Polymorphisms in the Estrogen Receptor Alpha Gene and Mammographic Density Result Study in Brazilian Women

The estrogen receptor (ER) is a ligand-activated transcription factor that mediates the actions of the estrogen in target tissues. Several ERα gene polymorphisms are associated with changes in the receptor expression and function. The aim of this study is to verify the hypothesis that the ERα gene polymorphisms could be associated with high mammographic density(HMD), a well known independent risk factor for breast cancer ( BC ) in a casecontrol study carried out in the city of São Paulo (SP, Brazil) from 2010 to 2013 . Two ERα gene polymorphisms named PvuII and XbaI was examined in 308 cases and 155 controls. The PvuII polymorphism was associated with an increased risk of having high mammographic density (HMD) post menopause, after adjustment for other risk factors, the odds ratio for pp genotypes was 1.75 (confidence interval of 95% CI 95%=1.10-2.79 ) compared with the genotypes PP and Pp. The XbaI polymorphism was also associated with a high risk of HMD, but not statistically significant, odds ratio for xx genotype was 1.31 (95% CI=0.7 to 1.9). No apparent synergistic effects of these two polymorphisms were identified. It was concluded that the Pvull polymorphism in the gene ERα was associated with an increasing chance of have HMD, a strong risk factor for BC. Thus recognizing these risk factors will be of great importance in the analyses of individual susceptibility to BC, in both the study of the response to various drugs and the prognosis. Journal of Cancer Science & Therapy J o u r na l o f C anc er ence & Tera p y

history and physical examination: age at menarche and menopause, parity, age at first birth, family history of breast cancer (FHBC), smoking, alcohol intake and body mass index (BMI). Peripheral blood samples were obtained for genomic DNA extraction and determination of polymorphisms in question.
The genomic DNA was extracted from peripheral blood using QIAamp DNA Blood Mini Kit (Qiagen), following manufacturer instructions. After DNA quality and integrity evaluation were performed PCR-RFLP assay for Estrogen Receptor PvuII and XbaI polymorphism analysis as described by Herrington et al. 2002 [26]. The laboratory was blinded on the subject identification.
Hardy-Weinberg Equilibrium (H_W) was used to verify if the genotype frequencies of PvuII and XbaI polymorphisms in our population were in genetic equilibrium. Simple mathematical model [(p+q) 2=p2 +2 pq+q2] used to calculate genotype frequencies from allele frequencies.
Statistical analysis: The data were described using average, standard deviation (sd), absolute frequency (n) and relative frequency (%). To verify the association between qualitative variables with the mammographic density was used the chi-square (X 2 ). For comparison between the groups of HMDand controls as quantitative variables was used the Kolmogorov-Smirnov Test to verify the normality of the data, as these are not normally distributed in all groups, was used the nonparametric Mann-Whitney Test for comparison between groups. To verify the relationship of the variables with the occurrence of high mammographic density was used Multivariate Logistic Regression model stepwise backward. The variables that entered in the model were those presented in the bivariate analyzes values p<0.20. In all statistical tests were adopted a significance level of 5% (p<0.05).

Results
The distributions of selected demographic characteristics that are the main risk factors for breast cancer are presented in Table 3. It was observed elevated risk of HMD for the main risk factors that have been reported in other previous studies [20,16].
The risk of having HMD was also elevated for younger women, lower WC, fewer pregnancies, higher age at having first birth, high number of women with a FHBC. It was not observed any apparent modification effect for other indicators of exposure to endogenous estrogens and lifestyle factors. In our study, the sample consisted of postmenopausal women, aged 45-65 years.
Regarding the genotype, 21.5% of controls and 32.5% of cases were mutated homozygous (pp), and 28.5% of controls and 19.8% of cases were wild homozygous (PP), with statistically significant difference (X 2 =7.42, p=0.024). OR adjusted for pp genotype was 1.75 (95% CI=1.10 to 2.79) compared with PP and Pp genotype. There were not any significant difference in allele frequency or genotype polymorphism ERα-XbaI between controls and cases (Table 4).
When the two ERα polymorphisms were analyzed together, no synergistic effect was consistently noted.
Additional analyzes (Multivariate Logistic Regression) were performed to evaluate the independent risk factors for HMD. From all clinical characteristics analyzed, entered only those presented in the bivariate analyzes values p<0.20 were included in the model: the Pvull polymorphism of the ERα gene; indicators of exposure to endogenous estrogen: age at menarche, menopause, time after menopause,   Likewise, it was performed the association between the x allele of the XbaI and the same independent variables that characterize the sample, but no results confirmed the association.

