Association of estrogen, progesterone and follicle stimulating hormone receptor polymorphisms with in vitro fertilization outcomes

ABSTRACT

Despite the advances in in vitro fertilization (IVF), the implantation success rate for infertile women remains approximately only 15%. In this study, we sought to determine whether implantation failure after repeated IVF treatments is influenced by the presence of common variants in estrogen α, progesterone and follicle stimulating hormone receptor genes. The study population included three groups of women: group 1 were 50 women who had the transfer of ≥3 high-quality embryos during the IVF procedure without ever having had a clinical pregnancy; group 2 were 50 women who achieved a clinical pregnancy after ≤3 high-quality embryos transfers and group 3 were 50 control subjects who achieved a clinical pregnancy without any fertility therapy that resulted in a one live-born infant. Genotype analysis was performed using polymerase chain reaction and Sanger sequencing for rs6165, rs6166, rs2234693, rs9340799. While progesterone receptor single nucleotide polymorphism (SNP)  was genotyped based on the amplicon size, the repeats for the ESR1 TA-repeat polymorphism were calculated based on the fragment length. A higher frequency of the heterozygote AG genotype was observed in the infertile groups when compared to controls. Significantly, an allele combination of T of rs2234693, A of rs9340799; S of ESR1 (TA), A of rs6166, G of rs6165 and del of PROGINS had a higher frequency in women who had a successful IVF outcome compared to women who had an unsuccessful IVF outcome, indicating a possible protective combined genotype that could reduce a negative outcome during IVF. This study has demonstrated that combining several candidate genes is needed to assess which may play a role in fertility.

Abbreviations: CI: confidence interval; COH: controlled ovarian hyperstimulation; DNA: deoxyribonucleic acid; ESR: estrogen receptors; FSH: follicle stimulating hormones; FSHR: FSH receptor; IVF: in vitro fertilization; PGR: progesterone receptors; SNP: single nucleotide polymorphism

KEYWORDS:

• Estrogen receptor
• progesterone receptor
• follicle stimulating hormone receptor
• in vitro fertilization outcomes
• haplotype analysis

Introduction

Nearly 15–20% of couples in the reproductive age experience infertility, which is defined as ‘the incapacity to conceive naturally after 1 year of regular unprotected sexual intercourse’ (Rowe and Comhaire 2000). In approximately 40% of these cases female factors like age, tubal state, frequency of ovulation, luteal phase deficiency, endometriosis or cervical and immunologic factors are responsible (Christofolini et al. 2001; Sen et al. 2013). However, to date only a few susceptibility factors for idiopathic female infertility have been reported. Given that many aspects of female reproductive function are strongly influenced by genetic factors, there is a need to identify more susceptibility loci in women for accurate management of idiopathic infertility.

Female human reproduction is controlled by a number of hormones with progesterone, estrogen and follicle stimulating hormones (FSH) being the three essential gonadotrophic hormones. Progesterone plays a crucial role in maintaining implantation and promoting uterine growth (Su et al. 2011). The major physiological effect of progesterone is mediated by progesterone receptors (PGR). FSH is involved in follicle development, oocyte maturation, steroidogenesis regulation and proliferation of granulosa cells (Trevisan et al. 2014). This hormone acts through the FSH receptor (FSHR) which is expressed on granulosa cells in the ovary. The third key hormone, estrogen, enhances the action of FSH by increasing the expression of FSHR in the granulosa cells thereby promoting its proliferation (Rod et al. 2014). This physiological response to estrogen is controlled by estrogen receptors (ESR) α and β (Liaqat et al. 2015).

