The ESR 1 and GPX 1 gene expression level in human malignant and non-malignant breast tissues

Background: The aim of this study was to establish whether the gene expression of estrogen receptor alpha (encoded by ESR1) correlates with the expression of glutathione peroxidase 1 (encoded by GPX1) in the tumor and adjacent tumor-free breast tissue, and whether this correlation is affected by breast cancer. Such relationships may give further insights into breast cancer pathology with respect to the status of estrogen receptor. Methods: We used the quantitative real-time PCR technique to analyze differences in the expression levels of the ESR1 and GPX1 genes in paired malignant and non-malignant tissues from breast cancer patients. Results: ESR1 and GPX1 expression levels were found to be significantly downregulated by 14.7% and 7.4% (respectively) in the tumorous breast tissue when compared to the non-malignant one. Down-regulation of these genes was independent of the tumor histopathology classification and clinicopathological factors, while the ESR1 mRNA level was reduced with increasing tumor grade (G1: 103% vs. G2: 85.8% vs. G3: 84.5%; p<0.05). In the non-malignant and malignant breast tissues, the expression levels of ESR1 and GPX1 were significantly correlated with each other (Rs=0.450 and Rs=0.360; respectively). Conclusion: Our data suggest that down-regulation of ESR1 and GPX1 was independent of clinicopathological factors. Down-regulation of ESR1 gene expression was enhanced by the development of the disease. Moreover, GPX1 and ESR1 gene expression was interdependent in the malignant breast tissue and further work is needed to determine the mechanism underlying this relationship.


INTRODUCTION
Breast cancer is the most common cancer among women worldwide.The number of diagnosed breast cancer cases among women has continued to rise since the 1980's, and now it constitutes 20% of all malignant tumors.Women aged 45-69 years are at the highest risk of developing breast cancer, and the incidence rate in that group is 50% of all the diagnosed breast cancer cases (Bojar et al., 2012).
Pathogenesis and development of breast cancer is often related to estrogen receptors (ERs) and their estrogen ligands.ERs belong to a large family of nuclear receptors that play a role of a transcription factors in cells.There are two types of ERs: alpha (ERα) and beta (ERβ) encoded by the ESR1 and ESR2 genes, respectively, and presenting opposite roles.Activation of ERα is associated with proliferation and growth of tumor cells (Au et al., 2007;Lin et al., 2007), while ERβ promotes apoptosis, suppresses malignant transformation and inhibits growth of tumor cells (Ström et al., 2004;Paruthiyil et al., 2004;Behrens et al., 2007).ERs regulate transcription by direct interaction and binding to DNA (Klinge, 2001) or indirectly through other transcription factors (e.g.AP-1 activator protein-1) (Kushner et al., 2000).ERs owe the ability to bind to DNA to specific zinc finger structures located in their DNA-binding domain.One zinc finger is responsible for binding to DNA, while the function of the other one is to stabilize the ER-ER homodimer (Schwabe et al., 1993).Zinc fingers are highly susceptible to oxidation, which for example may occur due to accumulation of reactive oxygen species (ROS) (Webster et al., 2001).Oxidation of cysteine thiol groups results in the release of zinc ions, causing change in the tertiary structure and loss of the protein ability to bind to DNA (Liang et al., 1998).
Cells are protected from oxidizing agents, such as ROS, by antioxidant enzymes: catalase, superoxide dismutases (soluble and extracellular Cu/ZnSOD and Mn-SOD) and selenoproteins, such as the family of glutathione peroxidases (GPx's) and thioredoxin reductases (TrxRs) which, with glutathione and thioredoxin (Trx), respectively, form an active ROS-reduction system and ensure redox homeostasis in a cell (Schafer & Buettner, 2001;Valko et al., 2007).Cu/ZnSOD is the first line of defense against ROS by catalyzing the dismutation reaction of the superoxide anion radical to hydrogen peroxide.TrxR, on the other hand, utilizes NADPH to reduce and activate Trx, as well as other proteins (Mustacich & Powis, 2000).Reduced Trx is an oxidative stress response protein that activates transcription factors in order to alter the expression of peroxiredoxin genes, so that cellular hydrogen peroxide can be diminished (Webster et al., 2001).H 2 O 2 is also subsequently enzymatically reduced to water by peroxidases, including GPx-1 (encoded by GPX1) and catalase.GPx-1 is found in the cytosol, in mitochondria, and also in peroxisomes.It uses reducing equivalents of glutathione to detoxify organic and hydrogen peroxides, and its activity depends on the selenium availability (Lubos et al., 2011).It was previously reported by Shultz-Norton ( 2008) that TrxR and Cu/ ZnSOD are closely related to ERα by being a part of a large ERα-ERE (estrogen response element) protein complex in the nucleus, where they influence regulation of estrogen-responsive genes in the target cell (Rao et al., 2009;Rao et al., 2008).Apart from this, TrxR is involved in maintaining a reduced cellular environment and active transcription factors (Arnér & Holmgren, 2000).That observation provides evidence of its special function in protecting ERα against oxidative agents in the nucleus.
Due to strong antioxidant properties of GPx-1 and high sensitivity of zinc finger structures to ROS, and the presence of Cu/ZnSOD and TrxR in the nucleus protein complex, we decided to investigate the relationship between the GPX1 mRNA level and the ESR1 mRNA level in human breast tissue.
More specifically, the differences in constitutive expression levels of the above mentioned genes between the healthy non-malignant and paired tumorous breast tissue specimens, as well as their mutual associations in the healthy and/or tumorous breast tissues, were analyzed.Moreover, the effect of tumor grading and staging on the above mentioned differences and/or associations was determined.The investigated relationships between the expression levels of the targeted genes may give fur-ther insights into breast cancer pathology with respect to the estrogen receptor status.

