Lack of replication of four candidate SNPs implicated in human male fertility traits: a large-scale population-based study

Abstract

Introduction

Many cases of male infertility are caused by spermatogenic failure such as azoospermia, oligozoospermia or asthenozoospermia; in addition, decreased semen quality can also result in an elevated risk of male infertility. However, the genetic determinants for human semen quality are poorly understood.

To date, four genome-wide association studies (GWASs) regarding male fertility and infertility have been reported. A pilot GWAS in Caucasians (92 cases and 80 controls) showed that 20 single nucleotide polymorphisms (SNPs) were significantly (P < 10⁻⁵) associated with azoospermia or oligozoospermia (Aston and Carrell 2009). Furthermore, two GWASs in Chinese men have revealed common variants located near PRMT6 (which encodes protein arginine N-methyltransferase 6), PEX10 (which encodes peroxisome biogenesis factor 10), and SOX5 (which encodes SRY related HMG-box gene 5) and within the HLA region that are associated with risk for nonobstructive azoospermia (Hu et al., 2012; Zhao et al., 2012). The findings from these two Chinese GWASs have been evaluated in independent Japanese cohorts (Jinam et al., 2013; Sato et al., 2013). Lastly, a GWAS of 269 married Hutterite men in the USA, a culture that traditionally proscribes contraception and uniformly desires large families, revealed 41 SNPs that are significantly correlated with family size or birth rate (P < 1 × 10⁻⁴). In the subsequent validation study using 123 ethnically diverse men composed mainly of Hispanics and African Americans, nine of the 41 SNPs were also reported to be associated with reduced sperm quantity and/or function (Kosova et al., 2012). The associations of these nine SNPs with reduced fertility and sperm parameters remain to be confirmed in additional, larger cohorts.

Four of the nine SNPs detected in the GWAS for male fertility traits were found to be associated with sperm concentration, semen volume, total sperm count, total motile sperm count, and/or sperm motility (Kosova et al., 2012; also see Table I), and they are thought to be common in the Japanese population because of their minor allele frequencies (MAFs) > 0.05 in the HapMap-JPT population. In this study, to further clarify the contribution of these four SNPs to semen quality in diverse populations, we conducted a replication study to assess whether the four SNPs were associated with sperm parameters in two large Japanese cohorts.

Materials and Methods

This study was approved by the ethics committees of the University of Tokushima and St. Marianna Medical University. All participants provided written informed consent.

Two Japanese cohort samples

Two Japanese cohorts, namely, 791 men of proven fertility and 1224 young men from the general population, were included in the replication study. Some of the subjects in this study have been described in previous reports (Iwamoto et al., 2013a,b). Briefly, fertile men were recruited from the partners of pregnant women who attended obstetric clinics in four cities in Japan (Sapporo, Kanazawa, Osaka and Fukuoka) (Iwamoto et al., 2013a). The eligibility criteria for the male participants were as follows: the participants had to have been aged 20–45 years at the time of invitation by the hospital at which they were recruited, and both the man and his mother had to have been born in and living in Japan. In addition, the pregnancy of the female partner had to have been the result of conception by sexual intercourse and not by fertility treatment. Young men from the general Japanese population were recruited from university students in three study centers based in the urology departments at university hospitals in Japan (Kawasaki, Kanazawa and Nagasaki), as previously reported (Iwamoto et al., 2013b). In addition, we recruited university students at a study center in Sapporo. Inclusion criteria were that the man was 18–24 years old and that both he and his mother had been born in Japan.

Physical examination and semen analysis in the two cohorts

Age, body weight, height and ejaculation abstinence period were self-reported. BMI (kg/m²) was calculated from body weight and height. Semen samples were obtained and analyzed as previously described (Iwamoto et al., 2013a,b). Briefly, semen samples were obtained once by masturbation after sexual abstinence for at least 48 h and were ejaculated into clean, wide-necked, sterile, nontoxic collection containers. The samples were protected from extremes of temperature and were then liquefied at 37°C prior to their examination. The sperm concentration of each sample was assessed using a Bürker-Türk hemocytometer. Semen volume was measured with a graduated 5-ml syringe (Terumo; Tokyo, Japan). Sperm motility was assessed from 10 μl of well-mixed semen placed on a clean glass slide, covered, and then examined at a total magnification of 400× at 37°C. The motility assessment was repeated twice, and the average value from two samples was calculated. The sperm were assessed using the World Health Organization (WHO) motility classes A, B, C and D (World Health Organization, 1999). In this study, sperm in classes A and B were considered as motile.

