Human Molecular Genetics, 2012, Vol. 21, No. 2 doi:10.1093/hmg/ddr462 Advance Access published on October 11, 2011

430–436

Association of 8q22.3 locus in Chinese Han with idiopathic premature ovarian failure (POF) Yingying Qin1, Han Zhao1, Jianfeng Xu2, Yongyong Shi3, Zhiqiang Li3, Jie Qiao4, Jiayin Liu5, Chunrong Qin6, Chune Ren7, Jie Li8, Shiling Chen9, Yunxia Cao10, China POF Study Group{, Joe Leigh Simpson11 and Zi-Jiang Chen1,∗ 1

Center for Reproductive Medicine, Provincial Hospital Affiliated to Shandong University, National Research Center for Assisted Reproductive Technology and Reproductive Genetics, The Key Laboratory for Reproductive Endocrinology of Ministry of Education, Shandong Provincial Key Laboratory of Reproductive Medicine, 324 Jingwu Road, Jinan 250021, China, 2Fudan-VARI Center for Genetic Epidemiology, Fudan University, Shanghai 200433, China, 3 Bio-X Center, Shanghai Jiao Tong University, Shanghai 200240, China, 4Department of Obstetrics and Gynecology, Peking University Third Hospital, Beijing, Key Laboratory of Assisted Reproduction, Ministry of Education, Beijing 100191, China, 5The First Affiliated Hospital with Nanjing Medical University, Nanjing 210029, China, 6Maternal and Child Health Hospital, Shenzhen 518028, China, 7Center for Reproductive Medicine, Hospital Affiliated to Weifang Medical College, Weifang 261053, China, 8The First Affiliated Hospital, Sun Yat-sen University, Guangdong 510080, China, 9 Nanfang Hospital, Guangdong 510515, China, 10Reproductive Medicine Center, First Affiliated Hospital, Anhui Medical University, Hefei 230032, China and 11College of Medicine, Florida International University, Miami, FL 33199, USA Received July 26, 2011; Revised September 27, 2011; Accepted October 4, 2011

Premature ovarian failure (POF) is a complex heritable disorder known to be caused by chromosomal abnormalities and to date a limited number of known mutations, often autosomal. We sought to identify additional genetic loci associated with POF by performing the first large-scale genome-wide association study (GWAS). GWAS, using Affymetrix SNP 6.0 chip, was conducted in an initial discovery set of 391 well-documented (follicle-stimulating hormone >40 IU/ml) Chinese Han POF patients, compared with 895 unrelated Chinese female controls. A replication study on the most significant loci was then performed in an independent set of 400 cases and 800 controls. Suggestive significant associations were observed at 8q22.3. Replication of eight single-nucleotide polymorphisms (SNPs) (rs10464815, rs10808365, rs3847152, rs3847153, rs3847154, rs3843552, rs10955242, rs3843555) (P ≤ 3.86 3 1026) was confirmed in verification sets. No specific candidate gene was found in the immediate region of 8q22.3. This GWAS, involving by far the largest sample of POF cases accumulated to date, revealed heretofore unrecognized association between POF and a novel genetic locus or region of unknown nature on 8q22.3. We speculate existence of a long-distance regulatory region that has relevance to the control of ovarian differentiation or oogenesis. Given failure to find association with any of the other autosomal regions known to harbor genes causing ovarian failure, our findings also underscore the likelihood of considerable genetic and etiologic heterogeneity in POF and the need for additional approaches like whole-genome sequencing.

INTRODUCTION Premature ovarian failure (POF) is defined as cessation of menstruation before the expected age of menopause.

Diagnosis is confirmed by elevated serum follicle-stimulating hormone (FSH) levels (.40 IU/l) prior to the age of 40 years. Approximately 1% of the Caucasian population is affected

∗

To whom correspondence should be addressed at: 324 Jingwu Road, Jinan 250021, China. Tel/Fax: +86 53187923887; Email:chenzijiang@hotmail. com The members of the China Premature Ovarian Failure Study Group are listed in Supplementary Material.

