of rs799905 genotypes with clinical and pathological features of breast cancers was assessed (Table 4). Results showed a higher frequency of GG than GC or CC genotypes (16.1% vs. 11.3% or 9.6% respectively) in patients with TNBC (triple negative BC), and CC than CG or GG genotypes (54.7% vs. 44.7% or 43.1% respectively) in patients with lymph node (LN) metastases. However, these observations were borderline statistically significant (p = 0.48 for TNBC and p = 0.46 for LN metastases). There was also a statistically significant association of rs799905 with presence of germline BRCA1 PV (p = 0.003). All correlations of rs799905 with BC characteristics are presented in Table 4.Table 4. Association of rs799905 with breast cancer characteristics in group of unselected casesCharacteristic****Genotypep**-value****GG ****GC ****CC Age of onset, years0.98 Mean5453.753.8 Median525253Age of onset below 50, n (%)0.62 Yes395 (45.2)344 (44.8)82 (41.4) No479 (54.8)424 (55.2)116 (58.6)Death, n (%)0.48 Yes146 (16.8)111 (14.6)30 (15.5) No723 (83.2)648 (85.4)163 (84.5) No data: 19Histopathological type, n (%)0.54 Ductal G1/G2276 (40.6)242 (40.9)72 (43.9) Ductal G3135 (19.9)100 (16.9)25 (15.2) Ductal Gx57 (8.4)50 (8.4)11 (6.7) Lobular87 (12.8)90 (15.2)31 (18.9) Medullar19 (2.8)18 (3)5 (3) Tubular10 (1.5)8 (1.3)1 (0.6) DCIS27 (4)18 (3)1 (0.6) Other69 (10.1)66 (11.1)18 (11) No data: 404TNBC, n (%)0.048 Yes79 (16.1)47 (11.3)11 (9.6) No411 (83.9)369 (88.7)103 (90.4) No data: 820Tumor size, n (%)0.71 <1cm71 (12.4)66 (13.2)15 (11.4) 1-1.9225 (39.3)210 (42.1)52 (39.4) 2-4.9250 (43.6)208 (41.7)61 (46.2) >527 (4.7)15 (3)4 (3) No data: 636Lymph nodes metastases, n (%)0.046 Yes248 (43.1)231 (44.7)76 (54.7) No328 (56.9)286 (55.3)63 (45.3) No data: 608Bilateral BC, n (%)0.47 Yes34 (5.2)22 (3.8)6 (4.1) No615 (94.8)556 (96.2)141 (95.1) No data: 466BC/OC among relatives, n (%)0.24 Yes36 (4.1)21 (2.7)7 (3.5) No838 (95.9)747 (97.3)191 (96.5)Germline BRCA1 PV, n (%)0.003 Yes30 (3.4)10 (1.3)1 (0.5) No844 (96.6)758 (98.7)197 (99.5)DCIS Ductal carcinoma in situ, TNBC Triple-negative breast cancer, BC Breast cancer, OC Ovarian cancer, PV Pathogenic variant

Discussion

Although DNA methylation has been investigated for many years, its precise molecular mechanism remain unclear. Various factors, including genetic predisposition and random environmental influence, appear to contribute to the methylation of multiple genes across different tissues and at various stages of life. This study aimed to explore and analyze the genetic basis of BRCA1 promoter methylation in peripheral blood DNA and its potential contribution to breast carcinogenesis.

It has been suggested that gene methylation resulting from germline alterations may contribute to a subset of inherited cancers in families without a detectable predisposing PV [7, 8]. This theory assumes that higher (~ 50%) level of (constitutional) DNA methylation detected in patients would indicate the modification of a single allele, likely caused by cis-acting germline variant localized in promoter region of the gene.

For our study, we selected group of 100 breast cancer patients who had previously been analyzed by MS-HRM and showed BRCA1 promoter methylation in peripheral blood and paired tumor tissue DNA [15]. In this group, we investigated DNA methylation using quantitative pyrosequencing to identify cases with possible allelic methylation of BRCA1 promoter. However, in our study group of 100 breast cancer cases, the detected methylation levels ranged from 4 to 25% and were lower than the expected ~ 50% (corresponding to allelic methylation of BRCA1 gene). Nonetheless, based on the combined results from pyrosequencing (quantitative methylation) and immunohistochemistry (BRCA1 protein expression), we sought to identify germline alterations that might be associated with BRCA1 promoter methylation and correlate with breast cancer risk. In 47 out of 100 tested patients with highest methylation level (> 14%) and absent or low BRCA1 expression in tumor tissue (< 10% cells with immunoreactivity), we searched for germline alterations in the BRCA1 promoter region. Sequencing of 643 bp fragment encompassing promoter of BRCA1 gene identified a single intronic variant in BRCA1 c.-20 + 101 C > G, located between alternative exons 1α and 1β of the gene. This variant was reported in NCBI databases (rs799905) and assumed to be clinically benign, primarily based on the minor allele frequency (MAF) of 28% in the European population.