Single nucleotide polymorphisms (SNPs) in any of these three genes, or in combination, could modify hormone action and the endocrine feedback systems (Ilgaz et al. 2015). This could result in interindividual variation in reproductive performance which in turn could influence the success of the assisted reproductive techniques, like in vitro fertilization (IVF), which are popular treatment options for infertile couples. For PGR, a functional SNP of PGR, PROGINS, was reported to be associated with adverse reproductive outcomes, including unexplained infertility and repeated IVF implantation failure (Su et al. 2011). PROGINS is a complex which includes a 306 bp Alu element in intron 7, a missense SNP in exon 4 and a silence SNP in exon 5, all of which are in complete linkage disequilibrium (Rowe et al. 1995). In addition, two SNPs in exon 10 of FSHR, p. Ala307Thr (rs6165) and p. Ser680Asn (rs6166) have been studied in the context of IVF outcome. Rod et al. (2014) have reported that the FSHR variant p. Ser680Ser is associated with poor response to FSH stimulation. Rs6165, which causes an amino-acid change of threonine to alanine, is located within the transmembrane region of FSHR protein (Yan et al. 2013) and hence is proposed to be involved in the hormone-binding ability of FSHR and FSH-mediated signal transduction during ovarian stimulation (Kene et al. 2004; Agrawal and Dighe 2009). Other predictive markers of ovarian response are polymorphisms of the alpha gene of ESR (ESR1); rs2234693, also known as the PvuII restriction polymorphism, and rs9340799, also known as the XbaI restriction polymorphism. Both these polymorphisms have been reported to be associated with the risk of infertility (Anagnostou et al. 2013). For the third SNP, the microsatellite length polymorphism ESR1 (TA)n, the longer ESR1 (TA) n microsatellite repeat polymorphism has been reported to be associated with an improved ovarian response to FSH. Using this information, we compared the distributions of these six polymorphisms between women who had ≥3 implantation failures versus women who succeeded in having at least one live-born infant after the first IVF attempt and women with a confirmed fertility.

Results

A total of 150 Indian women participated in this study and the infertile groups were compared with the fertile group for the six polymorphisms. The demographic and clinical characteristics of the study population are summarized in Table 1 and the genotype and allele frequencies of the polymorphisms for all three groups are summarized in Tables 2 and 3. Comparing the genotype distribution between the women who had an unsuccessful IVF outcome and the control group, both the heterozygote genotype and the dominant model of rs6165 (AA vs. AG + GG) showed a significant difference between the groups. In addition, the G homozygote mutant allele and the dominant model of rs6166 showed a significant difference between the women who had a successful IVF outcome and those that did not.

Table 1. Demographic and clinical characteristics of the study population.

Variable	Cases (n = 100)		Controls(n = 50)
	Group 1	Group 2	Group 3
	(unsuccessful IVF outcome) (n = 50)	(successful IVF outcome) (n = 50)	(fertile women)
Age in years	31.8.3 ± 5.4 (range 21–45 years)	30.4 ± 3.7 (range 21–41 years)	29.2 ± 3.5 (range 21–41 years)
Age of menarche	12.33 ± 1.17	11.63 ± 1.01	12.07 ± 1.26
Duration of infertility	4.37 ± 2.63	3.26 ± 1.62	–
Body mass index (BMI) (25 kg/m^2)	25.88 ± 4.75	24.2 ± 3.61	25.0 ± 3.23

Table 2. Comparison of genotype distribution between group 1 (unsuccessful IVF outcome) and 3 (controls).

