Introduction

Breast cancer is one of the most common cancers among women, accounting for 23 % of the total cancer burden and 14 % of cancer deaths worldwide [1]. It has become a great threat to public health and a major burden on the global economy. Therefore, new diagnostic markers are urgently needed for the early detection and prevention of breast cancer. A large number of studies have demonstrated that the pathogenesis of various tumors, including breast cancer, is associated not only with unlimited proliferation but also with the suppression of apoptosis [2]. Therefore, an in-depth and complete exploration of the abnormal proliferation and apoptosis associated with the pathogenesis of cancer may be of interest.

Apoptosis, a complex process in which cells neatly commit suicide, plays a critical role in the development and maintenance of homeostasis, and the elimination of malignant cells [3]. Accumulating evidence suggests that aberrant regulation of apoptosis can result in unchecked cell growth and proliferation during carcinogenesis and that malignant cells possess the ability to resist apoptotic stimuli [4, 5]. Fas and Fas ligand (FasL) are critical components of apoptosis and have special status in the apoptotic process. Moreover, they are involved in immune escape, by which malignant cells can protect themselves from attack by immune cells [6]. Fas is an apoptosis-signaling transmembrane receptor that belongs to the tumor necrosis factor receptor superfamily [5]. Upon ligation of its natural ligand, FasL, Fas triggers molecular interactions, resulting in initiation of a proteolytic cascade and eventually apoptosis [7]. Previous studies have suggested that decreased Fas or increased FasL expression may facilitate the development and progression of tumors by reducing tumor cell apoptosis or inducing immune cell apoptosis [8, 9].

The Fas and FasL genes are localized at chromosome 10q24.1 and 1q23, consisting of nine and four exons, respectively. Polymorphisms in the Fas and FasL genes have been linked to their differential expression. Thus, we hypothesized that polymorphisms in Fas and FasL could contribute to the variability in individual susceptibility to breast cancer. The most extensively investigated polymorphisms are −670A > G (rs1800682) and −1377G > A (rs2234767) in the promoter region of Fas, and −844C > T (rs763110) in the promoter region of FasL [10–12] (Fig. 1). Previous studies have investigated these three polymorphisms in the etiology of breast cancer [12, 2, 13, 11, 10, 14]. However, the association between Fas and FasL polymorphisms and breast cancer risk has not been conclusively established. Moreover, circulating soluble Fas (sFas), another form of Fas that has five variants because of alternative messenger RNA (mRNA) splicing, can inhibit Fas-mediated apoptosis by neutralizing FasL or anti-Fas antibodies [15–17]. Furthermore, increased concentrations of serum sFas have been observed in patients with breast cancer [18, 19]. On the other hand, sFasL, released from the cell surface by cleavage of matrix metalloproteinase-like enzyme, has been reported to have the opposite function to membrane-bound FasL and is involved in anti-inflammatory processes [20].

Fig. 1
figure 1

The mode chart of the human Fas and FasL polymorphism positions indicated

Full size image

In order to investigate the effects of Fas and FasL polymorphisms on breast cancer risk among the Chinese population, we conducted a population-based case-control study with 439 breast cancer patients and 439 healthy, age-matched controls. In addition, to assess the risk of potential functional polymorphisms in Fas for breast cancer, 22 tumor tissues and adjacent normal tissues were collected to measure differences in Fas expression. Furthermore, sFas and sFasL concentrations were measured in 180 breast cancer patients to elucidate the associations between sFas and sFasL expression levels and breast cancer risk. Finally, the clinical characteristics of patients’ tumors, such as estrogen receptor (ER), progesterone receptor (PR), and CerbB-2 status, were detected by immunohistochemistry (IHC).

Materials and methods

Study population

A total of 439 female patients with histologically confirmed and classified breast cancer (according to the World Health Organization classification) were recruited from January 2010 to December 2012 at Nanjing First hospital, Nanjing Medical University, China, after exclusion of their history of cancer, chemotherapy, or radiotherapy. A cohort of 439 healthy, age-matched women who visited the same hospital for a routine physical examination were enrolled as controls. All subjects were unrelated, of Han ethnicity, and residents of Nanjing and its suburbs. Upon receipt of written informed consent (before the initiation of any study-related procedure), each subject was interviewed to collect additional information about breast cancer risk factors. Meanwhile, three indexes (ER, PR, and CerbB-2) were detected using IHC. The protocol was approved by the ethics committees of Nanjing First Hospital, Nanjing Medical University, China.

