Background

Breast cancer (BC) is the second most common cancer in the world with approximately 209,000 new cases in 2018. Importantly, it is the leading cause of women cancer death [1].

It should be noted that high ratio of deaths occurs in low- and middle-income countries, where there are greater challenges in improving diagnosis and treatment due to limited resources [2, 3]. In contrast, developed countries have significantly reduced mortality rate due to the provision of appropriate facilities aiding early diagnosis such as regular screening programs with physical examination, accurate mammographic or ultrasound procedures and new effective treatment [4, 5, 6].

As a multifactorial disease, in addition to the environmental risk factors, the impact of genetic on the pathogenesis of breast cancer is well documented [7]. Age, anthropometric related properties, age of menarche and menopause, mammographic density, breast feeding, reproductive pattern and hormone replacement therapy are amongst known risk factors [8]. It is shown that mutations in few involved genes like BRCA1/2 are responsible for only 25% of family and inherited breast cancer patients while sporadic breast cancer risk modification in the general population is related to the underlying common genetic variations including SNPs [8,9,10]. Considering non-genetic factors, mammographic density is one of the independent, strong and well-confirmed risk factors for this cancer [11]. Mammographic density (MD) which has been categorized into 4 groups from A to D using the BI-RADS system, is a measure of the amount of radiopaque epithelial and stromal tissue or fibroglandular tissue as opposed to fat tissue in the breast [12]. Some studies have indicated that higher breast density is linked with the higher risk of breast cancer [11, 13, 14].

Breast density is a risk factor in which many SNPs are associated with this phenotype and thus cancer risk. The current knowledge about such association mainly comes from extensive genome-wide association studies (GWAS) [15,16,17,18,19]. Among several genetic loci examined in many GWAS for breast cancer, the association of the ESR1 (estrogen receptor1) gene locus has been shown with both breast cancer risk and mammographic density [20, 21]. The ESR1 gene has been identified as one of the sex hormones’ key receptors which also acts as a transcription factor for most essential genes and has a remarkable function in the secondary sexual traits development in women [22, 23]. The association of SNPs in 6q25.1 locus with breast density and/or cancer in various ethnicities has been studied with controversial results [24,25,26]. Therefore, a case–control study was designed to identify whether there is such an association with ESR1 locus SNPs in the Iranian population. Since there have been no published data on the association with rs6915267, rs2077647, rs1801132 polymorphisms and breast density, we aimed to assess these markers in association with breast density values and the risk of breast cancer.

Material and methods

Study population

The Ethical Committee of Mashhad University of Medical Sciences approved this study. 200 patients, who were referred to both Omid and Imam Reza Hospitals, Mashhad University of Medical Sciences, and 200 healthy controls who were referred to a medical center for screening and mammography, were recruited. The mammographic densities of cases and controls were determined by a radiologist. All participants signed a written consent form. Demographic data including age at diagnosis, weight, height, BMI, history of lactation and abortion, diet and history of screening were collected using questionnaires. The clinical information of all patients were obtained from the patients’ medical records. Tumor types of breast and mammographic density reports were in accordance with WHO and BI-RADS classifications, respectively.

Blood and DNA preparation

The salting out method was used for DNA extraction where 5 ml of peripheral blood was collected from all participants in EDTA-contained tubes. To ensure quality, the extracted DNA was loaded onto 1% agarose gel and the concentration was investigated by Epoch™ Microplate Spectrophotometer (BioTek Instruments Inc., Winooski, VT, USA,). Finally 100 ng per microliter concentration were prepared and stored at − 20° C temperature.

Genotyping

The genotyping of rs6915267 and rs2077647 and rs1801132, was performed using the amplification refractory mutation system (ARMS) method. The sequences of primers used for detection of SNPs alleles and PCR conditions have been shown in Table 1. The final PCR products separated by electrophoresis in 2% agarose gel.