Patients and tissue specimens
The study involved 37 breast cancer female patients aged 44-82 years (mean age 63.1 years; S.D. 9.9 years) undergoing a curative resection surgery without adjuvant chemotherapy or radiotherapy at the Department of Oncology Surgery, Regional Cancer Center in Lodz, Poland, between November 2011 andDecember 2013.Of all the enrolled patients, 9 reported themselves as current-smokers, 9 as ex-smokers, 18 as non-smokers and 1 subject did not specify her smoking-status in detail.At the time of the study, none of the subjects received hormonal replacement therapy, but 9 of them declared hormonal treatment for more than 1 year in the past.Detailed characteristics of the investigated group of patients with respect to various clinicopathological factors (the histological grade (G), the primary tumor site (T) and the regional lymph node involvement (N), estrogen receptor (ER) and progesterone receptor (PR) status, Her/neu-2 status) and their smoking status, are presented in Table 1.There were no statistically significant differences in the age and BMI between the above mentioned groups (data not shown).
Prior to analysis, a written and informed consent for participation in the study was obtained from each enrolled subject.The study was performed in accordance with the guidelines of the Helsinki Declaration for human research and was approved by the Local Bioethics Committee for Scientific Research (resolution no.01/2011).
Thirty-seven primary breast tumor specimens (including 25 ductal carcinomas and 12 breast tumors of different types: 5 lobular carcinomas and 7 not specific type carcinomas) with paired non-malignant surrounding breast tissue samples, were removed intra-operationally and placed immediately at -20°C for 24 h, transported to the Nofer Institute of Occupational Medicine and stored at -80°C until further processing.
Gene expression analysis.Total RNA was isolated from the malignant and adjacent non-malignant breast tissue specimens using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instruction.Genomic DNA contamination was removed by the on-column digestion with the RNase-free DNase set (Qiagen, Hilden, Germany).Total RNA was further quantified and analyzed with regard to protein content using an Eppendorf BioPhotometer instrument (Eppendorf, Germany) and stored at -80°C.An aliquot of 200 ng of purified RNA was then reverse-transcribed in a 20 μl reaction mixture using a QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions, on an MJ Research BioRad PTC-200 DNA Peltier thermal cycler (MJ Research, Watertown, MA, USA) and the cDNA samples were frozen at -20°C.
Expression levels of the ESR1 and GPX1 genes were evaluated by means of the quantitative real-time PCR (qPCR) technique with the BioRad's CFX96 Real Time PCR system (BioRad, Hercules, CA, USA) using an SsoAdvanced SYBR Green Supermix (BioRad, Hercules, CA, USA) and beta-actin (ACTB) as the reference gene.Real-time PCR reactions were performed in 10 μl reaction mixture containing 5 ng cDNA, 500 nM of each of the forward and reverse primers, and 1x SsoAdvanced SYBR Green Supermix.The primer sequences (Table 2) were designed by the Beacon Designer 7.0 (PREMIER Biosoft Int., Palo Alto, CA, USA) and cycling conditions comprised of 30 s of polymerase activation at 95°C, followed by 49 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 30 s and extension at 72°C for 30 s. Products of the PCR reaction were analyzed by means of the Melt Curve technique using the Bio-Rad CFX Manager Software.qPCR efficiencies were calculated using dilutions of 5 randomly selected and pooled cDNA samples.All of the samples were measured in duplicate and the paired malignant and non-malignant breast tissue specimens were always analyzed in one analytical run in order to avoid between-run variations.As confirmed by the initial data analysis, expression of the reference gene (ACTB) was stable under experimental conditions.
Normalized relative expression level (NRQ) for a given gene of interest in the tumorous versus the paired adjacent non-malignant sample and the expression level of genes of interest normalized to the expression level of the housekeeping gene ACTB (NQ) was calculated utilizing a method described previously by Pfaffl (Pfaffl et al., 2002), based on each sample's average CT value and each gene's average PCR efficiency.
Statistical analysis Normality of the data was evaluated by the Shapiro-Wilk's W-test.Experimental data showing departure from normality are presented as median and interquartile range (IQR; in parentheses).To test whether differences in the expression levels of genes of interest normalized to the expression levels of the reference gene between the non-malignant and tumorous breast tissues met the criterion of statistical significance, the Mann-Whitney U-test was utilized.The between-group differences in the measured parameters were tested by the Mann-Whitney U test or the Kruskal-Wallis test.Spearman's rank correlation coefficient (R S ) was used to assess simple associations between the variables.Analyses were performed using the STATISTICA 10 software package (StatSoft, Tulsa, OK, USA).