DNA extraction and genotyping

Genomic DNA was extracted from peripheral blood samples using the GoldMag-Mini Whole Blood Genomic DNA Purification Kit (GoldMag, Xi’an, China) according to the manufacture’s protocol, and DNA concentrations were measured using spectrometry (DU 530 UV/Vis spectrophotometer, Beckman Instruments, Fullerton, CA, USA). Genotyping was determined using the Sequenom MassARRAY RS-1000 (Sequenom, San Diego, CA, USA) according to the standard protocol. Sequenom Typer 4.0 software was used for data management and analysis [21, 22].

Quantitative real-time PCR assay

Fas mRNA levels were measured in 22 breast cancer and adjacent normal tissue sample pairs. The total RNA was extracted from the tissues using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol, followed by RNA quantification with spectrophotometry. Complementary DNA (cDNA) synthesis from total RNA was carried out in a final volume of 20 μL using a cDNA synthesis kit (Takara, Dalian, China) according to the manufacturer’s protocol. Expression of Fas and the housekeeping GAPDH gene was measured using quantitative real-time PCR (qRT-PCR). Samples were assayed in a 20-μL reaction mixture including 2 μL cDNA, 10 μL SYBR Premix Ex Taq II (TaKaRa, Dalian, China), 0.8 μL of each specific primer (10 μM), 0.4 μL ROX reference dye, and 6 μL RNase-free water. The qRT-PCR reactions were performed using the ABI 7500 System (Applied Biosystems, Foster, CA, USA) with an initial incubation at 95 °C for 30 s, followed by 40 cycles at 95 °C for 5 s and 60 °C for 34 s. The primers for qRT-PCR of Fas (forward: 5′-TGA AGG ACA TGG CTT AGA AGT G-3′; reverse: 5′-GGT GCA AGG GTC ACA GTG TT-3′) and GAPDH (forward: 5′-GTC AAC GGA TTT GGT CTG TAT T-3′; reverse: 5′-AGT CTT CTG GGT GGC AGT GAT-3′) were used. To ensure the reproducibility of the results, all measurements were conducted in triplicate.

Enzyme-linked immunosorbent assay

Serum samples were extracted from whole blood by centrifuging at 3,000 rpm for 10 min and stored at −20 °C until use. The concentrations of sFas and sFasL in serum were determined by ELISA using the sFas and sFasL kits (USCN, Wuhan, China). In brief, each serum sample was diluted and added to the microtiter wells, which were precoated with monoclonal antibodies specific for Fas and FasL, respectively, followed by additional incubation with enzyme-linked polyclonal antibodies specific for Fas and FasL. Color development was then produced by adding tetramethylbenzidine substrate solutions and stopped using sulfuric acid. Finally, absorbance was measured at 450 nm with a microplate reader (Bio-Rad, Hercules, CA, USA). The concentration analysis of sFas and sFasL was conducted using standard curves. The sensitivities of the sFas and sFasL assays were 12 pg/mL and 0.038 ng/mL, respectively.

Immunohistochemistry assay

An IHC assay was used to assess ER, PR, and CerbB-2 expression in paraffin-embedded tumor tissue. The IHC procedure has been described elsewhere [23]. In brief, 3-μm tissue sections taken from paraffin blocks were placed on adhesive-coated slides and incubated overnight at 37 °C. After deparaffinization by xylene, the slides were passed through graded alcohols for 15 min and rinsed in distilled water. The slides were then placed in citrate buffer (pH 6.0) and in a microwave oven for 20 min to retrieve antigen. Incubation with 3 % hydrogen peroxide for 10 min can block endogenous peroxidase activity. The three primary antibodies used in this procedure were polyclonal rabbit antihuman CerbB-2 oncoprotein (Dako, Copenhagen, Denmark; 1:10 dilution, 90 min), monoclonal rabbit ER antibody (Spring Bioscience, Pleasanton, CA, USA; 1:100 dilution, 90 min), and monoclonal mouse antihuman PR antibody (Spring Bioscience; 1:100 dilution, 90 min). The secondary antibodies were applied for 30 min with primary antibody enhancer (Spring Bioscience) for 30 min. The slides were then incubated for 30 min with polyvalent HRP polymer (Spring Bioscience). AEC chromogen system was used to slides for 15 min (Spring Bioscience). The results were recorded as the percentage of positively stained target cells, with positivity defined as samples with more than 10 % stained neoplastic cell nuclei.