Table 1 The sequences and other properties of used primers
Full size table

The standard form of ARMS-PCR for rs6915267 was performed in two 10 μl-reactions for wild-type and mutant alleles, separately. For the mutant allele this reaction contained 4 μl master mix (Taq 2x, Parstous), 1.2 μl of each forward and reverse primers (10 μM) and 2 μl (200 ng) genomic DNA. For wild-type allele the reaction containing 4 μl master mix, 1 μl of each forward and reverse primers (10 μM) and 2 μl (200 ng) genomic DNA.

Tetra ARMS-PCR amplification for rs2077647 was performed in a 10 μl reaction volume containing 5 μl master mix, 0.4 μl of each outer primers (10 μM), 0.8 μM of forward inner (10 μM), 1.4 μl (10 μM) of reverse inner (wild-type allele) and 2 μl (100 ng) genomic DNA.

Tetra ARMS-PCR amplification for rs1801132 was performed in a 10 μl reaction volume containing 5 μl master mix, 0.5 μl of each outer primers (10 μM), 1.3 μM of forward inner (10 μM), 1.2 μl (10 μM) of reverse inner (wild-type allele) and 1.5 μl (100 ng) genomic DNA.

Statistical analysis

The Pearson χ2 distribution with one degree of freedom was used to check any deviation of allelic frequencies from the Hardy–Weinberg equilibrium. The associations of breast density and cancer risk, risk factors and tumor characteristics with genotypes in different genetic analysis models were assessed using binary logistic regression. Association of independent variables with the breast density and cancer risk, reported as odds ratios (ORs) with 95% confidence intervals, was assessed using multivariate logistic regression.

Using the PHASE software version 2 and 2LD program version 1.00, haplotype frequencies and linkage disequilibrium (LD) were assessed respectively.

Other statistical analyses were done using SPSS 16.0 (IBM, USA). Significant P values were considered of values less than 0.05.

Results

Characteristics of the study population

The comparison of demographic data between the case and control groups plus the dense and non-dense groups has been listed in Table 2. Clinical characteristics of the disease have been summarized in Table 3.

Table 2 Demographic features in breast cancer/healthy groups and dense/ non-dense groups
Full size table
Table 3 Frequency distribution of tumor characteristics of Breast cancer cases
Full size table

Due to lack of access to medical records and mammograms of 22 patients, they were excluded from the study. The mean age was 46.79 ± 9.91 and 52.42 ± 8.96 years in breast cancer and healthy women, respectively, with a significant difference (p < 0.001).

Furthermore, age of menstruation (p = 0.019), age of first gestation (p = 0.027) and body mass index (BMI) (p = 0.002) indicated a significant difference between breast cancer patients and controls. Breast density was evaluated among the participants and it was found that the frequency of women with dense breast was higher in the patient group and as a result, there was a significant difference in mammographic density between cases and controls (p < 0.01).

The higher rate of positive history of breast self-exam and mammography (breast cancer screening) was observed in the control group (23.8%) than patients (11.9%) as a result, there was a statistical association between the two groups (p = 0.007).

The significantly higher frequency of no breastfeeding history was observed in the patients (8.8%) than healthy individuals (1.7%) (p = 0.007).

In this study, the women with class A and B mammographic density were categorized in non-dense breast group and women with class C and D breast density were placed in the dense group. A total of 216 individuals with non-dense breast (mean age of 52.35 ± 9.63) and 162 women with dense breast (mean age of46.36 ± 8.99) were included with a significant difference (p < 0.01). In addition, dense breast group had the lower age of menopause (p = 0.015).

cytopathological evaluation indicated 61.5% of patients were in early stages (stage I & II) and 54.6% of them had grade II. 84.1% of the breast tumors were invasive ductal carcinomas and the frequency of invasive lobular carcinoma was 4.1%. 48.1% of tumors sizes were between 20 to 50 mm. A large proportion of breast cancer women were hormone receptor positive/HER2 negative (78.2%) and 12.9% of cases were triple negative breast cancer (TNBC).

