Introduction

Breast cancer is the major cause of cancer-related mortality among women [1]. Double-Stand Breaks (DSB) are the most severe type of damage. Homologues Recombination pathway (HR) is an extensively regulated process that is involved in the repair of DSB. Deficiencies in DSB repair increase the risk of breast cancer [2]. ATM (Ataxia telangiectasia mutated), BRCA1 (Breast cancer 1), RAD51 are three key proteins involved in the control of HR pathway. ATM is activated in response to DSB and triggers cellular signaling pathways [3]. Once activated, ATM in turn activates its target genes by phosphorylation. BRCA1 is one of these target genes [4].

BRCA1 takes part in HR and Non Homologues End Joining (NHEJ) repair. BRCA1 deficient cells are inefficient in repairing DNA damage by HR [5] and the most used pathway in these cells is NHEJ, which is an error-prone repair pathway [6]. Association of BRCA1 with RAD51 provides the first evidence that BRCA1 participated in DNA repair [7].

RAD51 is a central recombinase in HR that in preparation for HR, forms nucleofilament on ssDNAFootnote 1 which then promotes strand invasion and homologous pairing between two DNA duplexes [8]. It has been reported there is a reduced expression of BRCA1 and ATM in breast cancer [6, 8, 27]. However, albeit the important role of ATM as a sensor of DSB and initiator HR repair, there is no investigative report examining the correlation of expression changes of ATM with BRCA1 and RAD51 expression.

Human BRCA1protein shows general cytoplasmic and nuclear expression. RAD51 has high nuclear expression in proliferating cells and low nuclear and cytoplasmic expression in most other cells. ATM protein is predicted to have nuclear localization [9].

In this study our purpose was to investigate the expression of these genes in breast cancer using Real Time PCR and examine whether there was a correlation among the expression changes of ATM/BRCA1/RAD51 in sporadic breast cancer. In addition, we aimed to assess the association of ATM, BRCA1, and RAD51 gene expression with prognostic factors.

Materials and methods

Tissue collection

Breast tumor and adjacent normal tissues were removed surgically from 50 women patients admitted to the Khatamolanbia Hospital and exposed to a mastectomy from 2011 to 2015. Fresh tumor and normal adjacent ones at the margin of the tumors (2–3 cm distance) containing normal mammary gland tissues were collected by the clinicians in separate sterile tubes. Tissue samples were frozen and stored at −70 °C. Two pathologists confirmed the cancerous tumors and normal tissues. Staging of the breast cancer was performed according to the Union for International Cancer Control (UICC) which is based on (AJCC-TNM) classification.

Primers and PCR consumables

The cDNA sequences of ATM, BRCA1, RAD51 and β-Actin genes were obtained from Gene Bank. After identifying the exon/intron junctions, upper primer was selected in exon/intron boundaries. The selected sequences were evaluated by OLIGO 6 for hairpin and duplex formation stability. Lower primers were selected using Primer3 program (http://frodo.wi.mit.edu/primer3/primer3_code.html). BLASTN searches against dbEST and nr were conducted to approve gene specificity of the primer sequences and the absence of DNA polymorphisms (Table 1).

Table 1 Primers used for RT-PCR amplification ATM, BRCA1, RAD51 and β-actin
Full size table

RNA Extraction and cDNA synthesis

TriPure Isolation Reagent (Roche Applied Sciences) was used for total RNA extraction of breast tissue samples. Electrophoresis, through agarose gels and ethidium bromide staining, was used to determine the quality of the RNA samples. The concentration of RNA was measured by Nano Drop spectrophotometer. After synchronizing all samples, 1 µg of RNA from each sample was used to synthesize cDNA using First Strand cDNA Synthesis Kit, Fermentas, USA.

Real time PCR

The Relative quantification of ATM, BRCA1, RAD51 and β-Actin transcripts were carried out in samples using Light Cycler TM system (Rotor gene, Corbett, Germany) and Fast-Start DNA Master SYBR-Green I kit (Roche Germany) with specific primer. β-Actin was used as an internal control. The PCR was performed in 10 µL of reaction mix, containing 2 µL of Master Solution, 0.3 µM of each primer and 2 µL of cDNA as a template which was placed into 0.1 vials. Thermal cycling consisted of an initial denaturation step, 95 °C for 5 min, followed by an amplification program. Cycling conditions were 95 °C for 5 min, 45cycles with 3 step (1) 95 °C for 10 s (2) 62 °C for 15 s for ATM and RAD51 and 61 °C for 15 s for BRCA1 (3) 72 °C for 15 s. To evaluate the efficiency of each reaction precisely, we use Linreg software.

