UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genetic factors in premature ovarian failure Bretherick, Karla Lucia 2008

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
24-ubc_2008_spring_bretherick_karla_lucia.pdf [ 11.88MB ]
Metadata
JSON: 24-1.0066666.json
JSON-LD: 24-1.0066666-ld.json
RDF/XML (Pretty): 24-1.0066666-rdf.xml
RDF/JSON: 24-1.0066666-rdf.json
Turtle: 24-1.0066666-turtle.txt
N-Triples: 24-1.0066666-rdf-ntriples.txt
Original Record: 24-1.0066666-source.json
Full Text
24-1.0066666-fulltext.txt
Citation
24-1.0066666.ris

Full Text

GENETIC FACTORS IN PREMATURE OVARIAN FAILURE  by  KARLA LUCIA BRETHERICK B.Sc., The University of Calgary, 2000  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES  (Medical Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2008  © Karla Lucia Bretherick, 2008  Abstract  Approximately 1% of women will experience menopause before the age of 40, a condition known as premature ovarian failure (POF). The goal of this thesis was to identify genetic causes of POF by examining a number of candidate factors in POF patients and control women. Carriers of FMR1 premutations (55-200 CGG repeats) are known to be at increased risk of POF. A higher prevalence of alleles between 35-54 repeats was found among POF patients (p=0.01), suggesting that risk for POF may extend outside the classic premutation range. There was no evidence for any difference in FMRI promoter methylation or gene expression between cases and controls. Allele distributions of gene polymorphisms in the androgen receptor (AR), estrogen receptor (3, sex hormone binding globulin, and FSH receptor genes did not differ between POF patients and controls. However, haplotype at the estrogen receptor a gene, ESRI, was found to be associated with POF in a simple dominant manner (RR=9.7; 95% CI=2.635.6). Although the functional effect of this haplotype could not be confirmed, it may confer a more active promoter that influences risk by increasing the rate of follicular atresia. X-chromosome inactivation (XCI) skewing can indicate an abnormal X chromosome and may therefore be increased in POF patients. There was no increase in skewed XCI >90% in patients with secondary amenorrhea, however, there was a significant increase in 4 POF patients with primary amenorrhea (p=0.001). No Xchromosome abnormalities were detectable by high resolution DNA microarray, and skewed XCI may be explained by a trisomic rescue event causing reduced follicular pool. Age-related chromosome factors were assessed to determine if POF patients demonstrate an increased rate of cellular aging. With age, XCI skewing and AR methylation increase  ii  and telomere length decreases. There was no difference in skewing or methylation between patients and controls. Surprisingly telomere length was increased in POF patients (p=0.04), a finding that may be explained by abnormal estrogen exposure. Genotype at the longevity-associated APOE gene was not associated with POF. In conclusion, these findings have illuminated several new areas of research in this field and provide background for future research into POF pathogenesis.  iii  Table of Contents Abstract ^  ii  Table of Contents ^  iv  List of Tables ^  viii  List of Figures ^  ix  List of Abbreviations ^  xi  Acknowledgements ^  xv  Co-Authorship statement ^  xvi  Chapter 1: Introduction ^  1  1.1 Diagnosis ^  2  1.2 Epidemiology ^  3  1.3 Pathophysiology ^  4  1.3.1 Follicular development and atresia ^  4  1.3.2 Regulation of cyclic follicle recruitment ^  6  1.3.3 Follicular depletion and ovarian dysfunction ^  7  1.4 Etiology of premature ovarian failure ^  9  1.4.1 Non-genetic causes of POF ^  10  1.4.2 Genetic causes of POF ^  12  1.4.2.1 Cytogenetic abnormalities implicated in POF ^ 15 1.4.2.2 Single genes implicated in POF ^ 1.5 Aging and Reproductive aging ^  18 24  1.5.1 Evidence for a link between chronological age and reproductive age ^24 1.5.2 Genetic factors associated with aging ^  26  1.5.2.1 Telomere length^  26  1.5.2.2 Skewed X chromosome inactivation ^  28  1.5.2.3 DNA methylation ^  31  1.5.2.4 Gene variants ^  32  1.6 Research objectives ^  33  1.7 References ^  41  iv  Chapter 2: FMR1 and premature ovarian failure ^  56  2.1 Introduction ^  56  2.2 Methods ^  59  2.2.1 Samples ^  59  2.2.2 FMRI repeat length ^  60  2.2.3 X chromosome inactivation skewing ^  62  2.2.4 FMRI genotype repeat size ^  63  2.2.5 DNA methylation ^  65  2.2.6 FMRI mRNA quantification ^  66  2.2.7 Statistical analysis ^  69  2.3 Results ^  69  2.3.1 FMRI allele size ^  69  2.3.2 FMR1 biallelic mean ^  71  2.3.3 X chromosome inactivation and genotype repeat size ^ 71 2.3.4 Methylation at the FMRI promoter ^  72  2.3.5 FMRI transcript level ^  74  2.4 Discussion ^ 2.4.1 FMRI CGG repeat length and POF ^  75 75  2.4.2 No association between FMRI expression and POF ^ 77 2.4.3 Possible mechanisms for FMRI in POF ^ 2.5 References ^  78 102  Chapter 3: Survey of hormone-related gene variants in POF ^ 108 3.1 Introduction ^  108  3.2 Methods ^  113  3.2.1 Samples ^  113  3.2.2 Genotyping ^  115  3.2.3 X chromosome inactivation skewing ^  116  3.2.4 ESR1 luciferase assay ^  116  3.2.5 ESR1 mRNA quantification ^  119  3.2.6 Statistical analysis ^  120  v  3.3 Results ^  120  3.3.1 ESR2, AR, SHBG, and FSHR are not associated with POF ^ 120 3.3.2 ESR1 is associated with POF ^  121  3.3.3 Mode of inheritance for ESR1 associated POF risk ^ 122 3.3.4 Interaction between ESR1 and FMR1 ^  123  3.3.5 Regulatory effects of ESR1 (TA)„ repeat ^  124  3.4 Discussion ^  125  3.5 References ^  149  Chapter 4: X chromosome inactivation skewing in primary and secondary POF ^159 4.1 Introduction ^  159  4.2 Methods ^  160  4.2.1 Samples ^  160  4.2.3 XCI skewing assay ^  161  4.2.4 DNA microarray ^  162  4.3 Results ^  164  4.3.1 XCI skewing in patients with secondary amenorrhea ^ 164 4.3.2 XCI skewing in patients with primary amenorrhea ^ 164 4.3.3 DNA microarray ^  165  4.4 Discussion ^  165  4.5 References ^  174  Chapter 5: Age-related chromosome factors and POF ^  177  5.1 Introduction^  177  5.2 Methods ^  181  5.2.1 Samples ^  181  5.2.2 Telomere length analysis ^  182  5.2.3 X chromosome inactivation skewing ^  185  5.2.4 DNA methylation at Androgen Receptor ^  185  5.2.5 Genotyping ^  187  5.3 Results^  188  vi  5.3.1 Telomere length ^  188  5.3.2 X chromosome inactivation skewing ^  190  5.3.3 AR DNA methylation ^  190  5.3.4 Correlation between age-related chromosome factors ^ 191 5.3.5 APOE genotype ^  192  5.4 Discussion ^  193  5.5 References ^  218  Chapter 6: Discussion ^  224  6.1 Genetic factors assessed in this thesis ^  224  6.2 Strengths and limitations of this study ^  226  6.3 Current knowledge and future research on genetics and POF ^229 6.4 Significance ^  231  6.5 Conclusion ^  233  6.6 References ^  234  Appendix 1 UBC Research Ethics Board Certificates of Approval C01-0460 ^236 Appendix 2 UBC Research Ethics Board Certificates of Approval H01-70460 ^239  vii  List of Tables Table 1.1 Stages of follicle development ^  35  Table 1.2 X chromosome genes implicated in POF ^  36  Table 1.3 Autosomal genes implicated in POF ^  37  Table 1.4 Genetic diseases associated with POF ^  39  Table 2.1 Studies examining incidence of POF in Fragile X Syndrome families ^ 82 Table 2.2 Studies