Discussion
The association of ERα gene polymorphisms with the risk of BC draws attention because the ER functions as a regulator of the hormone dependent transcription, which plays a significant role in the development of BC [10]. Indeed, approximately two thirds of BC expresses the estrogen receptor alpha. Epidemiologic studies also correlate steroid hormones to changes the MD, and have examined whether variations in genes that regulate the biosynthesis and hormonal metabolism could explain individual differences in MD. The ER gene, located in the long arm of chromosome 6q25, has been associated to HMD in many studies, due to its importance in the development, progression and prognosis of BC. Several polymorphisms of ERα gene have been reported, with the PvuII and XbaI -located in intron 1 of the ERα gene with 50 base pairs between them -being the most studied. Several diseases, including BC [6,9,10,18,30], endometrial cancer [31], Alzheimer's disease [32], obesity [33,34], endometriosis [35], leiomyomas [36] and bone mineral density [37], have been evaluated for a possible connection with ERα gene polymorphisms.
Our findings revealed a statistically significant difference between the two groups of mammographic density as the main risk factors for breast cancer that has been reported in previous studies. The risk of having HMD was increased in younger women, with a lower waist circumference, fewer pregnancies, a higher age at the first birth, and a high number of women with a family history of breast cancer. We observed that for the increase of 1 year in age, the probability of being classified as HMD as compared to controls, decreased 6.8%. The groups of women with HMD were 56.31 years old in average, while in the controls group the average age was 58. 16. This relationship was statistically significant (p<0.001), indicating that younger women are most likely to have HMD. These findings are in agreement with the literature, showing decrease in breast density patterns with increasing age (Stomper et al. 1996 andBoyd et al. 2007). Among the avoidable risk factors throughout life, obesity/overweight are the most important ones, since they are the main determinants of the estrogen level in women after menopause [34] and, exposure to estrogens during life is a well established factor of increased risk for BC. According to recent studies, there is a positive relationship between BMI and breast cancer, especially in postmenopausal women [38][39][40]. Nevertheless, women with high BMI are less likely to have HMD, even though the risk for BC development increases in women with HBD and who are obese (obese, OR=1.7: obese HBD, OR=2.01) [41]. It is also suggested that for every 5 kg of weight gain in adult life, the risk of developing BC [42] increases by 8%. Our data showed a high prevalence of obesity, but it was predominant in the control group (47.7%). Obesity was highly prevalence in the HMD group (36.7%), a surprising and worrying finding that deserves attention because of women with HBD, a factor considered as a high risk for developing BC (RR>4.0).
Our findings revealed a statistically significant difference between the two groups of mammographic density also with respect to the distribution of the PvuII polymorphism genotype (p=0.024). It was found that women with two mutated alleles (homozygous mutant) had 76% greater chances of being diagnosed with HMD, compared to those with one or two normal alleles for this polymorphism ( Table 4). The risk associated with the p allele of Pvull was shown to be an independent risk for other indicators of high exposure to endogenous estrogens. It was not observed any apparent effect modification for other indicators of exposure to endogenous estrogens and lifestyle factors. Similarly, van Duijnhoven et al. [17] studied the influence of ERα and HT on the MD, reporting an association between increased MD and the presence of the mutated genotype ERα-397 (OR=2.24) in women using HT when compared to those with wild genotype and not HT users. Parl et al. [30] found that the pp genotype of the PvullSNPs was higher in women with a diagnosis of BC at a younger age. Yaich et al. [9] examined the PvuII polymorphism in tumor tissue of 257 women with primary breast cancer and compared it to peripheral blood collected from 140 controls not affected by the disease. Women with BC and pp genotype had a diagnosis of BC at an earlier age compared with those with PP or Pp genotypes of the group that had cancer.
Investigating the genotype distribution of the XbaI polymorphism, The molecular mechanisms through which these polymorphisms alter the receptor activity are not clear because PvuII and XbaI are located in an intronic, and apparently nonfunctional, region of the gene. Possible explanations include: (a) the existence of a functional combination between polymorphic alleles, where the two markers in combination would alter the gene function, as well as the RNAm stability [23]. Thus, this study investigated if the combination of PvuII and XbaI polymorphisms increased the risk of having HMD, and did not observe any synergistic effect between them; (b) Another explanation would be that polymorphisms in intron could have a Linkage Disequilibrium (LD) with the exon which would affect the function of the ER as a whole [37]; in this regard, it has been investigated whether polymorphisms in intron 1 are in LD with repeat polymorphism (GT)n located at 6627 bp after the beginning of the transcription site of the exon 1 and to 144 kb before the exon 2; (c) Growth factors and their signaling molecules are important for cancer growth and its progression. There is considerable cross-talk between ER and growth factors such as insulin and IGF1 (growth factors like insulin), and the family of epidermal growth factors [43].
The allelic and genotypic frequencies obtained for the ERα-397 PvuII polymorphism (P=46.8%, p=53.2%) were similar to those found in other studies in which this polymorphism was correlated with MD or BC [11,13,17,19,27,29]. The genotype distribution was in Hardy-Weinberg equilibrium with the following frequencies: pp genotype 29.0%, Pp 48.5% and PP 22.6% (Table 4).
The allele frequencies for the Xbal polymorphism (x=56.9%, X=43.1%) were lower when compared to two other studies conducted in Brazil [20,27], but they were in accordance with the frequency found by van Duijnhoven et al. [17], Molvarec et al. [28] and Hsieh et al. [29]. The genotype distribution for ERα-351 XbaI was also in Hardy-Weinberg equilibrium with the following frequencies: xx genotype 31.6%, Xx 50.4% and XX 17.9% (Table 4).
In relation to the independent variables that showed associationwith MD (age, race, WC, menarche, menopause, time after menopause, number of pregnancies, age atthe first birth, number of births, number of abortions, FH, BMI, alcohol intake, smoking and PvuII and XbaI genotypes), after performing a multiple logistic regression, only the clinical factors (age, WC, number of pregnancies, age at the 1 st birth, FH and PvuII polymorphism in the ERα) showed to be independent risk factors for HMD (p<0.05).
In summary, the present case-control study concluded that the PvuII polymorphism in the gene ERα was associated with an increased chance of having HMD, a factor of high risk for BC. Thus, recognizing these risk factors will be of great importance in the analyses of individual susceptibility to BC, in both the study of the response to various drugs (for example HT) and the prognosis ( Table 5).