International Journal of

Introduction
 
Recurrent Pregnancy Loss (RPL) is defined as repeated occurrence of 3 or more miscarriages before 20th week of gestation and its occurrence affects about 1-5% of couples (1, 2). Diverse factors have been claimed to cause of RPL, including; uterine anomalies, chromosomal abnormalities, endocrine and immune defects as well as thrombophilia and infection. In most cases, however, a single cause for repeated abortions cannot be identified (3). However, the development and function of female reproductive system are controlled by hypothalamus-pituitary-ovarian (HPO) axis. In addition, fertility and pregnancy may be negatively affected by the levels of hormones like follicle stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E₂) and progesterone that are also controlled by the HPO axis (4, 5).
Estrogens modulate multiple reproductive functions, including progesterone production and uteroplacental blood flow (6). Factors like female secondary sexual characteristics, reproductive cycle, fertility, and maintenance of pregnancy may play a crucial role in embryonic and fetal development under the effect of estrogen (7, 8). Estrogen receptors (ERs), which in human consisted of ERα and ERβ, coded by ESR1 and ESR2 genes, respectively, are the mediator of estrogen signaling and function as ligand-dependent transcription factors (9, 10). ERsubtypes are members of nuclear hormone receptor superfamily. The ESR1 gene (140 kb) which is located on chromosome 6q25.1 consists of 8 exons; likewise ESR2 gene is located on chromosome 14q22 and contains 8 exons.
Both ESR1 and ESR2 genes are expressed in the human ovary, within the oocyte, in granulosa cells, in ovarian epithelial cells and in endometrial cells of uterine (11-14). In recent years, increasing efforts have been made to describe genetic causes of recurrent miscarriage at the molecular level. Several polymorphic variants of the estrogen receptor genes have been described. More than 2200 single nucleotide polymorphisms (SNPs) have been identified in the ESR1 gene, while this assessment for ESR2 gene is 720 (www.snpper.chip.org). ESR1 and ESR2 gene polymorphisms have been linked to a higher risk of unexplained infertility and to the outcome of controlled ovarian stimulation, as well as to the ovarian response to FSH (15, 16). In the present study, we analyzed two SNPs (-397 C/T and -351 A/G) on the ESR1 gene in Iranian women with RPL.
Both these polymorphisms lie on intron 1 and are recognized by PvuII (-397C/T) and XbaI (-351A/G) restriction enzymes and have been reported to be associated with RPL in Spanish and Taiwanese Han population (17, 18). We therefore conducted a case-control study to examine the role of ESR1 SNPs and circulating concentrations of estradiol, FSH and LH with RPL risk. To our knowledge, there are no studies on the association of these polymorphisms and RPL in Iranian women. Also, in this study we compared Iranian haplotype with previous studies on different ethnic groups, to understand the analogy of different population.
 
Materials and methods
 
Patients
In this case control study, the case group consisted of 244 women with at least three consecutive miscarriages before 20^th week of gestation who referred to Avicenna Infertility Clinic, Tehran, Iran, between the years of 2008-2010, for treatment of RPL. All patients had undergone standard diagnostic procedures, including a detailed history, chromosomal analyses of peripheral blood lymphocytes, ultrasonography and hysterosalpinogography to detect uterine anomalies, hormone profiles on day 2-3 of the menstrual cycles; FSH, LH, E₂ as well as thyroid hormone studies, investigation of infections and immunological diagnostics (including anti-phospholipid and anti-cardiolipin antibodies). Women with uterine anomalies, karyotype abnormalities, hormonal, infection and immunological disorder were excluded from the study.
The control group consisted of 104 healthy ethnically matched women with two or more successful pregnancies and live births and no history of complicated pregnancies, miscarriages, still births, small for gestational age fetuses, preeclampsia, ectopic pregnancy or preterm delivery. The study was approved by the Avicenna Research Institute local ethics committee for medical research, and written consents were obtained from participating subjects before entry into the study.
 
Biochemical assays
LH and FSH were measured by chemiluminescent immune assays (CLIA) (Diasorin, Italy). E₂ was measured by enzyme immunoassay (EIA) (Axis Shield, Dundee, UK).
 
Isolation of genomic DNA
Peripheral blood from each case and control subject was collected in vacutainer tubes containing ethylenediamine tetra acetic acid (EDTA) as anticoagulant followed by DNA extraction using salting out method (19).
 
PCR-RFLP analysis of gene polymorphisms
PCR-RFLP for the -397C/T and -351A/G polymorphisms were performed using specific primers and restriction enzymes (Table I). PCR reaction mixture contained 1 μg of total DNA, 1X Master Mix (Taq DNA Polymerase Master Mix Red, Ampliqon, Denmark) and 10pM of each specific primer. The amplification cycles were 95^oC for 5 min, and 35 cycles of 95^oC for 30s, 60^oC for 30s, and 72^oC for 30s, plus a final elongation at 72^oC for 5 min. The PCR products were then digested by restriction enzymes PvuII              (-397C/T), XbaI (-351A/G) (Fermentas, Lithuania) (Table I) followed by running on a Gel Red (Biotium, USA) stained 2-2.5% agarose gel.
 
Statistical analysis
Comparisons between groups for distribution of each polymorphism were made by Chi-square test. Student’s t- test was used to compare age, BMI and hormone levels between control and RPL groups. Correlations between different parameters were determined with Spearman's rank correlation test. Values of p<0.05 were regarded as significant. All statistical analyses were performed with SPSS version 13.0 (SPSS Inc., Chicago, IL). Haplotype analysis was determined with HaploView 4.2 (HaploView software Inc, USA).
 