Statistic analysis

Real Time PCR data analysis was performed by REST 2009 software. All data are demonstrated as the mean ± standard error (SE) of triplicate experiments. The association between the changes in gene expression and main clinicopathological items was assessed by SPSS software (version 22.0) with Student’s t test and ANOVA. In the current study, P value less than 0.05 (P < 0.05) was considered statistically significant.

Result

Patients Characteristics

Pathology examinations showed that 12 (24%), 23 (46%) and 12 (24%) patients were related to grade of I, II, III, respectively. Thirty-two (62%) of the patients expressed ER and 29 (58%) of them expressed PR. In addition, 19 (38%) of the patients were HER2 positive (Table 2). According to pathological reports, patients were P53 and BRCA1 mutation negative. No one was exposed to neoadjuvant chemotherapy before surgery.

Table 2 ATM expression levels and clinicopathological and paraclinicalparameters in breast cancer sample
Full size table

PCR optimization and validation

Melting curve analysis showed only one peak for each reaction. Gel electrophoresis also showed a single band product with the desired length (Supplementary data, Fig S1).

ATM gene expression analysis in normal and tumor tissues

ATM gene expression analysis in normal and tumor tissues

Analysis of Real Time PCR results by REST 2009 revealed a significant reduction in gene expression level of ATM by 7.1 fold in tumor tissues (0.1709 ± 0.0276) in comparison with corresponding adjacent normal samples (1.2240 ± 0.1386) (Fig. 1). One way ANOVAs revealed a significant association between down expression of ATM and the stage of tumor. (P value < 0.05). Patients in stage I demonstrated the lowest mean expression (0.1045 ± 0.0209) in comparison with stage II (0.1982 ± 0.0210) and stage III (0.2081 ± 0.0366) (Fig. 2a; Table 2). In addition, there was a significant association between down expression of ATM and ER status (P value ≤ 0.01). Mean expression in ER negative group showed significant decrease (0.1046 ± 0.0130) when compared with ER positive group (0.2496 ± 0.0347). Furthermore, there was a significant association between decreased expression level of ATM and PR status (P value < 0.05). Mean expression in PR negative group (PR−) was lower (0.1078 ± 0.0229) when compared with PR positive (PR+) group (0.2507 ± 0.0237) (Fig. 2b; Table 2). There was also a positive significant association between down expression of ATM and necrosis (P value < 0.05). Mean expression in groups with tumor necrosis was 0.0515 ± 0.0040 and in group without necrosis was 0.2794 ± 0.0184 (Fig. 2c; Table 2). T test analysis showed the significant association between down expression of ATM gene with perineural invasion (P value < 0.05) (Fig. 2d) vascular invasion (P value < 0.01) (Fig. 2e), and malignancy (P value ≤ 0.001) (Fig. 2f; Table 2).

Fig. 1
figure 1

Real-time PCR analysis of ATM, BRCA1 and RAD51 expression in breast cancer tumors. Using the 2−∆∆CT method, the data are presented as the fold change in gene expression normalized to an endogenous reference gene (βeta actin) and relative to normal control (adjacent normal tissue). ***P value < 0.001

Full size image
Fig. 2
figure 2

ATM mean expression and clinicopathological factors. a ATM mean expression level and tumor stage. Samples were grouped according to pathological reports, b ATM mean expression in breast cancer subtypes. Patients were grouped considering IHC studies commonly used in clinical practice. Results are expressed as fold number decrease versus control (adjacent normal tissues). ER estrogen receptors, PR progesterone receptors, HER2 human epidermal growth factor 2, TN triple negative (ER−, PR−, and HER2−), TP triple positive (ER+, PR+, and HER2+). HR (hormone receptors: estrogen/progesterone receptors). *P value < 0.05, **P value < 0.01. c ATM mean expression level and Tumor Necrosis. *P value < 0.05. d ATM mean Expression level and perineural invasion. *P value < 0.05. e ATM mean expression level and vascular invasion. **P value < 0.01. f ATM mean expression level and malignancy. ***P value = 0.001