Expression level of ESR1and GPX1 genes in malignant and non-malignant breast tissues
We observed a statistically significant down-regulation of expression level of the ESR1 and GPX1 genes in tumorous breast tissue when compared to the adjacent non-malignant one.In the tumorous tissue samples, expression level of the ESR1 gene was down-regulated when compared to the adjacent non-malignant one by 14.7% (NRQ(ESR1)=0.872,IQR: 0.691-1.154;p<0.05), whereas the expression level of GPX1 was reduced by 7.4 % (NRQ(GPX1)=0.931,IQR: 0.753-1.080;p<0.05) (Table 1).

Correlation between the expression levels of the investigated genes in the malignant and non-malignant breast tissue samples
We found significant positive correlations between normalized relative expression levels (NRQ) of ESR1 and GPX1 (R S =0.454, p<0.05) (Fig. 1), as well as the normalized expression level (NQ) of these genes in both, the non-malignant (R S =0.450, p<0.05) (Fig. 2) and malignant (R S =0.360, p<0.05) (Fig. 3), breast tissue samples analyzed separately.We also noted a positive correlation between mRNA level of the ESR1 gene and estrogen receptor status (R S =0.438, p<0.05).

DISCUSSION
This study analyzed association between mRNA expression level of the GPX1 gene and mRNA level of the ESR1 gene in both, non-malignant and tumorous breast tissues, and evaluated a possible role of such relationship in the development of breast cancer.
The antioxidant defense is very important for maintaining tertiary structure of ERα which has been previously described for human MCF-7 breast cancer cell line in the case of which antioxidant enzymes, like Cu/ ZnSOD and TrxR, interact with ERα to form a large protein complex, which migrates to the nucleus following the receptor activation (Schultz-Norton et al., 2008).This observation led us to investigate an association between mRNA level of the ESR1 gene and GPX1, yet another crucial antioxidant enzyme, even though it is not involved in the abovementioned protein complex.
In the study presented here we demonstrated that expression level of the ESR1 gene was significantly decreased in tumorous breast tissue when compared to the adjacent non-malignant one.This down-regulation of ESR1 was found to be related to the histological grade of the tumor and decreased significantly with increasing grade of cancer.It is known that receptor status can change as the tumor progresses (Amir et al., 2012).In very high-grade cancers this expression decreases or can even be lost (Huang et al., 2014;Stierer et al., 1993).Such trend was also observed in our study.In patients with the G1 breast cancer, the ESR1 expression levels did not differ significantly between malignant and paired non-malignant breast tissue, nevertheless, the ESR1 expression was significantly reduced in malignant breast tissue of the G2 and G3 breast cancer patients.The overall down-regulation of ESR1 expression observed when all samples were analyzed together can be explained by the structure of the group of patients examined in this study, with G2 and G3 patients accounting for 84% of all patients, which may have significantly influenced the level of ESR1 expression measured in the whole group of patients.The expression level of ESR1 varies depending on the progesterone receptor status.We found that the down-regulation of ESR1 expression in malignant breast tissue when compared to a paired non-malignant one was significantly much more pronounced among patients with a negative progesterone receptor status when compared to those with a positive progesterone receptor status.This observation may be related to the fact that activated estrogen receptor alpha induces transcription of the progesterone receptor (Kastner et al., 1990).Reciprocally, the lack of progesterone receptor may thus be seen as a consequence of reduced activity of ERα, which in turn may results from reduced amount of ERα due to down-regulated expression of ESR1.
It is noteworthy that the down-regulation of ESR1 expression relates particularly to the group of patients with positive estrogen receptor status, shortage of progesterone receptor activity, negative Her/neu-2 status and more advanced/bigger tumors (T2).We observed this down-regulation separately in all of the abovementioned subgroups, but considering the limited size of the study group we were unable to assess whether such ESR1 down-regulation would be also observed in a group of patients presenting all of these clinicopathological features together.Such an observation would be very interesting and would allow one to answer the question if, from a genetics point of view, the bigger tumors with negative HER/neu-2 status and lacking the progesterone receptor, tend to transform into the triple-negative subtype (TN) of breast cancer (with negative estrogen/ progesterone receptor and Her/neu-2 status) or into tumors with decreased expression of ERα (ER-) instead.Tumors transformed into the TN or ER-type are highly undesirable.These types of tumors are more aggressive than other subtypes of breast cancer and are characterized by poorer survival rates.This mainly follows from the fact that TN and ER-tumors are the most difficult ones to be treated because of the lack of benefits from the endocrine therapy and molecular targeted treatments for Her/nau-2 (Qiu et al., 2016).