Association of polymorphisms with breast cancer, pathological characteristics and mammographic density

The genotype frequencies for rs6915267, rs2077647 and rs1801132 polymorphisms were in Hardy–Weinberg equilibrium (p > 0.05). Genotypes were confirmed by randomly regetyping of 40 samples for each polymorphisms.

There was no significant association between these three polymorphisms and breast cancer in any of the genetics models before and after adjustment (p > 0.05). The allele and genotype frequencies as well as the association of the rs6915267, rs2077647 and rs1801132 with the risk of breast cancer have been shown in Table 4.

Table 4 The genotype and allele frequencies of the rs6915267, rs2077647 and rs1801132 variants and their association with breast density between cases and healthy controls after adjustment
Full size table

Considering pathological factors, rs6915267 AG and GG genotypes with 50% and 35.4% frequencies in HER2 positive patients compared with HER2 negative cases, were statistically different and these two genotypes were associated with being negative of HER2 receptor ([p = 0.018; OR = 4.88, 95% CI (1.31–18.18) and p = 0.040; OR = 4.12, 95% CI (1.06–15.93), respectively]. Also, there was no association between genotypes of rs2077647 and rs1801132 and clinical features in patients (p > 0.05). Results have been shown in Supplementary file.

Evaluation of the ESR1 variations in dense and non-dense breasts indicated rs6915267 AA genotype was associated with an increase in breast density in additive model (AA vs. GA) [OR = 2.55, 95% CI (1.08–5.99), p = 0.032] as the significantly higher rate of AA genotype was observed in subjects with dense breast (11.7%) than in women with non-dense breast (5.1%).

Moreover, our results showed GA genotype in the co-dominant model (GA vs. GG + AA) had a higher frequency in the non-dense breast group (60.2%) and there was an association in this genetic model. These findings remained unchanged after adjustment for demographic factors [OR = 0.57, 95% CI (0.36–0.90), p = 0.015]. No significant association was observed between rs2077647 and rs1801132 and breast density in different genetics models before and after adjustment (p > 0.05). Results have been shown in Table 5.

Table 5 The genotype and allele frequencies of the rs6915267, rs2077647 and rs1801132 variants and their association with breast density between women with non-dense and dense breast after adjustment
Full size table

Association of haplotypes with the risk of breast cancer and breast density

Haplotype analysis resulted in seventeen haplotypes including twelve 2-SNPs and five 3-SNPs haplotypes. G-T rs6915267-rs2077647, G-C rs6915267-rs1801132, T-C rs2077647-rs1801132 and G-C–C rs6915267-rs2077647-rs1801132 haplotypes were found to be more frequent combinations in all groups (breast cancer- healthy and dense- non-dense groups).

Our results indicated the G-T-C haplotype was associated with a decreased risk of breast cancer [OR = 0.54, 95% CI (0.31–0.92), p = 0.025]. Moreover, G-T/G-T diplotype of rs6915267-rs2077647 was associated with a decreased risk of breast cancer in comparison with other types (p = 0.019). However, no statistically significant association was found between other haplotypes or diplotypes with breast cancer and density (p > 0.05). The results have been shown in Supplementary file.

Discussion

The impact of ESR1 gene disruption in various cancers including breast, prostate, liver and endometrial cancers has been investigated in several studies. ESR1 protein works as a transcription factor and it is responsible for transcription of cell-critical genes such as those involved in regulating proliferation. Given the remarkable role of genetic susceptibility in cancer development, researchers have concentrated on the evaluation of the association of sex hormones, particularly estrogen receptors, polymorphisms with breast cancer risk. However, there are only few studies on genetic predisposition for breast density in general population.