examining FMR1 repeat length in women with POF ^ 83 Table 2.3 FMR1 alleles in POF patients and control populations ^ 84 Table 2.4 FMR1 repeat size, biallelic mean and genotype repeat size ^ 85 Table 2.5 Correlations with FMR1 methylation levels ^  86  Table 2.6 Correlations with FMR1 expression levels ^  87  Table 3.1 Published phenotypic associations with FSHR, ESR2, AR and SHBG ^ 132 Table 3.2 Published phenotypic associations with ESR1 polymorphisms ^ 133 Table 3.3. Published frequencies of EGFR and D1 3S317 alleles ^ 134 Table 3.4. Primer sequences used in genotyping ^  135  Table 3.5. Hormone receptor allele frequencies in POF patients and controls ^ 136 Table 3.6. Observed and Hardy-Weinberg-expected genotypes for ESR1 (TA)„ repeat ^ 137 Table 4.1 Clinical details for POF patients presenting with primary amenorrhea ^ 169 Table 4.2 FISH results for microarray follow-up ^  170  Table 5.1 Age-adjusted telomere length in POF patients and controls ^ 198 Table 5.2 Average telomere length and ESR1 genotype^  199  Table 5.3 Correlation between relative telomere length and FMR1 repeat length ^200 Table 5.4 Age-adjusted AR methylation in POF patients and controls ^201 Table 5.5 Correlations between age-related chromosome factors ^ 202  viii  List of Figures  Figure 1.1 Hormone regulation of oocyte maturation ^  40  Figure 2.1 Methylation quantification by single nucleotide primer extension (SNuPE) ^ 88 Figure 2.2 FMRI gene promoter region ^  89  Figure 2.3 Example of FMRI Southern blot results ^  90  Figure 2.4 RNA degradation over time ^  91  Figure 2.5 Validity of qPCR assay for FMRI with reference gene GUS ^ 92 Figure 2.6 Correlation between XCI skewing assayed at the AR and FMRI loci ^ 93 Figure 2.7 Correlation between FMRI methylation assayed at sites C5 and C6 ^ 94 Figure 2.8 FMRI promoter methylation assayed at CpG sites C5 and C6 ^ 95 Figure 2.9 FMRI promoter methylation at site C5 by age ^  96  Figure 2.10 Promoter methylation at site C5 and FMRI repeat length ^ 97 Figure 2.11 Reproducibility of FMRI quantitative PCR assay ^ 98 Figure 2.12 Relative FMRI transcript level in POF patients and controls ^99 Figure 2.13 Relative transcript level by FMRI repeat length ^ 100 Figure 2.14 Relative FMR1 transcript level and methylation ^ 101 Figure 3.1 TaqMan allelic discrimination ^  138  Figure 3.2 ESR1 gene promoter region ^  139  Figure 3.3 ESR2 allele distribution in POF patients and control group 1 ^ 140 Figure 3.4 AR allele and genotype repeat size distribution ^  141  Figure 3.5 SHBG allele distribution in POF patients and combined controls ^ 142 Figure 3.6 Genotype distribution of FSHR SNPs ^  143  Figure 3.7 ESR1 (TA) n repeat allele distribution ^  144  Figure 3.8 Estimated ESRI haplotype frequencies ^  145  Figure 3.9 ESR1 genotype distribution and FMRI repeat size in POF patients ^ 146 Figure 3.10 Relative expression of 801bp and 1346bp reporter constructs ^ 147 Figure 3.11 Relative expression of 1512bp reporter with (TA)14 or (TA)22 repeats ^ 148 Figure 4.1 Example of FISH results evaluated in microarray follow-up ^ 171 Figure 4.2 Degree of skewed XCI in secondary POF patients and controls ^ 172 Figure 4.3 DNA microarray data for select POF patients with skewed XCI ^ 173  ix  Figure 5.1 Telomere measurement by quantitative PCR ^  203  Figure 5.2 Validation of telomere length assay ^  204  Figure 5.3 Quantitative DNA methylation assay for AR ^  205  Figure 5.4 Validation of AR methylation assay ^  206  Figure 5.5 APOE genotyping assay ^  207  Figure 5.6 Average relative telomere length by age ^  208  Figure 5.7 Telomere length in POF patients, Control group 1 and Control group 2 ^209 Figure 5.8 Telomere length and CGG repeat length of the longer FMR/allele ^ 210 Figure 5.9 X chromosome inactivation skewing and age ^  211  Figure 5.10 AR methylation and age ^  212  Figure 5.11 AR methylation in POF patients and control women ^213 Figure 5.12 Correlation between telomere length and XCI skewing ^214 Figure 5.13 Correlation between telomere length and AR methylation ^215 Figure 5.14 Correlation between XCI skewing and AR methylation^ 216 Figure 5.15 APOE allele and genotype frequencies in POF patients and controls ^ 217  x  List of Abbreviations AGTR2  Angiotensin II type 2 receptor gene  AIRE  Autoimmune regulator gene  AMH  anti-Mullerian hormone  AMHR2  AMH type II receptor gene  ANOVA  analysis of variance  ANCOVA  analysis of covariance  APOE  Apolipoprotein E gene  APS  Autoimmune polyendocrine syndrome  AR  Androgen Receptor gene  313I-ISD  3-beta-hydroxysteroid dehydrogenase-isomerase  BAC  bacteria artificial chromosome  BERKO  estrogen receptor 13 knock out  BMD  bone mineral density  BMP15  Bone morphogenetic protein 15 gene  BPES cDNA  Blepharophimosis-ptosis-epicanthus inversus syndrome complementary DNA  CEPH  Centre d'etude du polymorphisme humaine  CGH  comparative genomic hybridization  CHM  Choroideremia gene  CI  confidence interval  CpG  cytosine guanine nucleotide sequence (in 5'--6' orientation)  CREB  Clinical research ethics board  Ct  cycles required to reach threshold  DACH2  Drosophila Dachshund homologue 2 gene  DAX-1  Dosage-sensitive sex reversal-congenital adrenal hypoplasia critical region on the X chromosome 1 gene  DFFRX  Drosophila fat facets related X-linked gene  DIAPH2  Drosophila diaphanous 2 homologue gene  DIG  digoxygenin xi  DMSO^dimethyl sulfoxide DNA^deoxyribonucleic acid Dnmt^DNA methyl transferase dNTP^deoxyribonucleotide triphosphate DZ^dizygotic E2^estradiol-17(3  EE 2^ethinyl estradiol EDTA^ethylenediaminetetraacetic acid EGFR^Epidermal growth factor receptor gene EIF2B^Eukaryotic initiation factor 2B gene EM^early menopause ERE^estrogen response element ERKO^estrogen receptor a knock out ERa^estrogen receptor a ERI3^estrogen receptor f3 ESR1^Estrogen receptor a gene ESR2^Estrogen receptor /3 gene FBS^fetal bovine serum FHA^Functional hypothalamic amenorrhea FISH^fluorescent in situ hybridization FIGLA^Factor in the germline alpha gene FMR1^Fragile X mental retardation 1 gene FMR2^Fragile X mental retardation 2 gene FMRP^Fragile X mental retardation protein FOXE1^Forkhead box El gene FOXL2^Forkhead transcription factor L2 gene FOX01A Forkhead box 01 gene FOXO3A Forkhead box 03 gene FSH^follicle stimulating hormone FSHPRIll FSH primary response homologue 1 gene FSHR^FSH receptor gene  xi i  FXTAS^Fragile X tremor ataxia syndrome GALT^Galactose-1 phosphate uridylyltransferase gene GDF9^Growth differentiation factor 9 gene GnRH^gonadotropin releasing hormone GRS^genotype repeat size GUS^/3-glucoronidase gene HLA^human leukocyte antigen HPO^hypothalamic-pituitary-ovarian HRT^hormone replacement therapy INHa^Inhibin a gene INH/3A^Inhibin 6' subunit A gene )  INH/3B^Inhibin /3 subunit B gene KAL^Kallman syndrome 1 gene LB^Luria-Bertani LH^Luteinizing hormone LHCGR^Luteinizing hormone/choriogonadotropin receptor gene LH/3^polypeptide gene LINE^long interspersed element LOD^log of the odds MEM-a^minimum essential medium a MHC^major histocompatibility complex mRNA^messenger RNA MZ^monozygotic NOBOX Newborn ovary homeobox gene NOG^Noggin gene ONPG^ortho-nitrophenyl-b-D-galactopyranoside PCOS^Polycystic ovarian syndrome PCR^polymerase chain reaction POF^Premature ovarian failure POLG^DNA polymerase gamma gene qPCR^quantitative PCR  RNA^ribonucleic