The level of enzymatic activity and protein concentration of GPx-1 in tumor tissue has been broadly investigated in relation to breast cancer (Tas et al., 2005;Kumaraguruparan et al., 2002;Punnonen et al., 1994;Portakal et al., 2000).Those studies have shown an increased activity in tumor tissue when compared to the normal one.Contrary to immunocytochemical research, we evidenced the down-regulation of GPX1 mRNA expression in tumorous breast tissue as compared to the paired non-malignant tissue samples.This finding seems analogical to the results of other previous studies that have reported lowered expression of GPX1 mRNA in colorectal (Nalkiran et al., 2015) and gastric (Min et al., 2012) cancer.In the case of gastric cancer, almost 25% of the cases even lacked the GPX1 expression.These outcomes were associated with an advanced gastric cancer, lymphatic invasion, aggressiveness of this cancer and poor patient survival (Min et al., 2012).In our study, the level of the GPX1 transcript was found to be down-regulated independently of the clinicopathological factors.
Down-regulation of the GPX1 gene in tumorous tissue may lead to decreased GPx-1 protein level and in consequence to reduction of its enzymatic activity.Shortage in the antioxidant defense may lead to increased oxidative stress in cells which may possibly have two mutually opposite effects: excessive levels of ROS may induce the carcinogenesis process and progression of cancer on one hand, but at the same time may be toxic to cancer cells on the other one (Barrera, 2012).
Induction of increased level of ROS in cancer cells is an often used chemotherapeutic approach.Chemotherapeutic agents, such as vinblastine, cisplatin, mitomycin C or doxorubicin, exert their anticancer activity by inducing the ROS-dependent apoptosis of cancer cells (Chiu et al., 2012;Casares et al., 2012;Kim et al., 2012).Hence, declined antioxidant response in cancer cells, due to down-regulated GPX1, for example, may be of benefit for further treatment.
In addition to a separate analysis of ESR1 and GPX1 transcript levels, we also analyzed the relationship between these two genes.We found a significant positive correlation between the levels of GPX1 and ESR1 transcripts, regardless of the tissue type.These results allow us to hypothesize that expression levels of the GPX1 and ESR1 genes are mutually inter-related, even though GPx-1 has not been previously identified among proteins involved in the formation of the protein-ERα-ERE complex.Moreover, research studies performed up to date have not defined any interaction mechanism between ERα and GPx-1 at the protein level, as well as any molecular relationships between genes encoding these proteins (e.g.mediated by common transcription factors).Based on the data presented here, we hypothesize that the down-regulation of GPX1 expression may lead to increased oxidative stress in tumorous breast tissue, which in turn may lead to a decreased expression of the ESR1 gene.This may, however, be contradictory to a previous study, according to which the oxidative stress induced by hydrogen peroxide, the main substrate of GPx-1, has only a minimal effect on the ERα level in MCF-7 cells (Tamir et al., 2002).On the other hand, interaction between the GPX1 and ESR1 gene expression can be opposite.In MCF-7 cells, it was observed that physiological concentration of 17-β-estradiol acts through the membrane-located estrogen receptors on activities of the MAP kinase (MAPK) and NFκB.It was shown that activation of MAPK and NFκB by estrogen, up-regulates expression of the Mn-SOD and GPx-1 antioxidant enzymes (Borras et al., 2005).Thus, this aspect definitely remains open and deserves further investigation.
The major weakness of the study concerns a relatively small sample size, which limited the possibility to perform a more advanced statistical analysis of the data.Also, the lack of information about the further course of treatment does not allow us to draw extensive conclusions about the influence of the ESR1 and GPX1 genes' expression level on the effectiveness of the therapy.
In summary, our study provides evidence in favor of the significant down-regulation of ESR1 and GPX1 expression in malignant breast tissue when compared to the adjacent non-malignant breast tissue.The correlation between these genes was significantly positive regardless of the type of tissue.The extent of down-regulation of ESR1 in tumorous tissue as compared to the paired nonmalignant breast tissue was dependent on clinicopathological factors and was mostly related to the histological grade and progesterone receptor status, while the GPX1 expression was reduced in tumorous tissue when compared to the surrounding non-malignant one, independently of the clinicopathological breast cancer features.
Based on our data, it seems evident that further research is needed in order to fully elucidate the mechanism underlying association between expression level of the ESR1 and GPX1 genes in the malignant and adjacent non-malignant breast tissue.