†

# The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

Human Molecular Genetics, 2012, Vol. 21, No. 2 before age 40; ≤0.1% are affected prior to age 30 and 0.01% ,20 years of age (1). Unfortunately, there are no accurate incidence data concerning POF in Chinese population. Diverse etiologies have been claimed, including infection, chemotherapeutic treatments, pelvic surgery and autoimmune diseases; however, genetic factors are generally considered paramount. Heritability has been well established in POF, and many genetic causes recognized (2). Loci on the X chromosome and on autosomes are pivotal for ovarian development and oogenesis. Deletions resulting in loss of certain regions on the X chromosome have long been accepted as causative. Terminal deletions arising at Xp11 or Xq13 more often cause complete than POF, whereas more distal deletions (Xp21  pter; Xq27  qter) usually cause POF (3). Interstitial X-deletions are also reported (4). Reciprocal (X-autosomal; autosomal) translocations explain certain cases, but whether reflective of a perturbed gene or reflective of non-specific meiotic breakdown remains unclear. Certain regions of the chromosome (designated POF1, POF1B, POF2 and POF3) have been identified for association with POF, but these associations are now realized as possibly unreliable and further do not unambiguously identify causal genes (5 – 7). Copy-number-variant duplications as well as deletions have also unexpectedly been observed in near-equal numbers (8,9), although usually without parental studies that are necessary to exclude familial transmission indicative of a nonfunctional familial polymorphism. Among several X-candidate genes, only premutation involving the fragile X FMR1 locus (Xq27) explains a significant minority of cases, 1 – 5% sporadic and only still slightly more familial cases (10,11). Autosomal genes play significant roles in ovarian development than traditionally believed. In fact, the autosomal gene FOXL2 is considered the locus most pivotal for female sex differentiation (12). Ovarian failure is one component of many syndromes caused by mutant autosomal genes. Here we focus on nonsyndromic (isolated) POF, for which several dozen autosomal genes have already been subjected to detailed candidate gene analysis by us and others. Except for FSHR in Finnish women (13), none explains .1 – 2% of POF in any given ethnic group. Other genes interrogated by candidate gene approaches and considered perturbed include TGF family members like BMP15 and GDF9 (14,15), inhibin A (16), homeobox genes like NOBOX (17), transcription factors FIGLA (18) and POU5F1 (19), meiotic genes MSH5 and DMC1 (20), and RFLP4 (21). We failed to show perturbations in the RNA-binding protein NANOS3 (22), the G protein receptor GPR3 (23) and homeobox LHX8 (24). Other groups found no perturbations in nonsyndromic POF when interrogating PTEN (25), NOG (26), KIT (27), FOXO1A (28) and AMH (29). A single FOXL2 mutation was found by Harris et al. (30) among 290 interrogated nonsyndromic POF cases (31). In aggregate, no more than 10% of POF cases can be explained etiologically. Given the above, POF becomes an ideal disorder to be subjected to genome-wide association studies (GWASs) for the detection of significant autosomal loci. In contrast to common adult onset disorders (e.g. bipolar disorder, schizophrenia), POF has the advantage of being defined by crisp laboratory-based diagnostic criteria. In normal women FSH is usually ,10 IU/l; levels exceeding 10.8– 17 IU/l are

431

Figure 1. Principal component analysis of GWAS sample. The plot presents the top two eigenvectors identified by principal component analysis using the Eigenstrat software.

indicative of diminishing ovarian reserve (32), whereas levels .40 IU/l invariably mean ovarian failure. As oocyte depletion begins to occur, FSH levels rapidly rise beyond 10 IU/l to reach 40 IU/l or higher, levels never observed in normal menstruating women. Thus, a distinct quantitative end point applies for the diagnosis of ovarian failure. Only two GWAS reports exist (Supplementary Material, Table S1). One showed association with PTH-responsive B1 gene (PTHB1) in a small discovery set of 24 women and 24 controls (33); the other showed association with ADAMTS19 in the discovery set of 99 Dutch women and 181 controls (34), although not in the replication set. In addition, 5q14.1 – q15 was found to be a susceptibility region in a single Dutch POF family subjected to genome-wide linkage scan (35). However, in no study did sample size exceed 100, all studies thus lacking statistical power sufficient to detect a modest association while evaluating hundreds of thousands of SNPs. We have thus conducted a GWAS using Affymetrix SNP 6.0 chip to identify predisposing genetic risk factors. Our report utilized a well-defined cohort encompassing in aggregate 791 POF patients and 1695 controls, all Chinese Han.

RESULTS Genome-wide association study Principal component analysis shows that the genetic substructure was comparable between our cases and controls (Fig. 1). Similarly, the Q-Q plot of adjusted P-value indicates no systematic bias, with an inflation factor of 1.019 (Fig. 2). The most significant locus was found to be located on 8q22.3. This locus includes eight SNPs in linkage disequilibrium (r 2 ¼ 0.98); P-values ranged from 1.6 × 1026 to 3.86 × 1026 (Table 1). We did not observe any SNP that reached a genome-wide significant level at 5 × 1028 (Fig. 3) in the

0.7237 0.7067 0.7198 0.7020 0.7199 0.7205 0.7189 0.7200 1.15E 2 06 1.71E 2 07 5.98E 2 07 8.85E 2 08 6.13E 2 07 7.12E 2 07 5.45E 2 07 6.08E 2 07 0.8177 0.7807 0.8083 0.7703 0.8003 0.8113 0.806 0.7958 0.0333 0.0087 0.0219 0.0053 0.0164 0.0254 0.0199 0.0139 0.3583 0.3578 0.3571 0.3583 0.3602 0.3601 0.3596 0.3593 0.3135 0.3031 0.3098 0.3008 0.3106 0.3135 0.3116 0.3086

a

101 877 327 101 879 667 101 882 250 101 882 332 101 882 354 101 882 444 101 882 806 101 884 409

F_A and F_U indicate the allele frequency in cases and controls, respectively. The Bonferroni correction was based on 419 326 SNPs that passed the QC.

