Reproductive BioMedicine Online (2010) 20, 588 593 www.sciencedirect.com www.rbmonline.com ARTICLE Frequency distribution of polymorphisms in the FSH receptor gene in infertility patients of different ethnicity EAM Kuijper a, *, MA Blankenstein b, LJ Luttikhof a, SJM Roek a, A Overbeek a, PG Hompes a, JWR Twisk c,d, CB Lambalk a a Division of Reproductive Medicine, Department of Obstetrics and Gynaecology, VU University Medical Center, 1007 MB Amsterdam, The Netherlands; b Department of Clinical Chemistry, VU University Medical Center, Amsterdam, The Netherlands; c Department of Clinical Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands; d Department of Methodology and Applied Biostatistics, Institute of Health Sciences, VU University, Amsterdam, The Netherlands * Corresponding author. E-mail address: e.kuijper@vumc.nl (EAM Kuijper). Esther AM Kuijper was born in 1977 in Alkmaar, The Netherlands. She studied Medicine at the VU University Medical Center in Amsterdam until May 2004. Since then she has been working as a Phd student on a thesis entitled Comparison of perinatal reproductive hormone status between familial dizygotic twins, non-familial dizygotic twins, monozygotic twins, singletons and their mothers. Current research interests are twins and basic endocrinology. Abstract Studies on the frequency distribution of follicle-stimulating hormone receptor (FSHR) polymorphisms report conflicting results. It has been suggested that ethnicity might influence these outcomes. Therefore, the aim of this study was to determine the frequency distribution of FSHR polymorphisms at position 680 of exon 10 within a large group of women with fertility problems from different ethnic backgrounds. A total of 1771 women of different ethnic origin (Caucasian, Asian, Hindustani, Creole and Mediterranean) were studied. FSHR single-nucleotide polymorphisms at codon 680 of exon 10 were determined by restriction fragment length polymorphism of amplicons generated by polymerase chain reaction. Genotypes were compared with serum FSH concentrations and between different ethnic groups. A significantly lower number of Asians (10.5%) were found to have the Ser680Ser receptor variant compared with Caucasians (21.5%) and Mediterraneans (22.3%) (P = 0.010). FSH concentrations did not differ between the various ethnic groups, or the different FSHR polymorphisms. In conclusion, the Ser680Ser receptor variant is less common in the Asian subgroup compared with Caucasians and Mediterraneans. This indicates that, when comparing allelic frequency distributions of the FSHR polymorphism variants, ethnic background should be accounted for. FSH concentrations did not differ between FSHR polymorphisms or between ethnic groups. RBMOnline ª 2010, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. KEYWORDS: ethnicity, FSH receptor, genotyping, infertility, polymorphism 1472-6483/$ - see front matter ª 2010, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.rbmo.2010.02.004 S60 Reprinted from Vol. 20, No. 5 (2010) pp. 588 593.