SNP selection and genotyping

Genomic DNA was extracted from the peripheral blood samples of subjects using a QIAamp DNA blood kit (Qiagen; Tokyo, Japan). From SNPs previously reported to show association with sperm concentration, semen volume, total sperm count, total motile sperm count, and/or sperm motility (Kosova et al., 2012), four SNPs (rs7867029, rs12870438, rs7174015 and rs724078) with MAFs > 0.05 in the HapMap-JPT population were selected for genotyping. These four SNPs were reportedly associated with two or more of the five sperm parameters of interest in this study at permutation-based P-values <0.05 (Table I). The rs12870438 SNP was detected by restriction fragment length polymorphism PCR using the following primer sets: 5′- GCAAACAGGAGAAGGGTGTT -3′ (forward) and 5′- GCTTTGGAGCATGTTTTCCC -3′ (reverse). DNA from each subject was amplified using Taq DNA polymerase (Promega; Tokyo, Japan) under the appropriate amplification conditions. The resulting PCR products were then digested using the HhaI restriction enzyme (New England Biolabs Japan, Inc.; Tokyo, Japan). The digested products were separated by electrophoresis on a 2.5% agarose gel. The following fragment sizes were used for allele identification on gels: 488 bp (A-allele) and 278 + 210 bp (G-allele). The rs7174015, rs724078 and rs7867029 SNPs were genotyped using TaqMan probes rs7174015 (C_32072246_10), rs724078 (C_2500858_10) and rs7867029 (C_31364474_20; Applied Biosystems; Tokyo, Japan) with the ABI 7900HT real-time PCR system (Applied Biosystems).

Statistical analysis

Hardy–Weinberg equilibrium (HWE) was assessed in two cohort samples by using an internet-based HWE calculator (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl).

The analyses for sperm concentration, semen volume, total sperm number and total motile sperm number were processed using square-root-transformed values to minimize deviation from a normal distribution. The associations between SNPs and semen parameters were assessed using multiple linear regression with adjustments for age, BMI and ejaculation abstinence in each of the two cohorts. Sperm motility and total motile sperm number were additionally adjusted for time from masturbation to test. The results from the two cohorts were combined in a meta-analysis using the meta package for the R version 3.0.2 statistical environment (http://www.R-project.org/). The extent of heterogeneity among studies was quantified by the I² statistic (Higgins et al., 2003) and statistically assessed by the Cochran's Q test. If there was no heterogeneity, as determined by the I² statistic <50% or a P-value >0.1, a fixed-effects model using the inverse variance method was used. Otherwise, the random-effects model using the DerSimonian and Laird method was employed.

All statistical analyses were performed using R version 3.0.2 (The R Project for Statistical Computing [http://www.r-project.org]). For replication purposes, only SNP-trait associations observed in the previous GWAS (Kosova et al., 2012) were tested assuming the specific genetic models reported (n = 11 tests; Table I). Statistical significance was considered at P-values <0.0045 (0.05/11) to account for multiple testing.

Results

Semen characteristics of the two cohorts

The characteristics of semen from fertile Japanese men and from young men from the general Japanese population are presented in Supplementary Table SI. As previously reported (Iwamoto et al, 2013b), except in the case of sperm motility, semen parameters for men from the general population were significantly lower than those for fertile men.

Association analysis of four SNPs and semen parameters in fertile men, and young men from the general population in Japan

To investigate the associations between the four SNPs (rs7867029, rs12870438, rs7174015 and rs724078) and semen parameters we genotyped these SNPs in a total of 2015 men. The allele and genotype frequencies of the four SNPs analyzed in each cohort are shown in Table II. The genotyping of the SNPs is complete except for rs12870438 (the missing genotyping rate is 0.1%), and the genotypes of all four SNPs were in HWE in the respective two cohorts (P >0.05). Then, we assessed the associations between the four SNPs and semen parameters using a multiple linear regression analysis of the two cohorts. In this study, we performed an association analysis with semen parameters that were related to the minor allele in Hutterites, as reported in a previous GWAS under the association model (Kosova et al., 2012). Multiple linear regression analysis revealed that rs7867029 showed a trend toward a negative association with sperm motility (β = −1.98, P = 0.026) in young men from the general Japanese population, and rs12870438 showed a trend toward a positive association with total sperm numbers (β = 7.80, P = 0.028) in fertile men (Table III). However, none of the four SNPs reached the adjusted P-value for multiple testing (P < 0.0045). Next, to assess the strength of the association, we conducted a combined analysis using a meta-analysis of the two Japanese male cohorts. However, unlike the results of the previous study (Kosova et al., 2012), none of the four SNPs displayed a significant association with semen parameters. Furthermore, there were no associations observed between the four SNPs and other semen parameters in three genetic models (additive, Supplementary Table SII; recessive, Supplementary Table SIII; and dominant, Supplementary Table SIV) in the combined analysis.