0.6424 0.6409 0.6401 0.6401 0.6466 0.6401 0.6398 0.6507 2.00E 2 06 1.76E 2 06 1.64E 2 06 1.64E 2 06 2.88E 2 06 1.64E 2 06 1.68E 2 06 3.86E 2 06 0.375 0.3756 0.375 0.375 0.3726 0.375 0.3744 0.3727

GWAS Case (n ¼ 391), control (n ¼ 895) F_A F_U P-value A2 A1

rs10464815 rs10808365 rs3847152 rs3847153 rs3847154 rs3843552 rs10955242 rs3843555

In the POF susceptibility region discovered—8q22.3—no known gene match exists. This region has not previously been related to POF, but it is noteworthy that two reports have described ovarian dysgenesis associated with balanced autosomal translocations involving chromosome 8, regions q22.1 and p11.2, respectively (36,37). Chromosomal rearrangement involving 8q22 is in particular consistent with

BP

DISCUSSION

SNP

In addition to the association test with the replication set, we also carried out a combined analysis with GWAS data using meta-analysis. The Mantel– Haenszel method was used to estimate the pooled odds ratio for all strata as a fixed effects model. After combining the GWAS and replication data, the P-value of one SNP (rs3847153) almost reached genome-wide significant level (8.85 × 1028). The combined P-values for the remaining seven SNPs range from 1.15 × 1026 to 1.71 × 1027. Results of the combined analysis strongly support the variants at 8q22.3 being associated with susceptibility to POF.

Table 1. The meta-analysis of the association of POF with SNPs on 8q22.3

Combined analysis

Odds ratio

To further validate the GWAS findings, we chose a smaller subset of SNPs for further replication in an independent sample set including 400 cases and 800 controls. The association results of the eight SNPs genotyped in the replication stage are presented in Table 1. All of the eight SNPs are significantly associated with POF (P , 0.05) and with the same direction of association.

0.2782 0.2782 0.2775 0.2775 0.2775 0.2775 0.2769 0.2788

Replication

A T C A C G G T

GWAS stage, and all SNPs with P , 5 × 1026 mapped to 8q22.3 (Fig. 4).

C C T G T T A C

Replication Case (n ¼ 400), control (n ¼ 800) F_A F_U P-value Odds ratio

Figure 2. Q– Q plots showing the distribution of expected compared with observed –log10 P-values for the association test results adjusting for population structure.

0.4839 0.0717 0.2510 0.0371 0.2569 0.2984 0.2285 0.2547

Human Molecular Genetics, 2012, Vol. 21, No. 2

Meta-analysis Case (n ¼ 791), control (n ¼ 1695) P-value Odds ratio P-valuea (B-corr.)

432

Human Molecular Genetics, 2012, Vol. 21, No. 2

433

Figure 3. Genome-wide association scan for POF. Negative log10 unadjusted P-values are shown for SNPs that passed quality control.

Figure 4. Regional plots of 8q22 region. X-axis shows chromosomal positions. Y-axis shows – log10 P-values from the association test. Colors of the dots indicate the LD relationship between the top associated SNPs (shown in the plots) and remaining markers in the 200 kb upstream and downstream of the top associated SNP. The right Y-axis shows the recombination rate. Both the LD and recombination rates are estimated based on the Hapmap Phase II CHB+JPT population.

our GWAS findings because molecular perturbation could have occurred and led to the deletion of a causative gene. Aboura et al. (38) and Ledig et al. (39) used array CGH to detect copy number variants, finding no changes at 8q22.3 but deletions (8q24.3 and 8q24.13) in ovarian failure cases and duplications (8q11.23 and 8q12.1) in POF cases. However, parental transmission could not be excluded; thus, significance of these findings remains unclear. We speculate that the 8q22.3 region we have discovered plays an important regulatory role in oogenesis. Examples of long-distance regulatory effects exist, such as 17q24 – q25, the region containing