FSHR gene polymorphisms and ethnicity 589 Introduction The follicle-stimulating hormone receptor (FSHR; 678 amino acids) is expressed solely on Sertoli cells of the testis and granulosa cells of the ovary and is part of the G protein-coupled receptor family (Hermann and Heckert, 2007). It consists of an extracellular N-terminal (359 amino acids) responsible for follicle-stimulating hormone (FSH) binding, a transmembrane region and an intracellular C-terminal which acts in signal transduction (Gromoll and Simoni, 2005; Sudo et al., 2002; Wunsch et al., 2005). Following receptor-ligand binding, receptor dimerization activates an intracellular signal-transduction mechanism involving protein kinase A and related proteins. This sequence of interactions is essential for normal functioning of the ovary and testis. Genetic defects caused by mutations in the FSHR, which result in gain or loss of function, have been described but are rare (Gromoll and Simoni, 2005). However, several single-nucleotide polymorphisms (SNP) have been identified in the FSHR gene, some of which are very common. The majority of the known SNP are located in the introns and have no known effect on receptor activity. The two most common SNP in the coding region of FSHR are in exon 10, i.e. in the protein domain fundamental for signal transduction, but not necessary for ligand-binding. Of these, the Ser Asn variant in exon 10 at position 680 (rs6166), which is in linkage disequilibrium with the Ala Thr variant (rs6156) at position 307, is highly polymorphic (Ahda et al., 2005; Gromoll and Simoni, 2005). Heterozygote frequencies of Ser680Asn range from 0.311 0.523 dependent on the population studied (www.hapmap.org), but subject numbers are limited. Earlier studies showed that women with the Ser680Ser genotype are characterized by a higher ovarian threshold for FSH, resulting in higher serum FSH concentrations (de Koning et al., 2006; Falconer et al., 2005; Laven et al., 2003; Perez Mayorga et al., 2000) and a higher need for FSH during ovarian stimulation for IVF (Simoni et al., 2002) and an almost doubling of the risk of clomiphene citrate resistance in patients with polycystic ovary syndrome (Overbeek et al., 2009). Furthermore, Greb et al. (2005) observed a longer menstrual cycle in women with the Ser680- Ser genotype. However, the impact of the Asn680Ser polymorphism in fertile and infertile patients remains controversial. Apparently, studies have potentially been biased by differences in populations studied: for example, cohort size, fertile and infertile populations and, probably most important, ethnic background (Ferlin et al., 2006; Gleicher et al., 2007; Simoni et al., 1999, 2002). Since 2003, FSHR polymorphism determination of the 680 SNP in exon 10 has been part of the routine fertility screening in the study clinic. Therefore, this study was able to analyse a large group of fertility patients with respect to FSHR polymorphism and to determine the frequency distribution within different ethnic groups. Materials and methods Patients This study included 1771 women (mean age ± SD 33.1 ± 4.9 years) who visited the Reproductive Medicine Unit of the Obstetrics and Gynaecology department at the VU University Medical Centre in Amsterdam between January 2003 and April 2006. Five different ethnic groups were distinguished: Caucasian, Asian (Japanese and Chinese), Hindustani, Creole and Mediterranean. The study population characteristics are shown in Table 1. Genotyping FSHR polymorphism screening has been part of the routine infertility screening since 2003. For genotyping of the Asn680Ser variant in exon 10 of the FSHR gene, genomic DNA was isolated from buffy coats using automated isolation on the BioRobot MDX, according to the manufacturer s instructions (Qiagen, Germany). FSHR polymorphism determination was done by restriction fragment length polymorphism analysis with polymerase chain reaction (PCR), using a modification of the method described by Simoni et al. (2002). For this, a 307 bp DNA fragment containing rs6166 was amplified by standard PCR in the presence of betaine with 1 cycle of 95 C for 4 min, 35 cycles of 95 C for 1 min, Table 1 Characteristic Study population characteristics. Value No. of patients 1771 Mean age (years) 33.1 ± 4.9 Ethnicity Caucasian 80.8 Asian 3.8 Hindustani 1.9 Creole 3.7 Mediterranean 9.9 FSHR variants Asn680Ser 48.8 Ser680Ser 21.0 Asn680Asn 30.2 Reason given when making appointment a Cycle disorder 11.0 Hyperprolactinaemia 0.5 Recurrent miscarriage 1.3 Male infertility 2.7 Refertilization surgery 1.1 Hirsutism 0.3 Primary infertility 50.9 Secondary infertility 28.6 Other 3.6 Values are percentage, unless otherwise stated. a Diagnoses as follows: cycle disorder = oligo/ amenorrhoea; recurrent miscarriage = two or more miscarriages; male infertility = semen analysis showing oligo/aspermia; primary infertility = trying to conceive for over 1 year with no history of pregnancy; secondary infertility = a patient with a previous pregnancy that tried to conceive for over 1 year. Reprinted from Vol. 20, No. 5 (2010) pp. 588 593. S61