Genotype	Group 3 (n = 50)	Group 1 (n = 50)	OR (95% CI)	p Value*
rs2234693
CC	17	19	Ref	–
CT	28	24	0.77 (0.33–1.80)	0.69
TT	5	7	1.25 (0.33–4.69)	0.74
CC vs. CT + TT	33	31	0.84 (0.37–1.90)	0.68
CC + CT vs. TT	5	7	1.47 (0.43–4.97)	0.54
rs9340799
AA	15	16	Ref	–
AG	31	24	0.73 (0.30–1.76)	0.17
GG	4	10	2.34 (0.60–9.10)	0.22
AA vs. AG + GG	35	34	0.91 (0.39–2.13)	0.83
AA + AG vs. GG	4	10	2.87 (0.84–9.88)	0.08
ESR1 (TA) n
SS	45	47	Ref	-
SL	5	1	0.19 (0.02–1.70)	0.05
LL	0	2	4.78 (0.22–102.50)	0.32
SS vs. SL + LL	5	3	0.57 (0.13–2.55)	0.46
SS + SL vs. LL	0	2	5.20 (0.24–111.24)	0.29
rs6166
AA	28	22	Ref	–
AG	19	25	1.67 (0.74–3.79)	0.46
GG	3	3	1.27 (0.23–6.93)	0.78
AA vs. AG + GG	22	28	1.62 (0.74–3.57)	0.23
AA + AG vs. GG	3	3	1.00 (0.19–5.21)	1.00
rs6165
AA	30	19	Ref	-
AG	15	29	3.05 (1.31–7.13)	0.01
GG	5	2	0.63 (0.11–3.59)	0.60
AA vs. AG + GG	20	31	2.45 (1.10–5.47)	0.03
AA + AG vs. GG	5	2	0.38 (0.07–2.03)	0.23
PROGINS
del/del	50	47	Ref	–
ins/del	0	3	7.44 (0.37–147.93)	0.18

*A p value of <0.05 is considered to be statistically significant.

Table 3. Comparison of genotype distribution between group 2 (successful IVF outcome) and 3 (control).

Genotype	Group 3 (n = 50)	Group 2 (n = 50)	OR (95% CI)	p Value*
rs2234693
CC	17	18	Ref	–
CT	28	20	0.67 (0.28–1.62)	0.38
TT	5	12	2.27 (0.66–7.80)	0.19
CC vs. CT + TT	33	32	0.92 (0.40–2.08)	0.83
CC + CT vs. TT	5	12	2.84 (0.92–8.79)	0.06
rs9340799
AA	15	22	Ref	–
AG	31	20	0.44 (0.18–1.04)	0.06
GG	4	8	1.36 (0.35–5.35)	0.65
AA vs. AG + GG	35	28	0.55 (0.24–1.24)	0.15
AA + AG vs. GG	4	8	2.19 (0.61–7.80)	0.23
ESR1 (TA) n
SS	45	50	Ref	–
SL	5	0	0.08 (0.01–1.52)	0.09
LL	0
SS vs. SL + LL	5
SS + SL vs. LL	0
rs6166
AA	28	17	Ref	–
AG	19	25	2.17 (0.93–5.06)	0.07
GG	3	8	4.39 (1.02–18.86)	0.046
AA vs. AG + GG	22	33	2.47 (1.10–5.55)	0.03
AA + AG vs. GG	3	8	2.98 (0.74–11.98)	0.12
rs6165
AA	30	26	Ref	–
AG	15	21	1.62 (0.69–3.76)	0.26
GG	5	3	0.69 (0.15–3.18)	0.63
TT vs. TA + AA	20	24	1.38 (0.63–3.06)	0.42
TT + TA vs. AA	5	3	0.57 (0.13–2.54)	0.46
PROGINS
del/del	50	48	Ref	–
ins/del	0	2	5.21 (0.24–111.24)	0.29

*A p value of <0.05 is considered to be statistically significant.

Haplotype analysis was performed to determine the association of combined genotypes of all the six polymorphisms with IVF outcome. Table 4, shows the haplotype analysis for genotypes of women who had an unsuccessful IVF outcome (group 1) and women who had a successful IVF outcome (group 2). An allele combination of T of ESR1 rs2234693, A of ESR1 rs9340799; S of ESR1 (TA), A of FSHR rs6166, G of FSHR rs6165 and del of PROGINS had a higher frequency in group 2 compared to group 1(OR: 6.14, confidence interval (CI): 1.20–31.32, p value: 0.029). Another allele combination of C of ESR1 rs2234693, A of ESR1 rs9340799; S of ESR1 (TA), G of FSHR rs6166, A of FSHR rs6165 and del of PROGINS also had a higher frequency in group 2 compared to group 1. However, the significance of this genotype was only marginal.

Table 4. Comparison of genotype distribution between group 1 (unsuccessful IVF outcome) and 3 (controls).