SOX9. Haploinsufficiency of SOX9 causes XY sex reversal (male to female) (40), whereas duplication causes XX sex reversal (female to male) (41). Disturbance of a region up to 1 Mb upstream of SOX9 precludes action of SOX9 in both mice (42) and humans (43). That annotated genes do not exist in the immediate region of 8q22.3 thus does not exclude a significant effect on oogenesis. A great strength of our study is its well-defined inclusion criteria for both POF subjects (FSH . 40IU/l) and controls (menstruating). Misclassification is thus very unlikely to confound our results. Our discovery and replication sets also presented at a younger age than POF cohorts in other studies (Table 2). This may explain our finding—a significant association with only one SNP (rs6279 in ANKK1) found in the four reported GWAS related to age at natural menopause and early menopause (44– 47) (Supplementary Material, Table S2). Our study was further restricted to Chinese Han, minimizing misinterpretation due to ethnic stratification. That significance of association using GWASs was not greater probably reflects the known genetic heterogeneity, autosomal and X chromosomal, accepted as existing in POF (2). Despite the attractiveness of GWASs for POF, genome searches in a sufficiently large number of patients are practically absent in POF, doubtless reflecting rarity of POF. An unavoidable limitation of our study is its sample size. Although we recruited by far the largest population-based cohort of POF to date, our study still lacked the sample size of thousands required to provide sufficient statistical power to detect modest associations of autosomal loci. However, this limitation becomes almost unavoidable because POF is rare compared with common disorders like type 2 diabetes mellitus or hypertension more typically subjected to GWASs. Our study lacked optimal size in the initial stage to detect other

434

Human Molecular Genetics, 2012, Vol. 21, No. 2

Table 2. Clinical features of cases in discovery and replication sets, which are not significantly different (P . 0.05) Characteristic Cases Number of cases FSH concentrationa (IU/l) Familial clusteringb Mother affected Menopause , 40 years Menopause , 45 years Siblings affected Menopause , 40 years Age at diagnosisa (years) Amenorrhea primary Amenorrhea secondary Age at menarchea,∗ (years) Age at onset of menstrual dysfunctiona (years) Age of amenorrheaa (years) Controls Number of controls Agea (years) Age at menarchea,∗ (years)

Discovery set

Replication

391 78.40 + 28.44 41 39 9 30 2 2 30.28 + 4.46 32 359 14.71 + 2.17 23.41 + 6.03

400 73.88 + 30.23 32 27 13 14 5 5 29.79 + 5.31 29 228 14.82 + 2.11 23.85 + 6.78

24.85 + 6.07

24.87 + 5.88

895 30.68 + 4.68 14.52 + 1.42

800 38.77 + 11.58 14.75 + 1.45

a

+SD. Defined as two or more first- or second-degree female family members with either POF or early menopause (,45 years old), including the index POF patient. ∗ P . 0.05 (cases versus controls by Student’s t-test). b

novel genetic determinants having smaller effects on POF. Given that POF is a rare phenotype, it is possible that point mutations or copy number variants might contribute to the etiology of POF. Indeed, in previous studies, we had sought associations of rare variants with POF at various candidate genes (14,17,18). Future research may elucidate the potential contributions of rare alleles to POF also at 8q22.3. Providing the etiologic (genetic) explanation for women with POF would be valuable clinically. Identifying genetic causation may further be of diagnostic value to other family members, especially female offspring. If a younger relative has the same genetic perturbation, timeline for achieving pregnancy might be advanced and early child-bearing encouraged. Cryopreservation of oocytes or ovarian slices can be envisioned. Of course, the ultimate goal is generating a therapeutic gene product to replace a deficiency or alternatively ameliorate the deleterious effect of a dominant negative mutation. This GWAS study of POF, involving by far the largest sample studied to date, used SNP arrays to show for the first time that a region containing eight susceptibility SNPs on 8q22.3 was involved in POF. We speculate existence of an important yet undefined long-distance regulatory region affecting oogenesis. Identification of risk alleles on 8q22.3 justifies further investigations to better define the role played by this region and to identify populations at high risk for POF.