590 EAM Kuijper et al. 52 C for 1 min and 72 C for 2 min, followed by extension for 10 min at 72 C and cooling. Restriction digestion of amplicons was performed using BsrI (New England Biolabs, USA) in the presence of acetylated bovine serum albumin. Following digestion for 2 h at 37 C, restriction fragments were size-separated by gel electrophoresis, visualized after ethidium bromide staining and scored. In this assay, a single band of 307 bp represents homozygosity for asparagine (Asn680Asn), two bands (189 and 118 bp) represents homozygosity for serine (Ser680Ser), and three bands (307,189 and 118 bp) represents heterozygosity (Asn680Ser). Hormone assays FSH Serum samples for FSH measurements were taken on cycle day 3. FSH was measured using an immunometric assay (Delfia fluorescentie; Perkin Elmer Life and Analytical Sciences, Wallac Oy, Turku, Finland). The lower limit of quantitation for FSH is 0.5 IU/l (coefficient of variation 10%) and the inter-assay coefficient of variation was below 8%. Oestradiol Serum samples for oestradiol were taken on cycle day 3 and measured using a radioimmunoassay (Double Antibody Diasorin; Saluggia, Italy). The lower level of quantitation is 18 pmol/l (coefficient of variation 20%) and the inter-assay coefficient of variation was below 16%. Statistics Statistical analyses were conducted with Statistical Package for Social Sciences version 15.0 (SPSS Inc, USA). Because concentrations of FSH and oestradiol were positively skewed, a logarithmic transformation was performed before analyses. Chi-squared tests were used to search for different frequency distributions of the FSHR polymorphisms within the various ethnic groups. If the chi-squared test indicated a significant difference, Student t-tests were used to compare the groups individually and correct for multiple testing by Bonferroni. Furthermore, a correspondence analysis was used to decompose the overall chi-squared test statistics into cell-specific components (Krzanowski, 2000). Linear regression analyses were conducted to compare FSH concentrations between the various ethnic groups and the different receptor polymorphisms. When necessary, a correction was made for age and oestradiol concentrations. To be able to compare this study s data with data previously described by others, some additional subgroup analyses were done on women with regular cycles aside from the total population. Results FSHR polymorphism and ethnic background The chi-squared test showed a significant difference for the FSHR polymorphism distribution within the various ethnic groups (chi-squared = 16.5, 8 degrees of freedom; P = 0.036). After correction for multiple testing (Bonferroni), a significant difference was found between Caucasians and Asians (P = 0.010) and between Asians and Mediterraneans (P = 0.010). This suggests that it is the Asian group that provides the largest deviations from expectation on the assumption of homogeneity, which were further evaluated with the correspondence analysis. Distribution of FSHR polymorphism variants per ethnic group and results of the correspondence analysis, describing the relationship between these groups, are shown in Table 2. The largest absolute numerical values are clearly for the Asian group, i.e. ( 1.170) 2 + ( 1.893) 2 + (3.070) 2 = 14.38, indicating that this group is separate from the others. The remaining coefficients are rather small by comparison. The large positive value for the Asn680Asn component denotes that this receptor variant is grossly over-represented in the Asian group, when compared with the expected value. FSH concentrations and ethnic background Linear regression analyses were conducted with FSH as the dependent variable and ethnic background (put into the model as dummy variables with Caucasians as a reference) as the explanatory variable (model 1). This model was tested with age and oestradiol as possible confounders (model 2). The results are shown in Table 3. Creole women have 1.138 times higher FSH concentrations compared with Caucasians, but after correction for age and oestradiol concentrations, no significant differences between the various ethnic groups were found. Table 2 Distribution of the FSHR polymorphisms per ethnic group. Ethnic group FSHR polymorphism Signed chi-squared component a Asn680Ser Ser680Ser Asn680Asn Total Asn680Ser Ser680Ser Asn680Asn Caucasian 700 (48.9) 308 (21.5) 423 (29.6) 1431 0.071 0.381 0.408 Asian 26 (38.8) 7 (10.4) 34 (50.7) 67 1.170 1.893 3.070 Hindustani 17 (51.5) 7 (21.2) 9 (27.3) 33 0.224 0.019 0.301 Creole 31 (47.7) 12 (18.5) 22 (33.8) 65 0.126 0.457 0.542 Mediterranean 90 (51.4) 39 (22.3) 46 (26.3) 175 0.500 0.353 0.932 Values are number (%), unless otherwise stated. a The signed chi-squared components describe the relationship between the FSHR polymorphisms for each ethnic group. S62 Reprinted from Vol. 20, No. 5 (2010) pp. 588 593.