Statistical analysis

The allele frequencies of cases and controls were analyzed using the χ2 test. Odds ratios (ORs) and 95 % confidence intervals (CIs) were calculated using logistic regression analysis to assess the strength of the association between a polymorphism and breast cancer risk [24]. mRNA expression in tissue samples and sFas and sFasL concentrations in serum were calculated using ANOVA. In controls, genotypic frequencies for each polymorphism were tested for departure from Hardy–Weinberg equilibrium, and P < 0.05 was considered significant. All statistical analyses were conducted using the Statistical Analysis System (version 11, SAS Institute, Cary, NC, USA). Haplotype analysis, including frequency distribution, was performed using SHEsis software (http://analysis.bio-x.cn/myAnalysis.php).

Results

Characteristics of the study population

All of the participants in this study were women, with 439 breast cancer patients and 439 healthy, age-matched controls (mean ± standard deviation age at diagnosis: cases 52.89 ± 10.78 years; controls 52.95 ± 10.89 years; P = 0.933), and none drank alcohol or smoked cigarettes. Demographic data and clinical characteristics of the participants are summarized in Table 1.

Table 1 Clinical features of breast cancer patients and healthy controls
Full size table

Genotype distribution of the Fas and FasL polymorphisms

Genotype frequencies of the Fas −1377G > A, Fas −670A > G, and FasL −844 T > C polymorphisms among breast cancer patients and controls and their associations with breast cancer risk are presented in Table 2. Briefly, for all three polymorphisms, there was a significant difference among the genotype distribution between cases and controls. Moreover, the three polymorphisms were all associated with breast cancer risk. In addition, the genotype frequencies of the three polymorphisms among controls were consistent with estimates by Hardy–Weinberg equilibrium.

Table 2 Frequency distribution of Fas and FasL polymorphisms and their association with breast cancer risk in breast cancer patients and healthy controls
Full size table

Clinical features of cases and controls with Fas and FasL polymorphisms

The distributions of Fas −1377G > A, Fas −670A > G, and FasL −844C > T genotypes according to the clinical characteristics of patients and their associations with breast cancer risk are presented in Table 3. In brief, no significant association was observed between genotype or combined variant genotype distributions and any of the clinical parameters.

Table 3 Genotype frequencies of Fas (−1377G > A, −670A > G) and FasL (−844 T > C) polymorphisms among patients and their associations with breast cancer risk in relation to patient characteristics and clinical features of breast cancer severity
Full size table

Haplotype analysis between Fas −1377G > A and −670A > G polymorphisms

Because the two polymorphisms in Fas (−1377G > A and −670A > G) were in linkage disequilibrium with each other, the combined associations of these two polymorphisms with breast cancer were calculated by haplotype analysis. The results indicated that the −1377G/−670A haplotype was associated with an increased breast cancer risk as compared with other haplotypes, whereas the −1377A/−670A haplotype was associated with a reduced breast cancer risk compared with other haplotypes (Table 4).

Table 4 Haplotype frequencies of Fas −1377G > A and −670A > G polymorphisms in breast cancer patients and healthy controls, and associations with breast cancer risk
Full size table

Interaction effects of Fas and FasL polymorphisms on breast cancer risk

To further clarify the interaction among different genes, an analysis of gene–gene interactions between Fas and FasL in breast cancer risk was conducted (Table 5). Briefly, significant interactions between Fas −1377G > A and FasL −844 T > C (P < 0.001), as well as Fas −670A > G and FasL −844 T > C (P < 0.001), were found. The results showed that the Fas −1377GA/AA and FasL −844CC genotypes were associated with a reduced risk of breast cancer, as were the Fas −1377GA/AA and FasL −844TC/TT genotypes, the Fas −670AG/GG and FasL −844CC genotypes, and the Fas −670AG/GG and FasL −844TC/TT genotypes.

Table 5 Interaction effects between Fas (−1377G > A, −670A > G) and FasL (−844 T > C) polymorphisms, and associations with breast cancer risk
Full size table

Fas expression in tumor tissues and adjacent normal tissues

qRT-PCR was used to determine Fas mRNA levels in breast cancer and adjacent normal tissues. The results showed no significant differences in the mRNA levels of Fas between breast cancer and normal tissues (Fig. 2a; P = 0.532). There were also no significant associations observed in Fas expression levels by tumor stage (Fig. 2b; P = 0.866) or lymph node metastasis (Fig. 2c; P = 0.619). Furthermore, no significant differences were observed for Fas expression in tumor tissues when compared by genotypes of three polymorphisms (Fig. 3a–c).