MATERIALS AND METHODS

regional referral centers in China. Samples were collected from 18 provinces during 2002 – 10. Of the 391, 32 had primary amenorrhea, but were included given common ascertainment and given that in our sample the age of secondary amenorrhea was young (mean age 24.85 + 6.07 years) (Table 2). Our sample thus differed from others in which the mean age of POF was considerably older (Supplementary Material, Table S1). In our discovery set, 10.4% (41/391) of cases had a positive family history, i.e. at least two first- or second-degree female family members with POF or early menopause (,45 years of age), including the index POF patient. Most of the affected relatives (39/41; 9 ,40 years and 30 ,45 years) were mothers of the index patient; the remaining two were sisters. In the replicate cohort, 13/400 (3.3%) had mothers with POF (n ¼ 13) or early menopause (n ¼ 14) (Table 2). Inclusion criteria consisted of cessation of menstrual cycles before 40 years of age, with at least two serum FSH concentrations exceeding 40 IU/l. Women with known chromosomal abnormalities, previous ovarian surgery, chemotherapy or radiotherapy were excluded. Information was sought on associated somatic anomalies, specifically those previously associated with POF [e.g. neurosensory deafness (one case), blepharophimosis (no case)]. Somatic anomalies proved uncommon (,0.2%). In the interest of facilitating future meta-analysis, we neither excluded these few cases nor tested routinely for FMR1 permutations. Either direct visualization (laparoscopy) or ultrasound imaging was performed in all cases, predictably showing smaller than expected ovaries. The controls attended the clinic because of tubal factor infertility. They were ascertained by hormone assays and ultrasound imaging to be free of POF-related phenotypes. They were known to be menstruating regularly with normal FSH concentration. Thus, misclassification should be extremely low. The replication cohorts comprised an independent set of Chinese Han POF (n ¼ 400) and matched controls (n ¼ 800), which were ascertained the same way as the discovery set. The controls shared the same geographic origin as the cases. Informed consent was obtained from all subjects. The study was approved by the Institutional Review Board of Reproductive Medicine of Shandong University. GWAS genotyping and quality control Genome-wide genotyping was performed using the Affymetrix Genome-Wide Human SNP Array 6.0. Genotype data were called using the Birdseed algorithm (48). We applied the following criteria to perform quality control. For samples, a call rate of 0.95 was used to filter individuals. For SNP filtering (after sample filtering), SNPs with call rates ,95% in either case or control samples were removed. SNPs with MAF (minor allele frequency) ,1%, or deviated significantly from Hardy – Weinberg equilibrium (P ≤ 0.001), in controls were excluded. A total of 642 069 SNPs passed the quality criteria and were used for the subsequent analysis.

Participants and phenotypes Our GWA study was performed initially on 391 unrelated idiopathic POF patients and 895 female controls, all Chinese Han collectively gathered from a number of high-volume

SNP selection for validation and replication genotyping SNPs with a P-value ,5 × 1026 in the GWAS stage were selected. SNPs with unambiguous genotype scatter plots

Human Molecular Genetics, 2012, Vol. 21, No. 2

were excluded from the subsequent stage. We used the Sequenom MassARRAY iPLEX platform for genotyping in the replication stage. The Sequenom protocol involves a multiplex PCR reaction prior to a single-base primer extension reaction. The individual SNPs are identified by using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Statistical analysis The EIGENSTRAT software was applied for the evaluation of population stratification by a principal component approach (49). The top two eigenvectors were plotted in Figure 1 and were adjusted as covariates in the logistic regression analysis. Genome-wide association analysis of a single marker was performed using the PLINK software package (50). The associations of the SNPs with POF status were evaluated using a logistic regression model, assuming additive effects of the alleles (0, 1 and 2). Haploview was used to generate the Manhanttan plot (Fig. 3) (51). The quantile – quantile (Q– Q) plot (Fig. 2) was created using the R qq.plot function. The regional plot was generated using LocusZoom. For the replication study, allelic association analysis of the replication sample was conducted using SHEsis (52). The GWAS and replication data were combined using meta-analysis with the R ‘meta’ package. The Mantel– Haenszel method was used to calculate the fixed effect estimate (53).

SUPPLEMENTARY MATERIAL Supplementary Material is available at HMG online.

ACKNOWLEDGEMENTS We thank all the participants involved in this study. We also particularly thank the China POF Study Group for patients’ recruitment. Conflict of Interest statement. No potential conflict of interest relevant to this article was reported.

FUNDING Supported by grants from the National Basic Research Program of China (973 Program – 2012CB944700, 2010CB945002, 2007CB947403, 2007CB948103), the National Natural Science Foundation of China (30973170, 81000236, 81000238), science research foundation item of no-earnings health vocation from the Ministry of Public Health (201002013) and Shandong Major Scientific Innovation Foundation (2006GG1102021).

REFERENCES 1. Coulam, C.B., Adamson, S.C. and Annegers, J.F. (1986) Incidence of premature ovarian failure. Obstet. Gynecol., 67, 604– 606. 2. Simpson, J.L. (2008) Genetic and phenotypic heterogeneity in ovarian failure: overview of selected candidate genes. Ann. N. Y. Acad. Sci., 1135, 146–154. 3. Simpson, J.L. and Rajkovic, A. (1999) Ovarian differentiation and gonadal failure. Am. J. Med. Genet., 89, 186– 200.