Full size image

BRCA1gene expression analysis in normal and tumor tissues

A remarkable decrease by 20 fold was observed in the expression level of BRCA1 in tumor tissues (0.1979 ± 0.0458) when compared with its adjacent normal ones (2.0416 ± 0.2324) (Fig. 1; Table 3). One way ANOVAs confirmed the significant negative association between down expression of the BRCA1 gene with Ki-67 expression status (P value ≤ 0.001).

Table 3 BRCA1 mean expression levels and clinicopathological and paraclinical parameters of breast cancer samples
Full size table

Mean expression in patients with >35% Ki67 was 0.0640 ± 0.0069 in comparison with group with <15% ki67 (0.2725 ± 0.0303), group with 15–25% Ki67 (0.2711 ± 0.0351), and 25–35% Ki67 (0.2159 ± 0.0127) (S2) (Table 3).

RAD51 gene expression analysis in normal and tumor tissues

RAD51 was markedly down regulated by 9.9 fold in tumor breast samples (0.2013 ± 0.030) in comparison with corresponding adjacent normal tissues (1.8113 ± 0.2093) (Fig. 1; Table 4). Moreover, there was a significant association between down expression of RAD51 and lymph node involvement (P value < 0.01). RAD51 gene expression level in group with more than 10 lymph node involvement was 0.1088 ± 0.0154 when compared with group with 4–9 lymph node involvement (0.1940 ± 0.0435), 1–3 lymph node involvement (0.2482 ± 0.0230) and without any lymph node involvement (0.2509 ± 0.0217) (Fig. 3a; Table 4). In addition, there was a significant association between PR (Progesterone Receptor) status (P value ≤ 0.01). Mean expression in patients with PR positive was (0.1042 ± 0.0120) when compared with PR negative group (0.2967 ± 0.0378) (Fig. 3b; Table 4). In addition, there was a significant association between low expression of RAD51 and a patient’s age (P value ≤ 0.001). Patients with mean age >55 demonstrated the lowest expression (0.1213 ± 0.0247) in comparison with mean age 45–55 (0.1960 ± 0.0295) and mean age under the 45 years old (0.2944 ± 0.0253) (Fig. 3c; Table 4). There was a significant association between down expression of RAD51 gene and pathological tumor grade. (P value < 0.05). Mean expression in grade III tumor was 0.1252 ± 0.0324 in comparison with grade II (0.1910 ± 0.0208) and grade I (0.2893 ± 0.0491). (Figure 3d; Table 4).

Table 4 RAD51 mean expression levels and clinicopathological and paraclinical parameters of breast cancer samples
Full size table
Fig. 3
figure 3

RAD51 mean expression and clinicopathological factors. a RAD51 mean expression level and lymph node involvement. **P < 0.01. b RAD51 mean expression in breast cancer subtypes. Patients were grouped considering IHC studies commonly used in clinical practice. Results are expressed as fold number decrease versus control (adjacent normal tissues). ER estrogen receptors, PR progesterone receptors, HER2 human epidermal growth factor2, TN triple negative (ER−, PR−, HER2−). TP triple positive (ER+, PR+, HER2+). HR (hormone receptors: estrogen/progesterone receptors).*P < 0.05. c) RAD51 mean expression level and age. **P < 0.01. d RAD51 mean expression level and tumor grade. Patients were grouped according to pathological reports. *P < 0.05. e RAD51 mean expression level and auxiliary lymph node metastasis. *P value < 0.05

Full size image

There was a significant association between auxiliary lymph node metastasis and decreased expression level of RAD51. (P ≤ 0.01). Mean expression in group that showed metastasis was 0.1373 ± 0.0116 and in group without metastasis was 0.2763 ± 0.0379. (Figure 3e; Table 4).