The present study has evaluated the association of three common variations (rs6915267, rs2077647 and rs1801132) in ESR1 rejoin and their haplotypes, with the breast density value and the risk of breast cancer in an Iranian population. To the best of our knowledge, these polymorphisms have been investigated for the first time in this population. We did not observe any significant association between rs6915267, rs2077647 and rs1801132 and breast cancer risk. However, breast density was less in heterozygote carries of rs6915267, suggesting that rs6915267 may have an independent effect on breast density. Furthermore, we did not find any statistically significant difference between two other polymorphisms (rs2077647 and rs1801132) and the breast cancer risk or breast density. Additionally, according to haplotype analysis, G-T-C haplotype of rs6915267, rs2077647 and rs1801132 and G-T/G-T diplotype of rs6915267-rs2077647 were identified to be associated with risk of breast cancer, suggesting the allelic interaction and combination may influence the risk.

Also, the average frequency of rs6915267 in the world is about 20%; though our results showed that the frequency of rs6915267 in our population was 35.5%. Unfortunately, the association of this polymorphism with breast cancer and/or density has not been studied in other populations so we could not compare our results with those of other populations.

In a breast cancer GWAS conducted by Zheng et al. among Chinese women, a significant association was found for rs2046210 at 6q25.1. According to this study, there was a similar association in an independent study conducted among people with European ancestry (p = 0.01) [19]. Such evidence compromises the importance of ESR1 gene region in breast cancer susceptibility. In another study in Chinese women, Han et al. suggested the association of 6q25.1 genetic locus with breast cancer risk.

A meta-analysis study conducted by Li et al. showed that rs3798577 was associated with a lower breast cancer risk in an Asian population, although in Caucasians it was contradictorily associated with a higher rate of the disease risk. rs2228480 has been reported as a protective allele in Caucasians with a large effect, while similar to our results, rs2077647 did not indicate association with breast cancer risk in both ancestries [20]. However, Son et al. reported rs2881766, rs2077647, rs926778, and rs2273206 ESR1 polymorphisms might increase but rs3798377 might decrease the risk in a Korean women [27]. Similarly, Hsiao et al. suggested that the rs2077647 has reversed association with breast cancer in Taiwanese women. As a result, they proposed the correlation between ESR1 genetic variants and breast cancer in their population [28].

On the other hand, Sasaki et al. identified that the SNP rs2077647 (S10S) had a protective effect for endometrial cancer [29]. Moreover, some studies indicated that this SNP is associated with increased risk for prostate and renal‐cell carcinomas [30, 31]. For rs1801132 polymorphism, a meta-analysis study was conducted by Sun et al. and they found rs1801132 and rs2077647 may be protective factors for cancer in Caucasians. Also, rs2077647 may be closely associated with hepatocellular carcinoma [32].

So far, there is an insufficient number of studies on the association of genetic variants of ESR1 gene with mammographic density. In a study conducted by Brand et al. among European ancestral women, two novel mammographic density polymorphisms (rs9485370 in TAB2 gene and rs60705924 in CCDC170/ESR1 region) were indicated at 6q25.1 genetic locus, both of which were identified as breast cancer associated SNPs. Consequently, they suggested the 6q25.1 was a susceptibility region in breast density and provided some evidence about the related mechanism [33].

In addition, in a GWAS study, Lindström et al. identified genome significant loci for dense areas in which some regions like ESR1 were known breast cancer susceptibility loci [34]. Mariapun et al. evaluated the association of 36 SNPs with the density value in Malaysian-Chinese women and found three near ESR1 polymorphisms were associated with breast cancer risk and/or mammographic density [35]. These findings highlight the possible effect of this locus on breast density and further cancer risk.

Conclusion

Although ESR1 rs6915267, rs2077647 and rs1801132 polymorphisms did not represent association with the risk of breast cancer as well as mammographic density measures in the studied population, combination effects of these markers may influence the risk as reflected in haplotypes and diplotypes of these variants. However, further larger studies and evaluation other genetic markers in this region along with considered polymorphisms can improve our knowledge about the genetic basis of breast cancer pathogenesis.