435

4. Ochalski, M.E., Engle, N., Wakim, A., Ravnan, B.J., Hoffner, L., Rajkovic, A. and Surti, U. (2011) Complex X chromosome rearrangement delineated by array comparative genome hybridization in a woman with premature ovarian insufficiency. Fertil. Steril., 95, e9 –e15. 5. Bione, S., Sala, C., Manzini, C., Arrigo, G., Zuffardi, O., Banfi, S., Borsani, G., Jonveaux, P., Philippe, C., Zuccotti, M. et al. (1998) A human homologue of the Drosophila melanogaster diaphanous gene is disrupted in a patient with premature ovarian failure: evidence for conserved function in oogenesis and implications for human sterility. Am. J. Hum. Genet., 62, 533– 541. 6. Lacombe, A., Lee, H., Zahed, L., Choucair, M., Muller, J.M., Nelson, S.F., Salameh, W. and Vilain, E. (2006) Disruption of POF1B binding to nonmuscle actin filaments is associated with premature ovarian failure. Am. J. Hum. Genet., 79, 113– 119. 7. Sala, C., Arrigo, G., Torri, G., Martinazzi, F., Riva, P., Larizza, L., Philippe, C., Jonveaux, P., Sloan, F., Labella, T. et al. (1997) Eleven X chromosome breakpoints associated with premature ovarian failure (POF) map to a 15-Mb YAC contig spanning Xq21. Genomics, 40, 123–131. 8. Dudding, T.E., Lawrence, O., Winship, I., Froyen, G., Vandewalle, J., Scott, R. and Shelling, A.N. (2010) Array comparative genomic hybridization for the detection of submicroscopic copy number variations of the X chromosome in women with premature ovarian failure. Hum. Reprod., 25, 3159–3160. 9. Quilter, C.R., Karcanias, A.C., Bagga, M.R., Duncan, S., Murray, A., Conway, G.S., Sargent, C.A. and Affara, N.A. (2010) Analysis of X chromosome genomic DNA sequence copy number variation associated with premature ovarian failure (POF). Hum. Reprod., 25, 2139– 2150. 10. Wittenberger, M.D., Hagerman, R.J., Sherman, S.L., McConkie-Rosell, A., Welt, C.K., Rebar, R.W., Corrigan, E.C., Simpson, J.L. and Nelson, L.M. (2007) The FMR1 premutation and reproduction. Fertil. Steril., 87, 456– 465. 11. Jacquemont, S., Hagerman, R.J., Hagerman, P.J. and Leehey, M.A. (2007) Fragile-X syndrome and fragile X-associated tremor/ataxia syndrome: two faces of FMR1. Lancet Neurol., 6, 45– 55. 12. Uhlenhaut, N.H., Jakob, S., Anlag, K., Eisenberger, T., Sekido, R., Kress, J., Treier, A.C., Klugmann, C., Klasen, C., Holter, N.I. et al. (2009) Somatic sex reprogramming of adult ovaries to testes by FOXL2 ablation. Cell, 139, 1130– 1142. 13. Aittomaki, K., Lucena, J.L., Pakarinen, P., Sistonen, P., Tapanainen, J., Gromoll, J., Kaskikari, R., Sankila, E.M., Lehvaslaiho, H., Engel, A.R. et al. (1995) Mutation in the follicle-stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell, 82, 959– 968. 14. Zhao, H., Qin, Y., Kovanci, E., Simpson, J.L., Chen, Z.J. and Rajkovic, A. (2007) Analyses of GDF9 mutation in 100 Chinese women with premature ovarian failure. Fertil. Steril., 88, 1474–1476. 15. Tiotiu, D., Alvaro Mercadal, B., Imbert, R., Verbist, J., Demeestere, I., De Leener, A., Englert, Y., Vassart, G., Costagliola, S. and Delbaere, A. (2010) Variants of the BMP15 gene in a cohort of patients with premature ovarian failure. Hum. Reprod., 25, 1581– 1587. 16. Prakash, G.J., Ravi Kanth, V.V., Shelling, A.N., Rozati, R. and Sujatha, M. (2010) Mutational analysis of inhibin alpha gene revealed three novel variations in Indian women with premature ovarian failure. Fertil. Steril., 94, 90–98. 17. Qin, Y., Choi, Y., Zhao, H., Simpson, J.L., Chen, Z.J. and Rajkovic, A. (2007) NOBOX homeobox mutation causes premature ovarian failure. Am. J. Hum. Genet., 81, 576– 581. 18. Zhao, H., Chen, Z.J., Qin, Y., Shi, Y., Wang, S., Choi, Y., Simpson, J.L. and Rajkovic, A. (2008) Transcription factor FIGLA is mutated in patients with premature ovarian failure. Am. J. Hum. Genet., 82, 1342– 1348. 19. Wang, J., Wang, B., Song, J., Suo, P., Ni, F., Chen, B., Ma, X. and Cao, Y. (2011) New candidate gene POU5F1 associated with premature ovarian failure in Chinese patients. Reprod. Biomed. Online, 22, 312– 316. 20. Mandon-Pepin, B., Touraine, P., Kuttenn, F., Derbois, C., Rouxel, A., Matsuda, F., Nicolas, A., Cotinot, C. and Fellous, M. (2008) Genetic investigation of four meiotic genes in women with premature ovarian failure. Eur. J. Endocrinol., 158, 107– 115. 21. Suzumori, N., Burns, K.H., Yan, W. and Matzuk, M.M. (2003) RFPL4 interacts with oocyte proteins of the ubiquitin-proteasome degradation pathway. Proc. Natl Acad. Sci. USA, 100, 550– 555. 22. Qin, Y., Zhao, H., Kovanci, E., Simpson, J.L., Chen, Z.J. and Rajkovic, A. (2007) Mutation analysis of NANOS3 in 80 Chinese and 88 Caucasian women with premature ovarian failure. Fertil. Steril., 88, 1465–1467.