S.No	SNP1	SNP2	SNP3	SNP4	SNP5	SNP6	Haplotype frequency		OR (95% CI)	p value*
							Group 1	Group 2
1	C	A	S	A	G	del	0.1689	0.1847	Ref	–
2	T	G	S	A	G	del	0.1146	0.0618	0.52 (0.16–1.70)	0.276
3	C	G	S	A	G	del	0.1072	0.1081	0.94 (0.32–2.74)	0.916
4	T	A	S	A	G	del	0.0174	0.1304	6.14 (1.20–31.32)	0.029
5	T	A	S	G	G	del	0.108	0.0663	0.60 (0.18–1.91)	0.38
6	T	A	S	A	A	del	0.0752	0.0517	0.67 (0.18–2.54)	0.560
7	C	A	S	G	G	del	0.0645	0.0581	0.94 (0.25–3.50)	0.931
8	C	A	S	G	A	del	0.0241	0.1134	5.19 (1.00–26.94)	0.049
9	C	G	S	G	A	del	0.0739	0	0.06 (0.003–1.18)	0.065
10	C	A	S	A	A	del	0.072	0.0154	0.13 (0.01–1.21)	0.074
11	C	G	S	G	G	del	0.0296	0.0623	1.88 (0.41–8.77)	0.417
12	T	G	S	G	A	del	0	0.0715	14.19 (0.75–267.4)	0.076
13	C	G	S	A	A	del	0.03	0.018	0.63 (0.09–4.24)	0.634
14	T	G	S	G	G	del	0	0.0382	8.51 (0.43–169.9)	0.160
15	T	A	S	A	G	ins	0	0.0199	4.73 (0.21–105.6)	0.326
16	C	G	S	A	G	ins	0.03	0	0.14 (0.01–2.81)	0.196

SNP1: ESR1 rs2234693; SNP2:ESR1 rs9340799; SNP3: ESR1 TA–repeat; SNP4: FSHRrs6166; SNP5: FSHR rs6165; SNP6: PROGINS.

*A p value of <0.05 was considered to be statistically significant.

Discussion

IVF is a complex process where oocytes are collected for fertilization and subsequent implantation after controlled ovarian hyperstimulation (COH) with FSH. In this multistep process either the response to COH or the implantation process is often influenced by interindividual variability. In this study, we examined the association of six polymorphisms of receptors of three key hormones with an increased risk of failed embryo implantation failure during IVF. While polymorphisms provide useful information on disease susceptibility, gene haplotypes have a better predictive power in the analysis of associations between these gene variants and on interindividual variability.

Interestingly, our results have demonstrated an allele combination of T of rs2234693, A of rs9340799; S of ESR1 (TA), A of rs6166, G of rs6165 and del of PROGINS had a higher frequency in women who achieved a clinical pregnancy after ≤3 high-quality embryos transfers compared to women who had the transfer of ≥3 high-quality embryos during the IVF procedure without ever having had a clinical pregnancy. Thus, this haplotype could indicate a protective combined genotype that could reduce the negative outcome of IVF. There was also some evidence supporting a statistical difference between individual infertile groups and the controls. While the heterozygote genotype and the dominant model of rs6165 showed a variation between the groups 1 and 3, the G homozygote mutant allele and the dominant model of rs6166 showed a difference between the women of groups 2 and 3. However, since neither the genotype distribution nor the haplotype analysis showed a significant difference between the 100 infertile women and the 50 fertile women, this result needs to be explored in a larger population to evaluate the implication of these findings.

FSHR rs6165 and rs6166 polymorphisms have been reported to be involved in ovarian response to hyperstimulation with FSH (Wunsch et al. 2007). rs6166, which is linkage disequilibrium with rs6165 affects the baseline FSH level and increases the amount of gonadotrophin required during COH. We found a higher frequency of the heterozygote AG genotype of rs6165 and the G homozygote mutant genotype of rs6166 in the infertile group. Considering that both these SNPs are linked, the presence of the mutant allele in any of these genes may predispose the individual to infertility and influence ovarian function. In a similar study comparing the FSHR SNPs between infertile and fertile women (Sever et al. 2014) a significantly higher frequency of the GG homozygote polymorphism for rs6166 was observed in the women who responded well to ovarian hyperstimulation. Both these findings are in contrast with that of de Castro et al. (2003), where the homozygote mutant was observed more in women who were classified as poor responders to hyperstimulation. The interaction of FSH with its receptor is crucial for the follicular development and maturation. Any variation in the genotype of FSHR may contribute to the altered ability of the receptor to bind FSH and to induce signal transduction pathway, and could affect reproductive ability, especially in women.