Fig. 2
figure 2

Relative Fas expression in breast cancer tissues and adjacent normal tissues. a Fas expression in 22 breast cancer tissue and 22 matched adjacent normal tissue samples were measured by qRT-PCR. Fas expression according to b tumor stage and c lymph node metastasis

Full size image
Fig. 3
figure 3

Relative Fas expression in breast cancer tissues in three genotypes of Fas and FasL polymorphisms. Fas expression in three genotypes of a Fas −1377G > A polymorphism, b Fas −670A > G polymorphism, and c FasL −844 T > C polymorphism in breast cancer tissues

Full size image

Concentrations of circulating sFas and sFasL in different genotypes carriers

Circulating levels of sFas and sFasL were measured in 180 patients, who were meticulously selected from 439 cases. The associations among Fas −670A > G, Fas −1377G > A, and FasL −844 T > C genotypes and sFas and sFasL are summarized in Table 6. There was a significantly increased sFas expression in the Fas −1377GG genotype compared with those in the GA and AA genotypes (Fig. 4a and Table 6), and in the Fas −670AA genotype compared with those in the AG and GG genotypes (Fig. 4b and Table 6). However, no significant differences were found in sFas levels among the three genotypes of the FasL −844 T > C polymorphism (Table 6). In addition, no significant differences in the levels of circulating sFasL were found among the genotypes of any of the three polymorphisms (Table 6).

Table 6 Associations of Fas and FasL polymorphisms and serum concentrations of sFas and sFasL
Full size table
Fig. 4
figure 4

Association of Fas −1377G > A (a) and Fas −670A > G (b) polymorphisms and concentrations of sFas in patients’ serum. a The GG genotype was associated with higher sFas concentrations than GA or AA (P = 0.001 and P = 0.008, respectively). b The AA genotype was associated with higher sFas concentration than AG and GG (P = 0.001 and P = 0.010, respectively)

Full size image

Discussion

Fas and its endogenous ligand FasL trigger the death signal cascade and lead to apoptosis [25]. Earlier studies suggested that polymorphisms of Fas (−1377G > A, −670A > G) and FasL (−844C > T) were associated with breast cancer risk [2–14]; however, these studies reached inconsistent conclusions. In the current study, we observed that the Fas −1377A and −670G alleles were linked to a lower breast cancer risk, which is inconsistent with the findings of other studies [2, 10, 12, 14], and that FasL −844TC genotypes were also significantly associated with a lower breast cancer risk.

Shao et al. reported that the genotype of Fas −1377GA and the Fas −670AG or GG genotype were associated with decreased prostate cancer risk [26]. Such associations were earlier identified by Li et al. in cutaneous malignant melanoma [27]. Similarly, the current study found that the Fas −1377AA and −670GG genotypes were associated with decreased breast cancer risk. However, contradictory conclusions have been reported by other studies, as follows. For the Fas −1377G > A polymorphism, previous studies reported that people who carry the A allele have a higher breast cancer risk than those carrying the G allele [2, 10, 11, 14], and for the Fas −670A > G polymorphism, Hashemi et al. reported that the Fas −670GG genotype was associated with an increased risk of breast cancer [12].

The aforementioned inconsistent study results might be connected to previous evidence that the Fas −1377G > A change can damage its binding element Sp1 and lead to decreased levels of Fas, while the Fas −670A > G change can damage its binding element STAT1 with the same result [28, 29]. Therefore, polymorphisms in Fas may be associated with an increased risk of breast cancer by downregulating the capacity for apoptosis. However, recent evidence indicates that Fas is a double-edged sword, because it can transmit proliferation and activation signals and convert a tumor suppressor into a tumor promoter at certain levels of FasL [30]. Moreover, activated Fas can mobilize and accumulate immune suppressive cells (e.g., myeloid-derived suppressor cells, regulatory T cells) in tumor tissue, where they can exert a tumor-promoting effect by inhibiting T and natural killer cell proliferation and activation, and contribute to tumor angiogenesis [30–33]. In addition, the tumor-promoting effect of Fas has been confirmed in vitro and in vivo [34–38].