436

Human Molecular Genetics, 2012, Vol. 21, No. 2

23. Kovanci, E., Simpson, J.L., Amato, P., Rohozinski, J., Heard, M.J., Bishop, C.E. and Carson, S.A. (2008) Oocyte-specific G-protein-coupled receptor 3 (GPR3): no perturbations found in 82 women with premature ovarian failure (first report). Fertil. Steril., 90, 1269–1271. 24. Qin, Y., Zhao, H., Kovanci, E., Simpson, J.L., Chen, Z.J. and Rajkovic, A. (2008) Analysis of LHX8 mutation in premature ovarian failure. Fertil. Steril., 89, 1012–1014. 25. Shimizu, Y., Kimura, F., Takebayashi, K., Fujiwara, M., Takakura, K. and Takahashi, K. (2009) Mutational analysis of the PTEN gene in women with premature ovarian failure. Acta Obstet. Gynecol. Scand., 88, 824–825. 26. Laissue, P., Christin-Maitre, S., Bouchard, P., Fellous, M. and Veitia, R.A. (2007) Mutations in the NOG gene are not a common cause of nonsyndromic premature ovarian failure. Clin. Endocrinol. (Oxf.), 66, 900. 27. Shibanuma, K., Tong, Z.B., Vanderhoof, V.H., Vanevski, K. and Nelson, L.M. (2002) Investigation of KIT gene mutations in women with 46,XX spontaneous premature ovarian failure. BMC Womens Health, 2, 8. 28. Watkins, W.J., Umbers, A.J., Woad, K.J., Harris, S.E., Winship, I.M., Gersak, K. and Shelling, A.N. (2006) Mutational screening of FOXO3A and FOXO1A in women with premature ovarian failure. Fertil. Steril., 86, 1518– 1521. 29. Wang, H.Q., Takakura, K., Takebayashi, K. and Noda, Y. (2002) Mutational analysis of the mullerian-inhibiting substance gene and its receptor gene in Japanese women with polycystic ovary syndrome and premature ovarian failure. Fertil. Steril., 78, 1329– 1330. 30. Harris, S.E., Chand, A.L., Winship, I.M., Gersak, K., Aittomaki, K. and Shelling, A.N. (2002) Identification of novel mutations in FOXL2 associated with premature ovarian failure. Mol. Hum. Reprod., 8, 729– 733. 31. Laissue, P., Vinci, G., Veitia, R.A. and Fellous, M. (2008) Recent advances in the study of genes involved in non-syndromic premature ovarian failure. Mol. Cell. Endocrinol., 282, 101– 111. 32. Bukman, A. and Heineman, M.J. (2001) Ovarian reserve testing and the use of prognostic models in patients with subfertility. Hum. Reprod. Update, 7, 581 –590. 33. Kang, H., Lee, S.K., Kim, M.H., Song, J., Bae, S.J., Kim, N.K., Lee, S.H. and Kwack, K. (2008) Parathyroid hormone-responsive B1 gene is associated with premature ovarian failure. Hum. Reprod., 23, 1457–1465. 34. Knauff, E.A., Franke, L., van Es, M.A., van den Berg, L.H., van der Schouw, Y.T., Laven, J.S., Lambalk, C.B., Hoek, A., Goverde, A.J., Christin-Maitre, S. et al. (2009) Genome-wide association study in premature ovarian failure patients suggests ADAMTS19 as a possible candidate gene. Hum. Reprod., 24, 2372– 2378. 35. Oldenburg, R.A., van Dooren, M.F., de Graaf, B., Simons, E., Govaerts, L., Swagemakers, S., Verkerk, J.M., Oostra, B.A. and Bertoli-Avella, A.M. (2008) A genome-wide linkage scan in a Dutch family identifies a premature ovarian failure susceptibility locus. Hum. Reprod., 23, 2835– 2841. 36. Sills, E.S., Harmon, K.E. and Tucker, M.J. (2002) First reported convergence of premature ovarian failure and cutis marmorata telangiectatica congenita. Fertil. Steril., 78, 1314– 1316. 37. Tupler, R., Barbierato, L., Larizza, D., Sampaolo, P., Piovella, F. and Maraschio, P. (1994) Balanced autosomal translocations and ovarian dysgenesis. Hum. Genet., 94, 171–176. 38. Aboura, A., Dupas, C., Tachdjian, G., Portnoi, M.F., Bourcigaux, N., Dewailly, D., Frydman, R., Fauser, B., Ronci-Chaix, N., Donadille, B. et al. (2009) Array comparative genomic hybridization profiling analysis

39. 40.