acid ROS^Resistant ovary syndrome RR^relative risk RT^reverse transcription SHBG^Sex hormone binding globulin gene SNP^single nucleotide polymorphism SNuPE^single nucleotide primer extension SOX3^SRY-box3 gene SWAN^Study of women across the nation SYM1^Proximal symphalangism 1 gene TERT^telomerase reverse transcriptase t/s^telomere/single copy gene ratio TSIX^XIST antisense gene UV^ultraviolet Xa^active X chromosome XCI^X chromosome inactivation Xi^inactive X chromosome XIC^X inactivation centre XIST^X inactivation specific transcript XPNPEP2 X-propyl aminopeptidase 2 gene ZFX^X-linked zinc finger gene  xiv  Acknowledgements I would like to thank my supervisor, Dr Wendy Robinson, for giving me both direction and freedom and investing so much time in me. Thank you to my committee members: Dr Carolyn Brown, Dr Geoff Hammond, and Dr Barb McGillivray, for their advice and support. Thank you to Ruby, Maria, Cathy, Jane, and Christy for training and assistance with experimental protocols. To Sara, Luana, Monica, Lauren, Courtney, Danielle and Jenn, thank you for keeping me entertained. I can't imagine how I would have done this without you, and I will miss you dearly when I leave. To my family: Mom, Dad, Suzannah, Norman, Liza, Ben, Pam and Mark, and my in laws: Ross and Elaine, thank you for your continual love and support. And to David, you saw the worst of it all and you stuck around, thank you for believing in me and keeping me sane.  XV  Co-Authorship statement Sections of Chapter 2 have been published by Bretherick KL, Fluker MR, and Robinson WP (2005). For this manuscript I performed all the data collection and analysis and wrote the manuscript. MRF ascertained patients. WPR designed the research project, supervised the research, and edited the manuscript. A version of Chapter 3 has been published by Bretherick KL, Hanna CW, Currie LM, Fluker MR, Hammond GL, Robinson WP (2008). This project was designed primarily by WPR and myself, with input from GLH. I performed —80% of the data collection and supervised two undergraduate students, CWH and LMC on the remainder. I analyzed the data in collaboration with WPR and personally wrote the manuscript. MRF ascertained patients. WPR supervised the research and edited the manuscript. Chapter 4 has been published by Bretherick KL, Metzger DL, Chanoine JP, Panagiotopoulos C, Watson SK, Lam WL, Fluker MR, Brown CJ, Robinson WP (2007). I collected and analysed 70% of the data and wrote the majority of the manuscript. SKW and WLL collected and analysed the remainder of the data and wrote the methods and results pertaining to this work. DLM, JPC, CP, and MRF ascertained patients. CJB and WPR developed the hypothesis and provided guidance and advice on data collection and analysis. WPR supervised the research and edited the manuscript. Portions of Chapter 5 are currently being prepared for publication by Hanna CW, Bretherick KL, Gair JL, Fluker MR, Stephenson M, Lansdorp P, Robinson WP (2008). CWH and I contributed equally to the collection of the data, analysis of results and preparation of the manuscript. MRF and MS ascertained patients. PL provided advice on data collection. WPR developed the hypothesis with JLG and supervised the research.  xvi  Chapter 1: Introduction The average age at menopause for women in the western world is 51 years, however approximately 1% of women will experience menopause before the age of 40, a condition known as premature ovarian failure (POF). POF has considerable physiological and psycho-social consequences. Women experiencing POF will live a significant portion of their lives with a post-menopausal hormone profile, putting them at greater risk for health consequences normally associated with aging. They also must contend with the devastation of infertility and its effects on psychological health and family planning. Because of a trend for women in many parts of the developed world to delay childbirth until later in life, an increasing number of women affected with POF will experience infertility before they have had the chance to make decisions regarding childbearing. As an introduction to this thesis, this chapter will provide basic background information required to understand the condition of POF and the genetic factors discussed in the following chapters. First, the diagnosis and prevalence of POF will be reviewed. Subsequently, the pathophysiology of ovarian failure will be covered, with a review of normal follicle development and regulation of follicular recruitment. This will be followed by a discussion of the known and postulated causes of POF, including nongenetic causes briefly and genetic causes in detail. Finally, the relationship between aging and reproductive aging will be presented, with a review of genetic factors known to be associated with aging. The introduction will conclude with the research objectives targeted by this thesis.  1  1.1 Diagnosis  Premature ovarian failure is characterized by hypergonadotropic amenorrhea prior to age 40. Diagnosis is made following a period of amenorrhea lasting at least 6 months by classic criteria (Vegetti et al. 2000), although it has been argued that to facilitate treatment in a timely manner a cut-off of 4 months may be more appropriate (Nippita and Baber. 2007). To confirm anovulation and hypergonadotropism, two serum follicle stimulating hormone (FSH) values of >40mIU/m1 must be obtained more than one month apart (Nippita and Baber. 2007). FSH stimulates follicle development in the ovary and is normally negatively regulated by inhibin and estrogen released from the maturing follicle. In ovarian failure, however, the absence of a maturing follicle results in a lack of negative feedback and elevated FSH levels. In addition, POF patients will also have low estrogen levels and elevated luteinizing hormone (LH) levels characteristic of postmenopausal women. POF and early menopause (EM; defined as menopause before age 45) can be considered part of a continuum of the spectrum of age of menopause onset (Vegetti et al. 2000), however this inclusion remains contentious (Kalantaridou and Nelson. 2000; Nelson et al. 2005). Both normal menopause and POF are characterized by amenorrhea and elevated serum FSH. However natural menopause is an irreversible cessation of ovulation, whereas nearly half of all POF patients have intermittent ovarian function that may persist for several years following diagnosis (Conway et al. 1996, Nelson et al. 1994). Furthermore, sporadic spontaneous ovulation has been reported in up to 20% of POF patients (Nelson et al. 1994) and pregnancy has been reported in approximately 510% of those diagnosed (Rebar and Connolly. 1990). For this reason many clinicians  2  favor the more symptom-specific labels "hypergonadotropic amenorrhea" or "hypergonadotropic hypogonadism" or the patient friendly term "ovarian insufficiency" which better describe the intermittent nature of POF (Nelson et al. 2005). Most patients with POF present with secondary amenorrhea, in which menstruation spontaneously ceases after a period of normal menses, and many studies of POF in the literature confine the term to this group (Simpson and Rajkovic. 1999; Vegetti et al. 2000). Others, however, include patients with primary amenorrhea, who do not undergo menarche, within the POF diagnosis (Anasti. 