On the other hand, previous studies have suggested that the −1377G and −670A alleles may upregulate the production of sFas, which can protect malignant cells from Fas-mediated apoptosis by inhibiting Fas receptor–ligand binding in the extracellular space [26, 39–42], consistent with the results in this study. In addition, the Fas −1377G and −670A alleles were associated with higher levels of sFas, which inhibits Fas-mediated apoptosis by neutralizing FasL or antibody [15]. However, the present study did not observe any association of Fas in breast tissue with Fas polymorphisms. Thus, we believe that the association of Fas polymorphisms with breast cancer risk may be the result of altered concentrations of sFas in serum, rather than of Fas expression in tumor tissue. Furthermore, the association of the Fas −1377G > A and −670A > G diplotypes with breast cancer risk revealed a markedly lower risk for developing breast cancer in carriers of Fas −1377A/−670A, whereas the diplotype −1377G/−670A was associated with an increased risk of breast cancer in this study. These results might indicate that these two polymorphisms are in linkage disequilibrium, consistent with previous studies [10, 14].

For the FasL −844 T > C polymorphism, the present results showed a lower risk of breast cancer in TC compared with CC genotype carriers. Hashemi et al. reported an increased risk of breast cancer in CC compared with TT genotype carriers (OR = 3.18, 95 % CI = 1.21–8.33) in an Iranian population [12]. Zhang et al. found the TC genotype to be associated with a decreased breast cancer risk (OR = 0.76, 95 % CI = 0.62–0.94) in a Chinese population [10]. Furthermore, Wang et al. found the CC genotype to be associated with an increased breast cancer risk (OR = 1.92, 95 % CI = 1.46–2.54) compared with the TT/TC genotype in a Chinese population [2]. As previously mentioned, the −844 T > C mutation has been associated with elevated expression of both sFasL and its membrane types [10] due to connection with its binding motif CAAT/enhancer-binding protein β transcription factor, which was associated with a higher breast cancer risk [43]. The study of gene–gene interactions between Fas and FasL showed a markedly lower risk of breast cancer in those carrying the Fas −1377GA/AA genotype and the FasL −844TC/TT genotype compared with −1377GA/AA and −844TC/TT alone. The same phenomenon was observed for the Fas −670AG/GG and −844TC/TT genotypes, indicating a strong synergistic effect between these two genes with respect to breast cancer.

Due to different genetic backgrounds between individuals, it is of great importance to choose the anticancer drugs [44, 45]. It was reported that GSTs not only participate in drug detoxification but also is involved in the control of apoptosis by inhibition of JNK signaling pathway [46]. As thus, they have become the focus of chemotherapy resistance research, in which GSTP1 was the most extensively investigated in breast cancer [47, 48]. Previous studies showed that GSTP1 expression might play a vital role in drug resistance in breast cancer [47–49]. The GSTP1 − 313A > G (rs1695) polymorphism is reported to reduce enzyme activity in removal of chemotherapy agents [50–52]. Moreover, the −313GG or AG genotype was reported to correlate significantly with unfavorable prognosis for breast cancer patients treated with alkylating agents [53]. Besides, the P-glycoprotein encoded by the multidrug resistance gene MDR1 is important in efflux transport of the chemotherapeutic agent, doxorubicin, used in breast cancer treatment. The polymorphism in exon 26, −3435C > T, was found to be a considerable advantage in neoadjuvant chemotherapy. Kafka et al. [54] reported that −3435TT was associated with a clinical response for treating breast cancer patients with anthracycline alone or in combination with taxane. In contrast, a recent study reported that the −3435CC was related to an enhanced response after neoadjuvant therapy and longer time to progression after anthracycline-based chemotherapy.

This study has some limitations. First, it is difficult to determine the effect of environmental factors such as diet, use of oral contraceptives, occupational exposures, and physical exercise, which can influence breast cancer risk via gene–environment interactions. Second, although our patient population was not small, it was still insufficient for further subgroup analysis and may limit the statistical power of our study.

Conclusions

In conclusion, this study provides evidence that Fas −1377G > A, Fas −670A > G, and FasL −844 T > C polymorphisms are associated with altered breast cancer risk in the Chinese population. Two polymorphisms of Fas were in linkage disequilibrium, and two Fas polymorphisms and a FasL polymorphism had an interaction effect on breast cancer risk, respectively. Moreover, the Fas −1377GG and −670AA genotypes were associated with higher sFas concentrations than other genotypes, indicating that Fas polymorphisms may be associated with altered breast cancer risk mainly by regulating the concentration of sFas. These conclusions need to be validated in large and multiethnic populations.