41. 42. 43.

44.

45.

46.

47.

48.

49. 50.

51. 52. 53.

reveals deoxyribonucleic acid copy number variations associated with premature ovarian failure. J. Clin. Endocrinol. Metab., 94, 4540–4546. Ledig, S., Ropke, A. and Wieacker, P. (2010) Copy number variants in premature ovarian failure and ovarian dysgenesis. Sex Dev., 4, 225–232. Kwok, C., Weller, P.A., Guioli, S., Foster, J.W., Mansour, S., Zuffardi, O., Punnett, H.H., Dominguez-Steglich, M.A., Brook, J.D., Young, I.D. et al. (1995) Mutations in SOX9, the gene responsible for Campomelic dysplasia and autosomal sex reversal. Am. J. Hum. Genet., 57, 1028–1036. Huang, B., Wang, S., Ning, Y., Lamb, A.N. and Bartley, J. (1999) Autosomal XX sex reversal caused by duplication of SOX9. Am. J. Med. Genet., 87, 349–353. Qin, Y. and Bishop, C.E. (2005) Sox9 is sufficient for functional testis development producing fertile male mice in the absence of Sry. Hum. Mol. Genet., 14, 1221–1229. Leipoldt, M., Erdel, M., Bien-Willner, G.A., Smyk, M., Theurl, M., Yatsenko, S.A., Lupski, J.R., Lane, A.H., Shanske, A.L., Stankiewicz, P. et al. (2007) Two novel translocation breakpoints upstream of SOX9 define borders of the proximal and distal breakpoint cluster region in campomelic dysplasia. Clin. Genet., 71, 67–75. He, C., Kraft, P., Chasman, D.I., Buring, J.E., Chen, C., Hankinson, S.E., Pare, G., Chanock, S., Ridker, P.M. and Hunter, D.J. (2010) A large-scale candidate gene association study of age at menarche and age at natural menopause. Hum. Genet., 5, 515– 527. He, C., Kraft, P., Chen, C., Buring, J.E., Pare, G., Hankinson, S.E., Chanock, S.J., Ridker, P.M., Hunter, D.J. and Chasman, D.I. (2009) Genome-wide association studies identify loci associated with age at menarche and age at natural menopause. Nat. Genet., 6, 724– 728. Murray, A., Bennett, C.E., Perry, J.R., Weedon, M.N., Jacobs, P.A., Morris, D.H., Orr, N., Schoemaker, M.J., Jones, M., Ashworth, A. et al. (2011) Common genetic variants are significant risk factors for early menopause: results from the Breakthrough Generations Study. Hum. Mol. Genet., 1, 186–192. Stolk, L., Zhai, G., van Meurs, J.B., Verbiest, M.M., Visser, J.A., Estrada, K., Rivadeneira, F., Williams, F.M., Cherkas, L., Deloukas, P. et al. (2009) Loci at chromosomes 13, 19 and 20 influence age at natural menopause. Nat. Genet., 6, 645– 647. Korn, J.M., Kuruvilla, F.G., McCarroll, S.A., Wysoker, A., Nemesh, J., Cawley, S., Hubbell, E., Veitch, J., Collins, P.J., Darvishi, K. et al. (2008) Integrated genotype calling and association analysis of SNPs, common copy number polymorphisms and rare CNVs. Nat. Genet., 40, 1253– 1260. Price, A.L., Patterson, N.J., Plenge, R.M., Weinblatt, M.E., Shadick, N.A. and Reich, D. (2006) Principal components analysis corrects for stratification in genome-wide association studies. Nat. Genet., 38, 904–909. Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A., Bender, D., Maller, J., Sklar, P., de Bakker, P.I., Daly, M.J. et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet., 81, 559–575. Barrett, J.C., Fry, B., Maller, J. and Daly, M.J. (2005) Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics, 21, 263–265. Shi, Y.Y. and He, L. (2005) SHEsis, a powerful software platform for analyses of linkage disequilibrium, haplotype construction, and genetic association at polymorphism loci. Cell. Res., 15, 97–98. Mantel, N. and Haenszel, W. (1959) Statistical aspects of the analysis of data from retrospective studies of disease. J. Natl Cancer. Inst., 22, 719–748.