1998; Goswami and Conway. 2005; Kalantaridou and Nelson. 2000; Nelson et al. 2005). The presence of primary or secondary amenorrhea does not necessitate a diagnosis of POF; only 10-28% of primary amenorrhea patients and 4-18% of secondary amenorrhea patients are found to have hypergonadotropism characteristic of POF (Anasti. 1998). Other common causes of secondary amenorrhea include: polycystic ovarian syndrome (PCOS), characterized by androgen excess; functional hypothalamic amenorrhea (FHA), distinguished by low FSH; and hyperprolactinemia characterized by high prolactin levels (Practice Committee of the American Society for Reproductive Medicine. 2006). Except where indicated, the results reported in this thesis are based on analysis of POF patients with secondary amenorrhea. POF as a result of primary amenorrhea likely has a distinct etiology and the results presented here should not be extrapolated to these POF patients. 1.2 Epidemiology  The overall prevalence of premature ovarian failure has been reported as 0.9-1.2% (Coulam et al. 1986; Cramer and Xu. 1996; Luborsky et al. 2003). Annual incidence increases with older age cohorts, such that for women ages 15-29 the incidence is only 10 3  per 100 000, but for women ages 30-39 the incidence is 76 per 100 000 (Coulam et al. 1986). In addition, there are ethnic differences that may be responsible for the differences in POF prevalence reported between study groups, an important consideration in genetic association studies. Prevalence of POF was found to be 1.0% in Caucasian women, slightly higher in African American women (1.4%) and Hispanic women (1.4%), and lower in Chinese (0.5%) and Japanese (0.1%) women, based on the longitudinal Study of Women's Health Across the Nation (SWAN), a United States survey collecting information based on self-report, regarding ethnicity and menopause along with other health and lifestyle factors (Luborsky et al. 2003).  1.3 Pathophysiology Premature ovarian failure may be caused by either a reduction in oocyte reserve, or an abnormality in ovarian function. Reduced oocyte reserve is a situation analogous to natural menopause; when a critical threshold of follicles is reached, there is a failure of response to signals directing follicle maturation, and an absence of ovulation. Women with POF may be reaching this threshold at an earlier age as either a result of accelerated follicular atresia, or a reduced complement of follicles at development. In contrast, dysfunctional ovarian response to signals of maturation may cause POF in women who possess a normal, age-appropriate oocyte reserve. An understanding of these two possible mechanisms for POF requires knowledge of both follicular development as well as the complex hormonal signals regulating follicle recruitment.  1.3.1 Follicular development and atresia In the developing embryo 1000-2000 primordial germ cells originating in the yolk-sac endoderm migrate to the genital ridge during weeks 5 -7 of human gestation and 4  proliferate to nearly 600 000 germ cells, known as oogonia (reviewed in Anasti. 1998). A subset of the oogonia is lost by atresia, while the remainder continues to proliferate by mitosis or differentiate by entering meiosis and becoming non-dividing primary oocytes. Germ cell number peaks at approximately 6-7 million by gestational week 20, after which primordial follicle formation begins as oocytes are surrounded by a layer of pregranulosa cells. The primordial follicles then remain inert, arrested after the pachytene stage of meiosis I, until lost by atresia or recruited for maturation. The original pool of primordial follicles is reduced dramatically during later stages of gestation, so that by birth there are only 250 000 - 750 000 oocytes remaining (Forabosco et al. 1991). Primordial follicles remain in the ovary, dormant until conscripted by either initial or cyclic recruitment (reviewed in McGee and Hsueh. 2000). Initial recruitment causes primordial follicles to develop into primary follicles, and progress into secondary, and antral (also called tertiary) follicles (Table 1.1) after which they are lost through atresia if not enlisted by cyclic recruitment. Initial recruitment begins before 24 weeks of gestation and continues throughout development, childhood, puberty, and adult life. Puberty initiates the onset of cyclic recruitment, a process in which gonadotropins signal a few antral follicles to continue growth, one of which becomes the dominant Graafian follicle from which an oocyte will be ovulated. In the 35-40 reproductive years of a female's life only 400-500 of the original 6-7 million germ cells will become fully mature and undergo ovulation. For every follicle ovulated, nearly 1000 follicles undergo atresia. When the follicular pool has been reduced by ovulation and atresia to a critical threshold of only —1000 primordial follicles, cyclic recruitment will cease and menopause will commence (Faddy et al. 1992; Nikolaou and Templeton. 2004).  5  A notable feature in the process of ovarian follicle depletion is the parallel decline in the number and quality of the remaining oocytes (Nikolaou and Templeton. 2004). The mechanism for the decline in quality includes meiotic non-disjunction and accumulation of DNA damage in the oocyte over time. However it has also been postulated that a contributing factor may be that the best oocytes are selected for recruitment early in life and the decline in oocyte quality may merely reflect the relative decrease in the number of quality oocytes in the remaining follicular pool (te Velde and Pearson. 2002). Mathematical modeling of the rate of oocyte decline based on pathological data-sets has suggested that rate of atresia increases with age, so that as follicle number falls below 25 000 there is nearly a doubling in the rate of follicle loss (Faddy et al. 1992; Faddy. 2000) reflected as a rapid decline in fertility. This observation has led to the "fixed interval hypothesis" (Nikolaou and Templeton. 2004), that between this critical figure of 25 000 and menopause there is a roughly fixed time period of 13 years. Therefore, for women experiencing menopause prior to age 40, sub-fertility may begin before the age of 30 (Faddy et al. 1992; te Velde and Pearson. 2002).  1.3.2 Regulation of cyclic follicle recruitment From the cohort of follicles that are continually selected for initial recruitment a subset will be rescued from the atretic process by cyclic recruitment (reviewed in McGee and Hsueh. 