Introduction

Breast cancer (BC) is a highly heterogeneous disease with many different histological, molecular, and clinical subtypes. The molecular heterogeneity is observed at the genetic, epigenetic, and transcriptomic levels. This, combined with the high level of morbidity, makes BC an interesting subject of scientific studies focused on understanding the causes and mechanisms of BC development and progression [1,2,3].

Each BC is unique and differs at the molecular level. Thus, the classification of BC based on histological and immunohistochemical features does not reflect the complex genetic changes underlying tumour growth and development [4].

Gene expression analyses allow for identifying molecular BC subtypes, thus, enabling tailored therapy. Moreover, currently, available multigene prognostic panels, such as Oncotype Dx and MammaPrint, rely on the expression of various genes in the tumour tissue and, thus, differ in the range of prognostic and/or predictive capabilities. Therefore, developing new tests based on wide genetic analyses and introducing them into clinical practice are a promising area in the diagnosis and treatment of this cancer [5, 6].

Maintaining genetic stability is extremely difficult and is enabled by constantly active DNA-repair systems [7, 8]. An accumulation of DNA mutations resulting from ineffective DNA repair leads to genome instability, a key reason for cancer development and progression. Therefore, mutations in DNA-repair genes are frequent targets for tailored therapies [8, 9].

It has been revealed that in BC patients with BRCA1/2 mutation, the effectiveness of poly(ADP-ribose) polymerase (PARP) inhibitors contributes to the extension of disease-free survival. PARP1, an enzyme involved in DNA repair, participates in the correction of DNA single-strand breaks (SSBs), which, if unrepaired, may convert into double-stranded DNA breaks (DSBs) during DNA replication. DNA replication-derived DSBs result in chromosome breakages and translocations, leading to severe genome instability [10]. The pivotal mechanism in repairing DSBs is the homologous recombination (HR) leading to the replacement of DSBs with an undamaged sequence [11,12,13,14]. Inactivating mutations in HR genes leads to the accumulation of other mutations in the genome, causing changes that promote carcinogenesis. The therapy of HR-deficient BC patients with PARP1 inhibitors (PARPi) causing synthetic lethality results in deleterious accumulation of mutations and, thereafter, apoptosis [15, 16]. Thus, PARP inhibitors are effective in the therapy of patients with BRCA1/2mut breast cancer and homologous recombination deficiency (HRD) [5, 15, 16].

Many studies confirm that BRCA-dependent carcinogenesis is related to HR system inactivation [11, 17, 18]. Other HR genes involved in BC development include PALB2, BARD1, TP53, BRIP1, and RAD51C [19]. Previous research revealed that mutations in the above-mentioned genes significantly decrease DSB repair activity [20].

Our study aims to evaluate the alterations in the expression/transcript level of the following HR genes: BRCA1, BRCA2, ATM, BARD1, FANCA, FANCB, FANCI, RAD50, RAD51D, BRIP1, and CHEK2. As a result, we aim to develop a potential prognostic marker in patients diagnosed with BRCA-positive BC.

Material

Breast cancer tissue samples were secured by the core needle biopsy before the patient's systemic treatment to assess the expression level of selected HR genes in tumour cells.

Core needle and vacuum-assisted breast biopsy (CNB and VABB) samples from 45 BC patients were collected and preserved in formalin-fixed paraffin-embedded (FFPE) blocks for subsequent RNA extraction and analysis. A pathologist selected the most representative BC tissue sections containing approximately 50% of cancer cells (macrodissection). All samples were taken before the chemotherapy. Later, all patients received treatment based on standard chemotherapy regimens (anthracyclines, taxanes).

In all patients, partial (PR) or complete pathological response (CR) was observed after neoadjuvant therapy (11 PR vs 33 CR). The complete pathological response was defined as the disappearance of all invasive cancer tissue in the resected breast specimen and in all sampled regional lymph nodes after the completion of neoadjuvant chemotherapy.

All patients were referred for a consultation with a clinical geneticist regarding hereditary breast and ovarian cancer predisposition. In the case of the germline BRCA1/2 pathogenic or likely pathogenic variant, consultation with a clinical geneticist and the appropriate molecular test were also offered to the patient’s adult close relatives.

Ethics statement

All patients signed informed consent. The study was approved by the Ethics Committee of Wroclaw Medical University (Nos. 611/2019 and 65/2023). All patients were diagnosed and treated in the Breast Unit, Lower Silesian Oncology, Pulmonology, and Hematology Center, Wroclaw, Poland. All samples were taken as part of the patient’s diagnostic and therapeutic scheme. All procedures performed in this study followed the principles for medical research of the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Methods

The pathogenic BRCA1 or BRCA2 germinal variants were assessed from blood using NGS BRCA1/2 panel testing (Devyser BRCA NGS CE-IVD, MiSeq Dx Illumina). Polish recurrent pathogenic variants were estimated/assessed using a laboratory-developed PCR screening test, whose results were confirmed by Sanger sequencing (BigDye™ Terminator v3.1 Cycle Sequencing Kit, ThermoFisher Scientific; Termocykler C1000 Touch Thermal Cycler, BioRad; ABI 3500 Series Genetic Analyzer, ThermoFisher Scientific).