Results
 
The average age of the case group was 33.4 years (ranging from 20-48) and of the control group was 39.2 years (ranging from 28-52). The genotype frequencies of all investigated polymorphisms in case and control groups are shown in table II. Allele frequency and genotype distributions of the ESR1 polymorphisms were not significantly different between the control and case groups (397C/T; p=0.20 and -351A/G; p=0.09). Heterozygous and homozygous polymorphic frequencies of -397C/T and -351A/G in the RPL group did not differ significantly from those in the controls.
A significantly negative correlation was observed between the two ESR1 gene polymorphisms, -397C/T and -351A/G         (r=-0.852, p<0/001) in RPL women. For this reason we conducted a study with inferred haplotypes (Figure 1 and Table III), (D’: 0.959; r-square= 0.758, p<0.001) as a consequence of the short distance between them. Table IV summarizes endocrine parameters for the cases and controls. We found no statistically significant differences between groups in levels of E₂ and FSH. However, although women with increased the LH levels in the sera was excluded from the study, the level of LH (within the normal range) was significantly higher in RPL than that in controls (4.55±0.22 vs. 3.04±0.28; p<0.001). In addition, genotype distribution analysis showed no correlation with the levels of E₂, FSH and LH (data not shown).


Table I. Primer sequences and restriction enzymes used to detect ER gene polymorphisms


Table II. Genotype frequencies of RPL and control groups


Table III. Haplotype frequencies of -397C/T and -351A/G polymorphisms in control and RPL groups


Table VI. Endocrine parameters in women with RPL and healthy controls




Figure 1. Linkage disequilibrium comparisons of, -397C/T and -351A/G polymorphisms in settlement order in the ER1gene.


Discussion
 
Polymorphisms in the ERs influence vascular tone and flow and could result in a corresponding phenotypic difference in estrogen-dependent events, including those involved in pregnancy maintenance. Many studies have been performed to investigate the potential association between the -397T/C and -351A/G SNPs in ESR1 and recurrent miscarriage. In the present study, we found no significant differences in the allele distribution of the ESR1gene polymorphisms between RPL and healthy Iranian women. Our finding is in line with studies by Aléssio et al and Hanna et al who did not find any association between these two SNPs and RPL in Brazilian and Canadian population, respectively (5, 20). These findings may suggest that these two polymorphisms have little or no effect on maintenance of pregnancy. However, a recent study in Spain found an association between increased number of miscarriages and the ESR1 haplotype composed of the -397T and    -351A alleles (17).
These polymorphisms have opened an important area of research in reproductive disorders (RPL, infertility, endometriosis and age of menarche) and diseases such as osteoporosis, cardiovascular  and cancer (10, 21-26). Conflicting results from different studies may be due to the differences in ascertainment of women with sporadic and recurrent miscarriage. Moreover, the combinatorial effects of SNPs, particularly in extremely polymorphic genes, and gene environmental interactions is difficult to appropriately address in association studies. There are some hypotheses in the literature which describe functional significance of these two ESR1 gene polymorphisms. They are located on, 397 and 351 bp upstream of the start of exon 2, and possible functional mechanisms include changes in ESR1 expression due to altered binding of transcription factors and influence on alternative splicing of the ESR1 gene. These mechanisms can be caused directly by these polymorphisms or through linkage disequilibrium with a truly functional, but so far unknown sequence variation elsewhere in the ESR1 gene (27).
Schuit et al reported that the -397T/C polymorphism is associated with decreased plasma E₂ levels in an allele dose-dependent manner in postmenopausal women (27). They suggested that the potentially lower ESR1 expression caused by the T-allele leads to a lower expression of an enzyme, such as 17β- Hydroxysteroiddehydrogenases (HSD), in the estrogen synthesis pathway and subsequently, reduced E₂ synthesis. The fact that in the -351A/G polymorphism, the A-allele is also associated with E₂ levels may be due to linkage disequilibrium with the -397T/C SNP or another functional polymorphism, or due to functional significance of the -351A/G polymorphism itself (27). These findings suggest that the genotype-dependent differences in E₂ levels created by these polymorphisms may be clinically significant.
Similar to our study, Zofkova et al showed no association between ESR1 gene polymorphisms and sex hormone levels they found that the -397T/C and -351A/G polymorphisms were not significantly associated with E₂ levels in a study of 114 postmenopausal women. In the present study, the distribution of -397T/C and -351A/G polymorphisms (Table IV) was very similar to what was previously reported in Caucasian populations of European ancestry and differed significantly from what was observed in populations of Asiatic (28).
Haplotype Px was detected, albeit in a low frequency whereas haplotype pX was not observed. This frequency of haplotypes may result from the disequilibrium which is not complete and may be due to recombination or multiple mutations which have occurred between/ at these two polymorphic sites. Endocrinological abnormalities are present in about a quarter of women with unexplained recurrent miscarriage (29). Several studies have reported high basal levels of LH, with or without polycystic ovaries, as a risk factor for miscarriage (30-32).
In the present study we found that LH levels were significantly higher in RPL compared to controls. No significant differences were observed in the levels of FSH and E₂ between the two groups which may be due to sample size. Larger studies may be needed to warrant a conclusion as to whether Iranian RPL women should be screened for the ESR1 gene polymorphisms.
 
Acknowledgments
 
We are grateful to the women who entered this study. We greatly appreciate the assistance of the personnel of Avicenna Infertility Clinic in sample collection, notably Miss Elham Savadi. This work was supported by grants from Avicenna Research Institute and Qazvin University of Medical Sciences.
 
Conflict of interest
 
The authors declare that, there is no conflict of interest.