2000), a process controlled by hormonal feedback regulation (Figure 1.1). In response to increased pulses of gonadotropin releasing hormone (GnRH) from the hypothalamus, FSH is released from the anterior pituitary and stimulates 5-10 antral follicles to continue to develop. One of these antral follicles will become the dominant Graafian follicle by growing faster and producing higher levels of estradiol and inhibin B  6  than the rest of the cohort. This Graafian follicle will produce increasing levels of estradiol and inhibin B which provide negative feedback on the hypothalamus and pituitary, suppressing FSH secretion. It is likely that the Graafian follicle exerts its dominance by being more sensitive or responsive to FSH, and this feature will protect it as FSH levels drop, while the remaining antral follicles, unable to respond to decreasing FSH, fail to survive. In addition, the Graafian follicle produces local factors that enhance its own selection and negatively regulate the rest of the cohort. The Graafian follicle will stimulate expression of FSH and LH receptors in its granulosa cells and continue to release increasing levels of estrogen which trigger the release of LH from the anterior pituitary. The LH surge triggers ovulation by binding to receptors on the granulosa cells and weakening the follicle wall. Following ovulation, the LH and FSH cause the residual follicle to become the corpus luteum, which releases estradiol, inhibin A, and progesterone. These hormones cause thickening of the endometrial wall in preparation for implantation. After 2 weeks with the demise of the corpus luteum, estrogen and inhibin levels fall removing the negative feedback to the hypothalamus and pituitary, and allowing FSH levels to once again rise and stimulate recruitment of a new cohort of antral follicles.  1.3.3 Follicular depletion and ovarian dysfunction There is evidence to support both follicular pool depletion and ovarian dysfunction as mechanisms in POF. Studies in both mice (Baker et al. 1980) and humans (Cramer et al. 1995a) show that unilateral oophorectomy decreases reproductive lifespan, although the severity of the shortening may depend on when in reproductive life the procedure is performed. In addition, one study found POF patients to have significantly  7  lower counts of primordial, primary and secondary follicles on ovarian biopsy than age matched control women (Vital-Reyes et al. 2006) although by ultrasound the probability of detecting a follicle remains stable for many years after diagnosis (Nelson et al. 1994). The mechanism by which reduced follicular pool leads to abnormal follicle development likely involves inappropriate luteinization of the cohort of antral follicles selected for cyclic recruitment. If this cohort is too small or of poor quality, it may not produce sufficient negative feedback to down regulate gonadotropins (Nelson et al. 1994). This would precipitate elevated LH levels leading to premature luteinization of the dominant follicle impairing normal follicle function and inhibiting ovulation. Therefore it is a depletion of follicle number that precipitates ovarian dysfunction. The long-held dogma that mammalian females acquire a finite number of non-renewable germ cells in fetal development has been challenged by a study reporting follicular renewal in postnatal mouse ovary (Johnson et al. 2004). Johnson et al report the presence of mitotically active germ cells in the mouse ovary that are indispensable in restoring the follicular pool. Therefore it may also be a loss or a failure of these cells to continue to divide and repopulate the ovary that is the basis of POF, rather than merely a reduction of the follicular pool itself. POF patients with ovarian dysfunction have a normal follicle number but lack the ability to respond appropriately to signals directing oocyte maturation, suggesting that in these cases POF may have an endocrine basis. This condition is infrequently referred to as resistant ovary syndrome (ROS). Although some groups exclude this from the POF diagnosis on the basis that it is potentially treatable, it can only be distinguished from follicular depletion by ovarian biopsy, which is often not clinically warranted or  8  prognostic and is therefore not common practice (Laml et al. 2002). There are several causes for ovarian failure as a result of endocrine dysfunction which will be discussed in detail in the following sections. These include: deficiencies in enzymes required for steroid hormone synthesis as a result of either genetic abnormalities or autoimmune attack; mutations in steroid hormone or hormone receptor genes; and targeted damage of endocrine organs including the hypothalamus, pituitary, or ovaries by disease processes, autoimmune causes, or iatrogenic assault. Support for the presence of ovarian dysfunction as a mechanism in POF is seen in a number of women who have FSH receptor mutations and have immature follicles on ultrasound (Aittomaki et al. 1996; Beau et al. 1998). In summary, it is likely that some POF patients suffer ovarian dysfunction while maintaining age-appropriate numbers of follicles, whereas others have premature follicle depletion as a result of either diminished initial follicle pool or an increased rate of follicular atresia. The percentage of patients falling into either group likely differs between studies based on mode of ascertainment. It is probable that these distinct POF pathologies have different etiologies, an important aspect to keep in mind when doing genetic association studies on a phenotype that may more appropriately be divided into separate endophenotypes based on ovarian reserve.  1.4 Etiology of premature ovarian failure There are a number of known causes of POF, however, for more than 90% of patients no known cause can be identified (Nippita and Baber. 2007). These cases are considered idiopathic and typically described as "karyotypically normal spontaneous POF". This section will briefly review iatrogenic, environmental, viral and autoimmune  9  causes of POF, and then focus on genetic causes of POF, including the evidence for a genetic role in age at menopause, a discussion of chromosomal abnormalities associated with POF, and a review of X chromosome and autosomal genes implicated in POF.  1.4.1 Non genetic causes of POF -  Iatrogenic origins of POF are those that are a consequence or side effect of therapy for an unrelated medical condition. Chemotherapy and radiation treatment for malignant diseases can cause ovarian failure, although risk varies depending on the age of the patient at the time of treatment, the dosage required, and the region of the body targeted for treatment (Anasti. 1998). In general, chemotherapeutic agents exert their effects by targeted destruction of rapidly dividing cells or alteration of cellular DNA. Destruction of the rapidly dividing granulosa and theca cells of recruited follicles or DNA damage in non-proliferating primordial follicles may cause an increased rate of atresia resulting in a prematurely reduced follicular pool (Anasti. 1998). Similarly, radiation treatment may cause follicular apoptosis resulting in a reduced follicular pool, although the risk is reduced in prepubertal girls and in cases where the irradiation field does not directly target the pelvis (Goswami and Conway. 2005). As mentioned previously, unilateral oophorectomy results in increased risk of early menopause (Cramer et al. 1995a). Although this may be a result of directly reduced follicular pool, all pelvic surgeries carry a slightly increased risk of ovarian failure, suggesting the mechanism may be a result of indirect ovarian damage caused by reduced blood supply or inflammation of the ovary (Goswami and Conway. 