RNA extraction

Total RNA was extracted from breast cancer FFPE tissue sections using a Qiagen® RNeasy FFPE Mini Kit according to the manufacturer’s instructions. The RNA quality was determined using a NanoPhotometer N60 (Implen). The samples with an RNA concentration of ≥ 60 ng/µl, an A260/A280 ratio of 1.8 ~ 2.0, and an A260/A230 ratio of 2 ~ 2.2 were accepted for the analysis.

PCR array

The Qiagen® Custom RT2 Profiler PCR Array was used to analyse the expression level of 11 HR genes (BRCA1, BRCA2, ATM, BARD1, FANCA, FANCB, FANCI, RAD50, RAD51D, BRIP1, and CHEK2) and two or three housekeeping genes (GAPDH and B2M, and for BRCA1/2 additionally ACTB). All genes were analysed in triplicates with respective housekeeping genes and controls included in the PCR plate according to the producer protocol using the real‐time cycler Bio-Rad CFX96 (Fig. 1).

Fig. 1
figure 1

Representations of real-time PCR amplification curves. Three replicates of the amplification reaction for each sample are shown either on a linear scale (left) or on a semi-log scale (right). Figure 1A The plot for BRCA1 mRNA expression in BRCA1-mutated tissue (Cq: 31.31/31.55/31.39); Fig. 1B The plot for BRCA1 mRNA expression in BRCA2-mutated tissue (Cq: 30.15/30.19/30.52); Fig. 1C The plot for BRCA1 mRNA expression in BRCA1/2-unmutated tissue (Cq: 36.65/34.89/35.82)

Full size image

Statistical analysis

Data analysis was performed using R/Bioconductor environment and HTqPCR and compareGroups packages [PMID: 19,808,880; https://www.jstatsoft.org/article/view/v057i12]. Raw data were normalised using the B2M gene as a housekeeping reference. Data distribution of continuous variables was assessed using the Shapiro–Wilks test. Depending on whether the variable was considered as continuously normal-distributed, continuous non-normal-distributed, or categorical t-test, the Kruskal–Wallis test or chi-squared test were used, respectively. The significance level was set at 0.05.

Results

The study group consisted of 45 women aged 35–76 years (average age 53 years), including 24 women with a confirmed diagnosis of breast cancer and BRCA1/2 germline mutation and 21 women diagnosed with invasive breast cancer and no BRCA1/2 germline mutations.

All patients received systemic neoadjuvant therapy. The receptor status (ER, PR, and HER2), Ki67 status, adjuvant therapy response, and histopathological classification are presented in Table 1

Table 1 Clinical data of the study group (24 patients with BRCA1/2 mutations)
Full size table

No significant difference has been found for ATM, BARD1, FANCA, FANCB, FANCI, RAD50, RAD51D, BRIP1, and CHEK2 expression in BRCA1/2mut cancer tissues as compared to BRCA1/2wt samples (Table 2, Fig. 2).

Table 2 Clinical data of the study group (21 patients without BRCA1/2 mutations)
Full size table
Fig. 2
figure 2

Box plots representing expression levels of selected HR genes considering germline BRCA1/2 mutation status

Full size image

The only statistically significant difference was observed for the BRCA1 gene expression. In BRCA1mut cancer tissues, the BRCA1 expression level was significantly higher than in BRCA2mut and BRCA1/2wt cancer tissues (Fig. 1). Apart from this observation, we found that BRCA1mut cancer tissues displayed significantly higher expression of Ki67 and a significantly higher proportion of ER and PR negative cases. Moreover, patients with BRCA1mut responded better to the applied therapy (higher proportion of complete response) than BRCA2mut and/or BRCA1/2wt patients (Table 3).

Table 3 Summary descriptives table by groups of BRCA1/2 status
Full size table

Discussion

Transcriptome profiling may be a powerful and effective method for searching for new molecular biomarkers of cancer prognosis, risk of progression, cancer-free survival, and other clinical features. Recent studies confirmed that a high homologous recombination deficiency (HRD) score is associated with poor survival among BC, prostate cancer, glioma, and head and neck squamous cell carcinoma (HNSCC) patients [21,22,23,24,25]. Hence, HRD, resulting from the loss of function of HR genes, including BRCA1/2, has been approved as an independent predictive biomarker of sensitivity to PARPi therapy [15, 16, 26]. Moreover, recent studies suggest a potential predictive value of HRD status in platinum-based chemotherapy in breast [27] and ovarian cancer (OC) patients [28].