FSHR gene polymorphisms and ethnicity 591 Table 3 Model FSH concentrations within the various ethnic groups. Ethnic group B P-value 95% CI 1 Asian 1.059 NS 0.948 1.183 Hindustani 0.973 NS 0.829 1.142 Creole 1.138 0.027 1.014 1.275 Mediterranean 1.010 NS 0.939 1.087 2 Asian 1.034 NS 0.931 1.148 Hindustani 0.989 NS 0.849 1.151 Creole 1.084 NS 0.972 1.208 Mediterranean 1.062 NS 0.989 1.140 Age 1.021 <0.001 1.017 1.025 Oestradiol 0.809 <0.001 0.773 0.846 B = transformed difference in FSH concentrations between the indicated ethnic group and the reference group Caucasians (and therefore represents a FSH ratio between these groups); NS = not statistically significant. FSHR polymorphism variants and FSH concentrations To analyse the relationship between FSH concentrations and FSHR polymorphism variants, which were put into the model as dummy variables, a linear regression model was used. Again, age and oestradiol concentrations were entered into the model as possible confounders. This was done for the total female population as well as for the women with regular menstrual cycles only (>11 menstruations per year). The results are shown in Table 4. In the population of women with regular menstrual cycles, significantly higher FSH concentrations were found in the Ser680Ser receptor variant group compared with women with an Asn680Ser receptor variant (P = 0.034). After correction for age and oestradiol concentrations, this difference was not significant. Discussion It has been suggested that, when comparing frequency distributions of FSHR polymorphic variants, ethnic background should be considered (Ferlin et al., 2006; Gleicher et al., 2007; Simoni et al., 1999, 2002). The present study is the first to report on the distribution of the SNP at codon 680 of the FSHR gene within such a large population of infertility patients from various ethnic backgrounds. The results indicate unmistakably that a significantly lower number of Asian women have a Ser680Ser FSHR variant and a significantly higher number have an Asn680Asn variant, compared with Table 4 Model FSHR concentrations and FSH polymorphism variants. n B P-value 95% CI Total female population 1 Asn680Asn 507 0.990 NS 0.942 1.041 Ser680Ser 359 1.035 NS 0.977 1.094 2 Asn680Asn 507 0.986 NS 0.941 1.035 Ser680Ser 359 1.019 NS 0.967 1.075 Age 1.020 <0.001 1.016 1.025 Oestradiol 0.806 <0.001 0.771 0.843 Women with regular cycles 1 Asn680Asn 237 1.018 NS 0.955 1.085 Ser680Ser 185 1.079 0.034 1.006 1.157 2 Asn680Asn 237 1.000 NS 0.943 1.063 Ser680Ser 185 1.054 NS 0.988 1.125 Age 1.023 <0.001 1.018 1.029 Oestradiol 0.782 <0.001 0.738 0.828 This table shows the results of the crude model (1) and the model corrected for age and oestradiol (2) used to describe the relationship between FSH concentrations and the various FSHR polymorphism variants within the total female population and women with regular menstrual cycles only. Asn680Ser is used as a reference (n = 829). B represents the transformed difference in FSH concentrations between the indicated FSHR polymorphism variant and the reference (Asn680Ser) and therefore represents a FSH ratio between the FSHR polymorphisms. NS = not statistically significant. Reprinted from Vol. 20, No. 5 (2010) pp. 588 593. S63
592 EAM Kuijper et al. Caucasians and Mediterraneans. The frequency distribution within the Asian population is in agreement with the data published by Sudo et al. (2002) reporting on infertile Japanese women. A study comparing Caucasian and Asian women also reported a different FSHR polymorphism distribution pattern between these populations, but unfortunately, they only focused on a SNP which is located in the promoter region of the FSHR (Wunsch et al., 2005). Asian women with polycystic ovary syndrome have greater sensitivity to gonadotrophin stimulation compared with Caucasians as shown by Palep-Singh