2005). Evidence for the influence of lifestyle and environment on fertility and reproductive aging remains inadequate, however this has become a rapidly advancing  10  area of research. Among the most studied of the environmental toxins is cigarette smoke. Although cigarette smoking has not been linked specifically to POF, epidemiologic and hormone studies of smokers suggest they have advanced ovarian age and a 1-4 year earlier onset of menopause than non-smokers (Kinney et al. 2007; Sharara et al. 1998). Alcohol and caffeine, in contrast, were found not to be related to any indicators of ovarian aging (Kinney et al. 2007). Furthermore, there is no evidence to support either a positive or negative influence for oral contraceptive pills on POF risk (Luborsky et al. 2003). Studies of other environmental toxins including endocrine disrupters, heavy metals, solvents, pesticides, plastics and industrial chemicals have so far been controversial and inconclusive (reviewed in Sharara et al. 1998). Although detrimental consequences of exposure in animal and cell-culture systems have been documented (Uzumcu and Zachow. 2007), a lack of conclusive data on disease trends, defined exposure limits, and putative mechanisms, limits the clinical utility of this data (Foster and Holloway. 2003). It remains to be determined how and if these ubiquitous agents have any affect on ovarian aging. POF may also result from a number of disease processes and infections, particularly those with a direct gonadal insult as part of the disease process. Mumps infection in women after adolescence tends to target the ovaries causing oophoritis and up to 3-7% of women contracting mumps develop POF (Anasti. 1998). In addition, there are several anecdotal reports of ovarian failure following a variety of different viral infections (Rebar and Connolly. 1990; Goswami and Conway. 2005). POF may also occur secondary to iron overload from hemochromatosis or blood transfusion treatment for thalassemia major (Davis. 1996). POF also occurs as part of the spectrum of  11  symptoms in a variety of rare genetic disorders which will be described further in the following section. It has been estimated that autoimmune etiology plays a role in 20%-30% of POF diagnoses (reviewed in Goswami and Conway. 2005). POF has been reported in association with both endocrine (diabetes, thyroid and adrenal disease) and nonendocrine autoimmune conditions (myasthenia gravis, Sjogren's syndrome, Crohn's disease, vitiligo, pernicious anemia, rheumatoid arthritis, and systemic lupus erythematosis). Endocrine autoimmune ovarian failure may occur on the backdrop of Addison's disease, a result of an autoimmune attack on the adrenal glands, or as one of a constellation of symptoms present in one of the autoimmune polyglandular syndromes (APS). The mechanism behind the association may be the similarity of autoantigens between the ovary and the adrenal, resulting in a concurrent destruction of both cell types. Most commonly POF with autoimmune etiology will occur in the presence of anti-ovarian antibodies. However, because of different methods of assessment and a lack of antibody specificity, studies examining anti-ovarian antibodies have reported their prevalence in POF patients to be anywhere between 7 and 69% (Goswami and Conway. 2005). Therefore, although the presence of anti-ovarian antibodies may provide a hint as to the etiology of the ovarian failure, their presence should not be taken as proof of autoimmune attack on the ovaries as the sole mechanism of ovarian destruction.  1.4.2 Genetic causes of POF An indication that POF may have a genetic component is suggested by a number of families in which POF occurs in multiple affected individuals from more than one generation (Davis et al. 2000). Classically, POF has been divided into familial or  12  spontaneous classes, based on the presence or absence of a family history of POF. However, the proportion of POF cases falling into either of these categories varies widely among studies. The percentage of familial POF cases has been reported to be as low as 4% (Conway et al. 1996) and as high as 31% (Vegetti et al. 1998) in studies examining incidence in family members. Discrepancies between reports may be due to differences in criteria for diagnosis of POF, ascertainment bias in recruitment, recall bias in collection of family history data, and the inclusion of all POF cases or confinement to solely idiopathic POF cases. Thorough evaluation of family members of karyotypically normal POF patients diagnosed with classic criteria of idiopathic POF before age 40, suggests that the incidence of familial cases is 12.7% for this group (van Kasteren et al. 1999). Interestingly, patients with sporadic POF were found to have a significantly earlier age of onset than patients with familial POF (Vegetti et al. 1998). In some cases although unaffected family members were not found to have POF before age 40, they did have an earlier than normal menopause (van Kasteren et al. 1999). Familial POF may be more likely to be part of the spectrum of menopausal age and when considered as a quantitative trait, genetic factors may be found to play a larger role than previously thought. Family and twin studies provide evidence for a genetic role in determining age at menopause. Based on interview data, women experiencing menopause before age 45 were 4-9 times more likely than controls to have a mother, sister, aunt or grandmother who had also experienced early menopause, with the odds increasing for those with immediate family members affected, multiple affected family members, or a family member with menopause before age 40 (Cramer et al. 1995b). Heritability (h 2 ), the  13  proportion of the variance of a phenotypic trait attributable to genetic factors, has been estimated at 63% in a study of age at menopause in monozygotic (MZ) and dizygotic (DZ) twin pairs, with a best fit found in a model in which additive genetic and unique environmental variance components are included (Snieder et al. 1998). Similarly, estimates from analysis of age at menopause in a study of singleton and twin sisters suggest a heritability of 71-87% (de Bruin et al. 2001). A recent twin study of age at menopause reports a higher concordance in age at menopause in MZ twins vs. DZ twins (Gosden et al. 2007). However they also find that the prevalence of POF in both MZ and DZ twins was 3-5 times higher than in singletons, a finding that could be confounding inheritance estimates based on twin studies. Based on the distribution, variation, and high heritability of age at menopause, menopausal age may be a complex quantitative trait to which a number of environmental and genetic susceptibility factors contribute (van Asselt et al. 2004). Genome-wide linkage analysis of sibling pairs considering age at natural menopause as a quantitative trait uncovered two chromosomal regions with suggestive linkage at 9q21.3 (LOD = 2.6) and Xp21.3 (LOD = 3.1), and a further 12 chromosome regions with elevated LOD scores > 1.0 (van Asselt et al. 2004), supporting the suggestion that multiple genes contribute to menopausal age. Further support for a genetic component to POF comes from a number of POF cases for which a single gene mutation or chromosome alteration has been implicated as causative. These causes will be discussed in detail in the remainder of this section. It is possible that for a subset of patients, POF is genetically heterogeneous, with a number of single gene disorders each responsible for a handful of cases. Families where POF segregation through the pedigree suggests an obvious or probable mode of inheritance,  14  likely fall into this category. For other patients, POF can be thought of as the extreme end of the spectrum of menopausal age, a complex trait for which multiple genes contribute. Families which have a preponderance of women with POF, menopause prior to age 45, and simply earlier than average menopause likely follow this mode of inheritance. The percentage of POF cases that can be attributed to each of these genetic classes and the percentage that are truly the result of non-genetic causes remains unknown.  