Correlation among ATM, BRCA1 and RAD51 expression levels in breast cancer patients

The 2−∆∆CT results for ATM, BRCA1 and RAD51 in tumor tissues were statistically analyzed for each two gene (ATM/BRCA1, ATM/RAD51 and RAD51/BRCA1) by Spearman’s correlation test. The results indicated there was a significant positive correlation among ATM/BRCA1 (r = 0.641, P value < 0.0001), BRCA1/RAD51 (r = 0.764, P value < 0.0001) and ATM/RAD51 (r = 0.619, P value < 0.0001) expression levels (Supplementary data, S3).

Correlation between ATM, BRCA1, and RAD51 expression and progress-free survival (PFS)

Clinical outcome of only 42 patients were available. Among them, only one passed away about 3 months after initial diagnosis. Two had metastasis: one to liver, one to lung. The rest are alive with no sign of recurrence of the disease or metastasis. Although, the expression levels of these three genes were low in these three patients, the number of the patients is too low to come to any conclusion regarding the correlation between the expression of these genes and PFS.

Discussion

Repair of DSBs is necessary to ensure genome integrity and cell viability [10]. ATM seems to be crucial as primary DSB sensor proteins. Activation of these proteins leads to posttranslational modification of downstream mediators, e.g. BRCA1. The mediators then reinforce the signal and trigger a signaling pathway that activates effectors proteins such as RAD51, P53 and checkpoint proteins which carry out cell cycle regulation, DNA repair and apoptosis [5].

Although some published reports indicated down expression of ATM in breast cancer [11, 12], some recent studies reported controversial results. Much higher expression level of ATM in ER-negative breast cancer was reported in one research [13]. In another study, it showed ATM expression is aberrantly reduced or lost in ER/PR/HER-2 negative breast cancers [12]. In two different previous research, no associations were found between low levels of ATM and different subclasses of breast cancer [14]. In addition, overexpression of ATM has been shown in myoepithelial carcinoma, prostate cancer and nasopharyngeal carcinoma [15,16,17]. Moreover, the association between ATM deregulation and prognostic factors in prognosis of breast cancer patients is limited.

In this study, our result showed decreased expression of ATM transcripts by 7.1 fold in tumor tissues in comparison with its normal adjacent ones (P value ≤ 0.0001). For the first time, we found a significant association between down expression of the ATM gene with vascular invasion (P value < 0.01), perineural invasion (P value < 0.05 and malignancy (P value ≤ 0.001) that indicate down expression of ATM correlates with aggressive behavior of tumors. In addition, for the first time a significant association has been found between ATM down-expression and tumor stage. Reduced expression of this gene would be due to LOHFootnote 2 [18, 19], aberrant methylation of promoter [2, 20] and post transcriptional dysregulation such as overexpression of regulating miRNA [11].

Genetic analyses indicated that BRCA1 is critical for HR pathway as well as the subnuclear assembly of RAD51 after DNA damage [21]. We have found a significant reduction in the expression of BRCA1 by 30 fold in breast tumors (P value ≤ 0.001). That is in line with other previous published report [22]. We found a significant association between down expression of the BRCA1 gene and higher expression of Ki-67 (P ≤ 0.001). Patients in group 4 (>35% proliferation rate) showed significant down expression, corroborating that BRCA1 deregulations plays a role in cell proliferation. That is in agreement with a previous paper which found a negative correlation between the expression of BRCA1 and Ki-67 [23].

Multiple mechanisms underlying inactivation and reduced expression of BRCA1 include (a) loss of heterozygosity (LOH) [2, 8] (b) methylation of the BRCA1 promoter region [24] (c) overexpression of some microRNA (miR-146a, miR-146b and miR-342 [24, 25]. (d) Mutations in transcription factors regulating the BRCA1 promoter (2). (e) Alterations in the signaling pathways upstream of the transcription factors [2].

RAD51 is a crucial recombinase, often dysregulated in tumors [26]. While several studies have reported overexpression of RAD51 in cell lines and tumors of different origins [26, 27] one study indicated down-regulation of RAD51 expression in different cancer cell lines grown under chronic hypoxic conditions [28]. The exact cause of overexpression is not clarified. However, it has been suggested TP53 deletions and some TP53 point mutations up regulate the expression of RAD51 [27, 28].