Discussion

Recently, 4 (rs7867029, rs12870438, rs7174015 and rs724078) of the 41 SNPs correlated with family size or birth rate (P < 1 × 10⁻⁴) in the GWAS of 269 Hutterite men in the USA were found to be associated with sperm concentration, semen volume, total sperm count, total motile sperm count or sperm motility in 123 ethnically diverse men from Chicago (Kosova et al., 2012). Additionally, we recently showed that of the four SNPs, rs7867029, rs7174015 and rs12870438 were significantly associated with the risk of developing oligozoospermia, and rs12870438 was also associated with azoospermia (Sato et al., submitted). In the present study, there was limited evidence of a significant association between these four SNPs and one or more of the five semen parameters in each of two Japanese replication cohorts, whereas none of the four SNPs displayed a significant association with any of the semen parameters in three genetic models in the meta-analysis of the two cohorts comprising a total of 2015 Japanese men. The current replication meta-analysis is well-powered to detect associations of semen quality trait loci with modest effect sizes because this provides >80% power for SNPs that explain 1% or higher of total phenotypic variance. The observed heterogeneity of the SNP-trait associations between a previous (Kosova et al., 2012) and this study, as well as between the two Japanese cohorts, may be attributed to potential biases in selection of the study subjects; in the previous study, most of the 123 subjects were Hispanic (58.5%) and had been referred for infertility evaluation at the University of Illinois Andrology Laboratory, Chicago, IL, USA, while 791 men of proven fertility and 1224 general controls were separately recruited for population-based assessment of semen quality in Japanese men.

Several limitations of this study should be noted. In this study, only the SNPs identified originally in the Hutterite GWAS were examined for associations with semen quality traits under three genetic models in a Japanese population. Owing to between-population differences in linkage disequilibrium (LD) structures around the SNPs examined, the tested SNPs may not be in high LD with unidentified true causal variants in Japanese subjects (Supplementary Figures S1–S4). The low LD between the genotyped SNPs and causal variants could increase the likelihood of false-negative findings, through the lowering of the statistical power of the analysis (Clarke et al., 2007). The differences in the extent of LD and the underlying haplotype structures between populations could also affect the fit of the specified genetic model to the data obtained, and the direction of the effect for the associated allele. This may account for apparent inconsistencies in the model fitting between previous and this studies. To overcome these limitations, fine-scale LD mapping of the fertility trait loci using a set of tagging SNPs across the loci will be necessary in populations with larger sample sizes. The locus mapping studies will allow for a better understanding of susceptibility genes contributing to semen quality in humans.

Authors' roles

Y.S. and A.Ta.: study design and data analysis; Y.S. and K.T.: genotyping; S.N., M.Y., E.K., J.K., M.N., K.M., A.Ts., K.K., N.I., J.E., and T.I.: cohort collection and characterization; Y.S., A.Ta., K.T., S.N., M.Y., E.K., J.K., M.N., K.M., A.Ts., K.K., N.I., J.E., I.I., A.Y., and T.I.: preparation and approval of the final version of the manuscript.

Funding

This study was supported in part by the Ministry of Health and Welfare of Japan (1013201) (to T.I.), Grant-in-Aids for Scientific Research (C) (23510242) (to A.Ta.) from the Japan Society for the Promotion of Science, the European Union (BMH4-CT96-0314) (to T.I.), and the Takeda Science Foundation (to A.Ta.).

Conflict of interest

None declared.

Acknowledgements

We thank all the volunteers who participated in this study. We are grateful to the late Prof. Yutaka Nakahori for collecting blood samples from the participants. We also thank Prof. Toyomasa Katagiri for his assistance with the AB GeneAmp PCR system 9700.

References