1.4.2.1 Cytogenetic abnormalities implicated in POF Ovarian failure for a portion of familial and spontaneous POF cases is attributable to inherited or de novo cytogenetic X chromosome abnormalities, including both structural and copy number aberrations. Determination of the frequency at which cytogenetic abnormalities are responsible for POF is complicated by differences in karyotypic resolution, mosaicism detection (Devi et al. 1998), and ascertainment bias between studies, but may be estimated at 5-13% (Devi and Benn. 1999; Vegetti et al. 2000). The X chromosome abnormality most commonly associated with POF is Turner syndrome, in which women have a single X chromosome. Turner syndrome patients commonly have ovarian dysgenesis and streak ovaries, and will in most cases experience severe POF presenting as primary amenorrhea. Mosaicism for a 45,X karyotype, in which some but not all cells in the body carry the abnormal karyotype, is also associated with increased risk for POF (Devi et al. 1998). Age of onset and severity of POF will likely depend on the percentage of abnormal cells and their distribution throughout organ systems. POF in women with trisomy X has been described in case reports (Holland. 2001; Itu et al. 1990), and population studies suggest up to 4% of patients may have  15  47,XXX karyotype (Goswami et al. 2003; Menon et al. 1984), however the risk of women with trisomy X developing POF remains unknown. A single X chromosome appears to be sufficient for initiation of ovarian differentiation, as oogenesis initially proceeds normally in a 45,X fetus. However two X chromosomes are required for ovarian maintenance, and when one is entirely or partially absent, massive fetal follicular atresia can result (Simpson and Rajkovic. 1999). Two possible mechanisms for failure of ovarian maintenance have been postulated, a disruption of meiotic pairing in developing germ cells (Ogata and Matsuo. 1995) or haplo-insufficiency for X chromosome or autosomal genes disrupted by the translocation breakpoint (Simpson and Rajkovic. 1999; Toniolo and Rizzolio. 2007). Although one of the two X chromosomes in every normal adult female is inactived in each cell, haplo-insufficiency could still occur for the subset of genes that escape X inactivation, or haplo-insufficiency in the oocyte where both X chromosomes are normally active could be critical. Which of these mechanisms is primarily responsible is still debated and evidence from POF patients with X chromosome structural abnormalities provides support for both. A number of X chromosome structural abnormalities have been reported to be associated with POF and analysis of these has provided insight into the chromosomal regions critical for normal ovarian function and mechanisms responsible for ovarian failure. Analysis of a number of cases of deletions and translocations of the X chromosome has led to the following suggestions: the proximal regions of the X chromosome short (Xp) and long (Xq) arms are most critical for ovarian maintenance, disruptions of these regions will usually cause primary amenorrhea, interruption of the distal regions of either arm are less severe, resulting in POF with a later age of onset  16  (Schlessinger et al. 2002; Simpson and Rajkovic. 1999). In general, the size of the deletion and extent of involvement of the proximal region will determine the severity of POF presentation. A number of cases of X chromosome deletion or X;autosome translocation in women with POF and normal fertility have been used to refine two "candidate regions" of the X chromosome that are necessary for normal ovarian function. The existence of these candidate regions suggests that chromosomal location of the X chromosome breakpoints is important, therefore dosage of genes in this region must affect reproductive capacity (Toniolo and Rizzolio. 2007). These candidate regions were termed POF1 at Xq26-q27, and POF2 at Xq13-q21, with the expectation that genes implicated in POF would be discovered in these regions (Davison et al. 1999; Marozzi et al. 2000). In addition to meiotic or gene disruption, a third possible mechanism, a position effect, in which deletion or chromosomal rearrangement causes changes in chromatin conformation altering gene regulation of X chromosome or autosomal genes near the breakpoint has also been proposed (Toniolo. 2006). The search for candidate genes disrupted by structural abnormalities revealed that of more than 40 cases with balanced translocations, only five genes were disrupted. In contrast, most breakpoints were found to fall within gene deserts (Rizzolio et al. 2006), supporting either the disrupted pairing hypothesis or the position effect hypothesis. For the region extending from Xq23-28 (termed critical region II) non-overlapping deletions suggest the presence of two or more X-linked genes required in double dose for normal ovarian function supporting the hypothesis of a position effect of the breakpoints on X-linked genes (Rizzolio et al. 2006). However, at Xq21 (critical region I) which is interrupted in over 80% of translocations associated with POF, a position effect for X chromosome genes is  17  unlikely, given that none of the genes within a 2Mb region were highly expressed in the ovary or had ovarian, follicle, or oocyte specific gene expression (Rizzolio et al. 2006). Alternatively, POF in patients with translocations at this breakpoint may result from a position effect on autosomal genes that have been relocated to the X chromosome (Toniolo and Rizzolio. 2007).  1.4.2.2 Single genes implicated in POF A variety of single genes have been implicated in POF, most often by the discovery of mutations found to be segregating with POF in isolated pedigrees. A number of these genes, including FMR1, the gene with the strongest association to POF, are located on the X chromosome. As expected, genes involved in hormonal signaling required for follicular development or function have also been associated with POF. In addition to these genes that exclusively cause POF, there are a variety of genetic diseases in which ovarian failure is a common clinical finding. Because of the association between ovarian failure and X chromosome abnormalities, X chromosome genes (Table 1.2) are natural candidates for involvement in POF. The most significant single gene association is with the Fragile X mental retardation 1 gene, FMRI, at Xq27.3. The 5'untranslated region of the FMR1 gene has a polymorphic CGG repeat and carriers of premutation size repeats (55-200 repeats) are at an increased risk of POF. Approximately 16-20% of premutation carriers experience POF (Allingham-Hawkins et al. 1999; Schwartz et al. 1994) and overall there is a shift towards an earlier menopause in premutation carriers (Allen et al. 2007; Hundscheid et al. 2001; Murray et al. 2000; Partington et al. 1996). Premutation size alleles are found in up to 7% of individuals with sporadic POF and 21% of familial POF cases but have a  18  prevalence of only 0.4% in control women (reviewed in Sherman. 