Although just one clinic research showed reduced expression of RAD51 in 30% of breast carcinoma at protein level, for the first time we are able to report down expression of RAD51 in breast cancer tissues at the mRNA level. The change rate was significant by 9.9 fold in tumor samples when compared with its adjacent normal ones. Our results are in contrast with two other published reports in which overexpression of RAD51 was observed. In one of these reports, patients were P53 mutation positive and were exposed to neoadjuvant chemotherapy before surgery [5, 29], two options that are suggested to underlie RAD51 overexpression. Patients in the other research have not above mentioned two options [30]. LOH is one of the mechanisms that leads to down expression of RAD51 in breast cancer [2]. Moreover, overexpression miRNA could lead to RAD51 down expression [31]. The association between RAD51 gene expression level and clinicopathological factors is limited. In this study, a significant association was clarified between down expression of RAD51 and clinical and preclinical parameters, such as lymph nodes involvement, auxiliary lymph node metastasis, age, grade and PR status. In this research, samples with PR positive showed significant decrease in comparison with samples with PR negative.

Finally, our data indicated a significant positive correlation among ATM/BRCA1 (r = 0.641, P < 0.0001), BRCA1/RAD51 (r = 0.764, P < 0.0001) and ATM/RAD51 (r = 0.619, P < 0.0001) expression levels.

It’s ostensible that targeted consideration of gene expression variation in specific cell pathways can exhibit some unknown biological properties that are not appeared in single gene changes. Although some studies reported down expression of ATM, BRCA1 [6, 8, 24]. However, albeit the major role of ATM in HR pathway there is no clinical research examining the correlation of expression changes of ATM with BRCA1 and RAD51 expression. We found significant correlation between down expression of these genes analyzed via SPSS software version 22. These results supported the hypothesis that upstream genes can regulate other downstream genes in the same pathway or genes in the same pathway, are regulated by same regulator through related mechanisms in a coordinated way.

(I) Downstream proteins are regulated by upstream ones. There are some interactions between BRCA1, ATM and RAD51 in HR repair such that ATM regulates BRCA1 [4] and BRCA1 regulates RAD51 [21]. ATM acts as the upstream sensor and in mammalian cells it is necessary for the initiation of a signaling pathway. Following DSB formation, ATM phosphorylates BRCA1 and TP53 to promote DSB repair and cell cycle regulation [4]. BRCA1 is a substrate of ATM in vitro and in vivo. It has been revealed that a part of the cellular response to DNA damage BRCA1 is regulated by an ATM dependent mechanism [4]. On the other hand, the major mechanism underlying down expression of BRCA1 in breast cancer is not clear. Methylation appears to be a significant factor in BRCA1 regulation only in a small proportion of breast tumors [32]. Since, LOH associated with low expression of BRCA1 in the minority of cases, it is an inadequate explanation as a cause for reduced expression of BRCA1 in breast carcinoma [33]. Previous research confirmed that BRCA1 regulate and activate RAD51 [21]. Therefore, down expression of ATM can be a cause for low expression of BRCA1 and RAD51 genes that act in concert downstream of ATM in the same pathway in breast cancer. In this hypothesis, down-expression ATM would underlie down expression of BRCA1 and down expression of the last gene leads to down expression of RAD51.

(II) Expression of these genes is regulated by related mechanisms. A same transcription factor regulates transcription of these genes. For example, E2F1 can elevate the expression of genes BRCA1 [34], RAD51 [28, 35] and ATM [36]. In addition, Epithermal growth factor receptor (EGFR) promotes DSB repair by interacting with ATM, RAD51 and BRCA1 [37] and regulate expression of these genes. Moreover, polo- like kinase 1 regulates transcription of both BRCA1 and RAD51 [38]. This raises the possibility co regulation of these genes, with the same transcription factors, can lead to down expression of these three genes in breast cancer. So up and down expression of regulators can underlie the expression all of target downstream genes. For future perspective it would very valuable to examine the whole tumorigenic pathway to achieve better insight into the molecular changes involved in breast cancer development and progression.

It has to be noted that in this study we only investigated the gene expression at mRNAs levels and not protein levels. This transcription level data can suggest that these proteins are probably present more in normal breast tissues compared to breast tumoral tissues roughly and at what level to expect to see these proteins. As a future work one might investigate the expression level of these genes at their protein levels.