In our study, rs799905 was detected in 18 out of 47 (38%) BC patients with detectable methylation and low expression of BRCA1; subsequently it was tested in 3 study groups to evaluate a possible association with breast cancer. The prevalence of GG, GC and CC genotypes of rs799905 between BC cases and controls in Group I (with detected methylation of BRCA1 in peripheral blood) and Group II (unselected) was similar and was not associated with breast cancer risk (Table 3). However, analysis of clinicopathological features of breast cancer patients from Group II showed a correlation of rs799905 with TNBC and lymph nodes metastases (Table 4). Both observations were of borderline significance (p = 0.048 and p = 0.046, respectively) and require further investigation. Interestingly, we found a relatively strong and statistically significant association of rs799905 with BRCA1 germline PV (p = 0.003) (Table 4). The prevalence of GG, GC and CC genotypes in carriers of BRCA1 germline PVs (Group III) differed substantially from Group I and II, showing 10-times lower frequency of CC genotype (Table 3).

Variant rs799905 has been previously studied mostly in the context of allele specific BRCA1 methylation and used to distinguish methylated and unmethylated alleles in heterozygotes [20–23]. Only one study performed in the Jordanian population analyzed the correlation of rs799905 with prognostic factors and clinical characteristics of breast cancer. However, their study group was relatively small (230 BC patients) and the results, including association of rs799905 with LN metastases and TNBC, were not statistically significant [24].

It has been suggested that rs799905 may be associated with BRCA1 expression [25]. Variant BRCA1 c.-20 + 101 C > G is located within a CpG site and the presence of the G allele creates a potential methylation site. Depending on the methylation status of this site, it may influence BRCA1 gene expression due to its proximity to transcription factor binding sites. Furthermore, it has been demonstrated that specific alleles of rs799905 are consistently linked with a second common variant rs8176071 (TGTT > T) located 818 bp upstream of rs799905. It was shown, that the C allele of rs799905 is associated with the GTT deletion of rs8176071, while the G allele is linked with the presence of GTT. Both variants appear to counterbalance each other, maintaining stable BRCA1 expression. Notably, alternative combinations of these polymorphisms (allele C/TGTT or G/T(delGTT) was reported to induce in cell lines a significant reduction in BRCA1 expression [25]. The rs8176071 was not included in our study, and we could not confirm these observations.

Results of our analyses suggest that variant rs799905 is not directly associated with an increased risk of breast cancer or BRCA1 promoter methylation, which is generally in line with previously published reports. However, we observed in our study a significant and novel correlation between rs799905 and germline BRCA1 PVs that remains to be explained. In previous studies it was shown, that methylation of BRCA1 gene is restricted to non-carriers of BRCA1 PVs, suggesting that the mutation and methylation of BRCA1 may be mutually exclusive [26, 27]. It was suggested that the G allele of rs799905 may create a CpG - a potential methylation site. However, we observed in our analysis a higher frequency of the G allele in BRCA1 carriers (Tables 3 and 4) that appears to contradict this notion, and suggests that this site probably remains unmethylated. Notably, no prior research has focused on the possible association between rs799905 and BRCA1 PV, and at present, we are unable to explain the underlying mechanism observed in our study between the association of rs799905 and germline BRCA1 PVs. Further studies, including larger groups of BRCA1 PV carriers are necessary to investigate and understand the connection between rs799905 and BRCA1 PVs.

Conclusions

In this study, we aimed to analyze the BRCA1 promoter region to identify germline alterations associated with BRCA1 methylation in peripheral blood DNA. We did not find a genetic variant which could correlate with a BRCA1 epimutation in the Polish population, and be used as a genetic marker for the identification of a breast cancer predisposition. Sequencing of 5’UTR of BRCA1 gene in pre-selected breast cancer patients revealed a common variant BRCA1 c.20 + 101 C > G (rs799905) located within a CpG site, of which the G allele potentially might create a methylation site. Association analyses did not reveal any correlation between rs799905 and BRCA1 promoter methylation and breast cancer risk. Nevertheless, this variant may be associated with certain clinical features of breast cancer, though further research is required to confirm these findings. In addition, rs799905 may be associated to some extent with BRCA1 PVs, but the nature and mechanism of this co-occurrence remains unclear. Further research is needed to clarify correlation and its potential biological significance.