Figure 2 .
Figure 2. Correlation between the expression level of ESR1 and GPX1 normalized to expression level of the housekeeping gene (NQ) in non-malignant breast tissue samples.Spearman's rank correlation analysis results: R s =0.450; p<0.05.(dotted lines: 95% Cl for regression line; n=35).

Figure 3 .
Figure 3. Correlation between the expression level of ESR1 and GPX1 normalized to expression level of the housekeeping gene (NQ) in tumorous breast tissue samples.Spearman's rank correlation analysis results: R s =0.360; p<0.05.(dotted lines: 95% Cl for regression line; n=35).

.944 (0.814-1.029)
Data presented as median normalized relative quantity (NRQ) of mRNA copies in paired tissue samples with respective interquartile range (in parentheses).In stratified analysis, NRQ values (i.e.ratio of normalized expression of a gene in tumorous breast tissue to paired nonmalignant breast tissue) within all strata were tested for significance by means of the Mann-Whitney U test.Statistically significant NRQs are indicated by asterisks (*p<0.05);Between-group comparisons of ESR1 and GPX1 expression levels were tested for significances by a the Mann-Whitney U test or b Kruskal-Wallis test.Statistically significant differences between individual strata are presented in bold (p<0.05).mRNA level of ESR1 and GPX1 in breast cancer tissues