Surprisingly, our investigation of ESR and PROGINS polymorphisms indicate that these SNPs do not appear to be good markers in predicting IVF outcome in our study groups. Again, there are mixed results with other populations; prior findings have indicated that the Pvull C allele frequency was lower among women who were poor responders (≤3 follicles) (Altmäe et al. 2007; Georgiou et al. 1997; de Castro et al. 2004; ÖÜ et al. 2009) and a better fertilization rate was observed in the XbaI GG genotype (ÖÜ et al. 2009). Though PROGINS has been found to be associated with endometriosis-associated infertility, Gimenes et al. (2010) and Pisarska et al. (2003) found an increase in the prevalence of PROGINS mutations among women with idiopathic infertility.

In conclusion, this study has shown an association between FSHR rs6165 and rs6166 genotypes and IVF outcome as summarized in Supplementary Tables 1 and 2. Although the number of patients studied was small, this study has demonstrated that combining several candidate genes is needed to assess which of them play a major role in the fertility process, allowing for a more precise evaluation of a patient before IVF.

Methods

Subject population

This prospective case control study enrolled 100 infertile female patients who had opted for assisted reproductive treatment at Kanmani Fertility Hospital and Sri Ramachandra Medical Hospital (Chennai, Tamil Nadu) between 2012 and 2014. Inclusion criteria included at least 1 year of infertility, a regular menstrual cycle of 25–35 days, the presence of morphologically normal ovaries and a normal karyotype. Based on the outcome after IVF, these 100 patients were then differentiated into Group 1, which included women who did not achieve a pregnancy after ≥3 (range 2–5) IVF cycles where at least two embryos were transferred in each attempt. Women who achieved a viable pregnancy by their second IVF attempt with the same conditions as the first group were placed in Group 2. A third group of women included 50 women who achieved a clinical pregnancy, without any fertility therapy, that resulted in a live-born infant. This study was approved by the Sri Ramachandra University ethics committee and all the participants were recruited with informed consent (IEC-NI/8 January 2002/03).

Genotyping

Genomic deoxyribonucleic acid (DNA) was isolated from peripheral blood leucocytes using the blood DNA isolation kit (QIAGEN, Hilden, Germany). Using previously reported primers and standard procedures, polymerase chain reaction was performed to amplify and analyze the SNP of PROGINS (Gomes et al. 2007), FSHR rs6165, FSHR rs6166 (Singhasena et al. 2014), ESR1 rs2234693, ESR1 rs9340799 (Kitawaki et al. 2001) and the ESR1 TA-repeat polymorphism (McIntyre et al. 2007).

Statistical analysis

For each polymorphism, the Hardy-Weinberg equilibrium was calculated for cases and controls using the online calculator available at dr-petrek.eu/documents/HWE.xls. To determine the difference in genotype frequencies between the groups, the Fishers exact test was used to calculate the odds ratio, CI and p value. A p value less than 0.05 was regarded as statistically significant. The online tool, SNPStats, was used to perform the haplotype analysis.

Acknowledgments

We are thankful to Ms. Priyanka Venugopal for her help in the haplotype analysis. We are also grateful to the study participants for their consent.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplemental Material

Supplemental data for this article can be accessed here.

Additional information

Notes on contributors

Solomon Franklin Durairaj Paul

Conceived and designed the experiments: VG, VV; Performed the experiments: TK, VV; Analyzed the data: VV, TK; Contributed reagents/materials/analysis tools: SRN, SM, SFDP; Wrote the Manuscript: TK, VV.