Our study assessed the expression of the selected main HR pathway genes in cancer tissue to check whether their mRNA levels are altered in the BRCA1/2mut BC. The rationale for this research was based on the extensive interconnections among HR proteins, where deregulation of a single but critical gene’s expression could potentially impact the expression of others.

The only statistically significant differences were observed for BRCA1, as its mRNA level was elevated in BRCA1-mutated tissues compared to BRCA2-mutated and BRCA1/2 wild-type tissues.

Our results are consistent with those published by Wang et al., who proved that BRCA1 and BRCA2 gene expression is upregulated in breast and ovarian cancer (OC) tissues. Moreover, these authors observed an increased expression of NF1 and SYCP2 genes, interacting with BRCA1/2 genes in the regulation of the cell cycle. Therefore, they suggested the importance of functional interrelations among the BRCA1/2 with the other genes involved in BC and OC development and progression, thus, influencing the clinical course of disease and treatment outcomes [29].

In their recent study on 38 ovarian cancer vs 11 fallopian tube tissues, Custódio et al. showed that BRCA1/2 mRNA expression varied between individual samples. Moreover, in tissues characterised by downregulated BRCA1/2 expression, the other 12 genes involved in the HR pathway also exhibited low mRNA levels. The analysis of 299 ovarian cancer samples from The Cancer Genome Atlas (TCGA) confirmed these findings [30]. It is important to emphasise that our study’s findings on BRCAmut breast cancer tissues differ, as we observed elevated mRNA levels for BRCA1. This discrepancy may arise from the varying schemes and methodologies employed in the studies: 1) the cancer type (BC vs OC) or 2) different HR genes studied (ERCC1, ERCC2, ERCC3, ERCC4, ERCC5, ERCC6, ERCC8, EXO1, FAN1, FANCA, FANCB, FANCC vs ATM, BARD1, FANCA, FANCB, FANCI, RAD50, RAD51D, BRIP1, CHEK2), 3) different BRCA1/2 mutational status (17/40 (42.5%) samples with BRCA1/2 pathogenic or likely pathogenic variant vs 24/45 (53%)), and finally, the most importantly 4) different laboratory methods (very sensitive and accurate NanoString Technology and droplet digital PCR (ddPCR) vs real-time PCR).

Another study on 96 fresh frozen ovarian cancer tissues obtained from chemotherapy-naïve patients and patients after neoadjuvant chemotherapy was performed using a tailored NanoString-based Pancancer Pathway Panel of 19 HR genes and showed a correlation between over-expression of C11osf30, NBN, FANCF, FANCC, FANCB, RAD50 and improved outcome in chemotherapy-naïve patients. Moreover, a correlation has been observed between over-expression of BRCA2, TP53, FANCB, and RAD51 and worse outcomes in chemotherapy-treated patients. When adding the extent of debulking as a covariate, the expression of NBN, FANCF, RAD50, and RAD51 was significant, respectively [31].

Also, in the bladder cancer cell lines, the expression of four DNA damage repair genes, including two MR genes, was evaluated respectively before and after chemotherapy. The authors of this study revealed that the increase of BRCA1 and RBBP8 expression induced by chemotherapy correlates with worse sensitivity to treatment in non-basal and non-luminal cell lines. In contrast, no significant differences in the basal-cell lines most sensitive to chemo and radiotherapy were observed. This observation revealed the high diversity of HR genes expression, which correlates with the histopathological characteristics of tumours [32]. Moreover, in 413 bladder cancer samples (data derived from TCGA), a significantly higher expression level of four genes, RAD21, RAD51, BARD1, and ERBB2, was observed in ERBB-low as compared to ERBB-high tumours. Also, the combined expression of two out of four tested genes has been shown to correlate with chosen clinical features. This confirms the interconnections among the expression of different HR pathway genes [33].

In sporadic gastric cancer patients after receiving postoperative adjuvant chemotherapy, BRCA1/BRCA2 expression assessed using IHC and mRNA tests displayed a correlation between BRCA2-elevated expression with advanced tumour stage but not disease-free and overall survival [34].

These findings indicate that the investigation should encompass not only the alterations in individual HR genes but also their interrelations, considering the clinical course of the disease alongside histopathological and biochemical variables characterising the studied cancer.

Conclusion

The mRNA expression of BRCA1 is upregulated in breast cancer tissues harbouring BRCA1 mutation. However, no relevant differences in the expression of other HR genes have been observed between BRCA1/BRCA2mut and non-mutated BC tissues. Hence, it appears that BRCAmut tissues do not exhibit crucial compensatory alterations in the mRNA expression of other HR genes.

Limitations

The prognostic value of HR gene expression in BC could not be assessed due to the limited sample size and the variability in histopathological classifications.