et al. (2007), which might be caused by a lower number of women with the Ser680Ser receptor variant within the Asian population. For Caucasian women, the distribution patterns were 29.6% for the Asn680Asn, 48.9% for the Asn680Ser and 21.5% for the Ser680Ser variant. Similar results were reported by Perez Mayorga et al. (2000) who focused on an ovulatory population (n = 161) with infertility caused by male or tubal factor. Ethnic background was not reported, but from another study performed by the same group (Behre et al., 2005), which in part employed the same population, it appeared that this study population was largely of Caucasian origin. Data on Caucasian women with no known fertility problems and regular menstrual cycles showed a frequency distribution not very different from ours (Greb et al., 2005). This suggests that there is no obvious difference in FSHR polymorphism variant distribution between fertile and infertile populations. In accordance, Laven et al. (2003) reported no difference in the genotype distribution of the FSHR (combination of the SNP) between fertility patients and controls. Nevertheless, they found a difference in the distribution of both SNP at position 307 and 680 between anovulatory patients and controls. This finding was confirmed and extended by Daelemans et al. (2004) who reported a significant difference in both the allelic frequencies as well as the genotypic frequencies between fertility patients and controls. Data were obtained from Caucasian subjects but also for Asian women a significant difference in the Asn680Ser receptor variant frequency between fertility patients and controls was found (Sudo et al., 2002). The present study s data indicate that, when comparing allelic frequency distributions of the FSHR polymorphism variants, one should account at least for ethnic background. Given the data from the literature presented, it would also be advisable to distinguish between fertile and infertile patients. Remarkably, this large series of data could not confirm the previous observations that the Ser680Ser receptor variant is associated with higher early follicular-phase FSH concentrations (Falconer et al., 2005; Laven et al., 2003; Sudo et al., 2002). The uncorrected data contained women with the Ser680Ser receptor variant who had significantly higher FSH concentrations compared with the Asn680Ser genotypes, but after correction for age and oestradiol concentrations, this difference was not significant. Therefore, the variance in outcomes between this study and others might be caused by correction for oestradiol concentrations and age. Because of the continuing feedback system in which FSH and oestradiol function, it is necessary that reported FSH concentrations are always corrected for oestradiol concentrations. Furthermore, this population has been analysed as one group irrespective of the influence of the specific causes of infertility on FSH concentrations, which could be a confounder within these analyses. It is concluded that, within this population, the Ser680- Ser receptor variant is less common in the Asian subpopulation compared with Caucasian and Mediterranean subpopulations. No difference in FSH concentrations between either the various FSHR polymorphisms or the different ethnic groups was found after correction for age and oestradiol concentrations. Acknowledgements The authors would like to thank Dr CBM Oudejans for his useful comments on the manuscript. References Ahda, Y., Gromoll, J., Wunsch, A., et al., 2005. Follicle-stimulating hormone receptor gene haplotype distribution in normozoospermic and azoospermic men. J. Androl. 26, 494 499. Behre, H.M., Greb, R.R., Mempel, A., et al., 2005. Significance of a common single nucleotide polymorphism in exon 10 of the follicle-stimulating hormone (FSH) receptor gene for the ovarian response to FSH: a pharmacogenetic approach to controlled ovarian hyperstimulation. Pharmacogenet. Genomics 15, 451 456. Daelemans, C., Smits, G., de, M.V., et al., 2004. Prediction of severity of symptoms in iatrogenic ovarian hyperstimulation syndrome by follicle-stimulating hormone receptor Ser680Asn polymorphism. J. Clin. Endocrinol. Metab. 89, 6310 6315. de Koning, C.H., Benjamins, T., Harms, P., et al., 2006. The distribution of FSH receptor isoforms is related to basal FSH levels in subfertile women with normal menstrual cycles. Hum. Reprod. 21, 443 446. Falconer, H., Andersson, E., Aanesen, A., Fried, G., 2005. Folliclestimulating hormone receptor polymorphisms in a population of infertile women. Acta Obstet. Gynecol. Scand. 84, 806 811. Ferlin, A., Arredi, B., Foresta, C., 2006. Genetic causes of male infertility. Reprod. Toxicol. 22, 133 141. Gleicher, N., Weghofer, A., Li, J., Barad, D., 2007. Differences in ovarian function parameters between Chinese and Caucasian oocyte donors: do they offer an explanation for lower IVF pregnancy rates in Chinese women? Hum. Reprod. 22, 2879 2882. Greb, R.R., Grieshaber, K., Gromoll, J., et al., 2005. A common single nucleotide polymorphism in exon 10 of the human follicle stimulating hormone receptor is a major determinant of length and hormonal dynamics of the menstrual cycle. J. Clin. Endocrinol. Metab. 90, 4866 4872. Gromoll, J., Simoni, M., 2005. Genetic complexity of FSH receptor function. Trends Endocrinol. Metab. 16, 368 373. Hermann, B.P., Heckert, L.L., 2007. Transcriptional regulation of the FSH receptor: new perspectives. Mol. Cell. Endocrinol. 260 262, 100 108. Krzanowski, W.J., 2000. Principles of Multivariate Analysis. Clarendon Press, Oxford. Laven, J.S., Mulders, A.G., Suryandari, D.A., et al., 2003. Folliclestimulating hormone receptor polymorphisms in women with normogonadotropic anovulatory infertility. Fertil. Steril. 80, 986 992. Overbeek, A., Kuijper, E.A., Hendriks, M.L., et al., 2009. Clomiphene citrate resistance in relation to follicle-stimulating hormone receptor Ser680Ser-polymorphism in polycystic ovary syndrome. Hum. Reprod. 24, 2007 2013. Palep-Singh, M., Picton, H.M., Vrotsou, K., Maruthini, D., Balen, A.H., 2007. South Asian women with polycystic ovary syndrome S64 Reprinted from Vol. 20, No. 5 (2010) pp. 588 593.
FSHR gene polymorphisms and ethnicity 593 exhibit greater sensitivity to gonadotropin stimulation with reduced fertilization and ongoing pregnancy rates than their Caucasian counterparts. Eur. J. Obstet. Gynecol. Reprod. Biol. 134, 202 207. Perez Mayorga, M., Gromoll, J., Behre, H.M., Gassner, C., Nieschlag, E., Simoni, M., 2000. Ovarian response to follicle-stimulating hormone (FSH) stimulation depends on the FSH receptor genotype. J. Clin. Endocrinol. Metab. 85, 3365 3369. Simoni, M., Gromoll, J., Hoppner, W., et al., 1999. Mutational analysis of the follicle-stimulating hormone (FSH) receptor in normal and infertile men: identification and characterization of two discrete FSH receptor isoforms. J. Clin. Endocrinol. Metab. 84, 751 755. Simoni, M., Nieschlag, E., Gromoll, J., 2002. Isoforms and single nucleotide polymorphisms of the FSH receptor gene: implications for human reproduction. Hum. Reprod. Update 8, 413 421. Sudo, S., Kudo, M., Wada, S., Sato, O., Hsueh, A.J., Fujimoto, S., 2002. Genetic and functional analyses of polymorphisms in the human FSH receptor gene. Mol. Hum. Reprod. 8, 893 899. Wunsch, A., Ahda, Y., Banaz-Yasar, F., et al., 2005. Singlenucleotide polymorphisms in the promoter region influence the expression of the human follicle-stimulating hormone receptor. Fertil. Steril. 84, 446 453. Declaration: The authors report no financial or commercial conflicts of interest. Received 13 May 2009; refereed 10 August 2009; accepted 17 December 2009. Reprinted from Vol. 20, No. 5 (2010) pp. 588 593. S65