2000) and therefore account for a small but significant proportion of POF cases. Given the association with FMR1, a related gene, FMR2, also containing a trinucleotide repeat, was a natural  candidate to examine for association to POF. Although one group reported finding increased frequency of FMR2 microdeletions in a population of POF patients (Murray et al. 1998; Murray et al. 1999) there has been no further confirmation of these findings in the literature. A number of X chromosome genes including POFIB, the Drosophila melanogaster diaphanous gene, DIAPH2, the X-propyl aminopeptidase 2 gene, XPNPEP2, the homologue of Drosophila Dachsund, DACH2, and the choroideremia  gene, CHM, have been suggested as candidates based on case reports of gene disruption by X;autosome translocations in POF patients. Although at the outset these genes are interesting candidates, a paucity of studies implicating them further in ovarian failure suggests that they are unlikely to be common causes of POF. Additional X chromosome candidate genes have been suggested due to their putative roles in the ovary. One of these, the Bone morphogenetic protein 15 gene, BMP15, was postulated to be a candidate gene based on a naturally occuring mutation in the Inverdale sheep, for which homozygotes experience primary ovarian failure (Galloway et al. 2000). BMP15 is expressed in the oocyte during folliculogenesis and acts to regulate the proliferation and differentiation of granulosa cells (Otsuka et al. 2000). Subsequent case reports (Di Pasquale et al. 2004) and population studies (Di Pasquale et al. 2006; Dixit et al. 2006; Laissue et al. 2006) have reported additional BMP15 mutations in POF patients and association of POF with a number of BMP15 sequence variants. Other X chromosome genes including Drosophila fat facets related X-linked gene, DFFRX, X linked zinc finger  19  protein, ZFX, the X-inactivation specific transcript, XIST, FSH primary response homologue 1, FSHPRH1, Angiotensin II type 2 receptor, AGTR2, and SRY related HMGbox 3, SOX3, have been suggested as candidates based on their putative function but have not been implicated in any specific cases of POF. A couple of these, DFFRX and ZFX are known to escape X chromosome inactivation (XCI) and therefore could conceivably cause POF due to haploinsufficiency if one copy is mutated or deleted. However, until genetic alteration at these genes is found in POF patients, their pathogenic role in ovarian failure cannot be confirmed. Normal ovarian activity requires proper endocrine regulation; therefore autosomal genes known or thought to play a role in hormone signaling (Table 1.3) are natural contenders for involvement in POF, particularly in cases where ovarian failure results from ovarian dysfunction rather than follicular depletion. FSH is necessary to recruit follicles for maturation, and inhibins A and B are responsible for negative feedback regulation of this process. POF patients have been screened for mutations in the FSH receptor gene, FSHR, the FSH/3 subunit gene, FSH/3, the Inhibin a gene, INHa, and Inhibin fi subunit genes, INH/JA, and INH,GB, and although mutations with suspected functional significance have been found, population screens have not revealed these to be common causes of POF. Luteinizing hormone (LH) is required to stimulate ovulation and therefore indirectly provides negative feedback on FSH production. A failure of response could therefore cause infertility and elevated FSH. Mutations and polymorphisms in the LH receptor gene, LHCGR, and LH/3 subunit gene, HO, have been reported in association with POF. Studies in mice have revealed that anti-Mullerian hormone (AMH) is expressed in the ovary from the onset of primordial follicle  20  recruitment and may negatively regulate the rate of recruitment to limit premature follicular exhaustion (Durlinger et al. 1999). Although no functional mutations were found in the AMH gene, AMH, or the AMH type II receptor gene, AMHR2, in a study of 16 Japanese POF patients (Wang et al. 2002), an association between age at menopause and a common variant in AMRH2 has recently been reported (Kevenaar et al. 2007). Deficiencies in enzymes required for estrogen synthesis such as cholesterol desmolase, 17a-hydroxylase, 17-20 desmolase, and aromatase, can result in primary amenorrhea and elevated gonadotropins (Anasti. 1998; Kalantaridou and Chrousos. 2000). These deficiencies may result from gene mutation although only a few case reports in patients with hypergonadotropic amenorrhea have been described. Genes encoding ovary-specific transcription factor genes such as Newborn ovary homeobox (NOBOX), Forkhead box 0 transcription factors (FOXO3A and FOX01), and Factor in the germline alpha (FIGLA),  have been implicated in POF due to their roles in regulating follicle development. Additional genes have been implicated due to their similarities to other genes with roles in the developing ovary. The Forkhead box El transcription factor gene, FOXE], has a similar polyalanine tract to that in FOXL2, a gene that has been implicated in blepharophimosis-ptosis-epicanthus inversus syndrome type I (BPESI) in which POF is a common feature. The Growth differentiation factor 9 gene, GDF9, is a bone morphogenetic factor related to BMP15, and is thought to have a similar expression pattern and role in regulating oocyte development (Elvin et al. 1999). GDF9 deficient mice are infertile (Dong et al. 1996) however mutation screens of POF patients have revealed few pathological mutations (see Table 1.3).  21  A variety of genetic disorders have been described in which POF is a commonly occurring feature, and the genes implicated in these cases (Table 1.4) may therefore have a role in ovarian failure. In some of these conditions the onset of POF is related to gonadal assault or endocrine dysfunction resulting from the disease process, in others the mechanism of ovarian failure remains unknown. Genetic disorders associated with POF include: galactosemia, a rare disorder resulting from deficiency in galactose-l-phosphate uridyl-transferase due to recessive mutations in the GALT gene; BPESI, an autosomal dominant condition caused by mutations in a winged helix/forkhead transcription factor, FOXL2; progressive external opthalmoplegia, a mitochondrial disease caused by  mutations in the gene for mitochondrial DNA polymerase y, POLG ; proximal symphalagism (SYM1) a condition caused by mutations in the Noggin gene, NOG; central nervous system hypomyelination/vanishing white-matter leukodystrophy a condition that can be caused by mutations in at least three of the five Eukaryotic initiation factor 2B genes, EIF2B 2, EIF2B 4, EIF2B 5. Screens of women with isolated -  -  -  idiopathic POF have revealed putative functional sequence mutations in a handful of cases for some of these genes; however they do not appear to be common causes of POF. Two X chromosome genes suggested to play a role in POF are KAL, associated with Kallman's syndrome and DAX 1, associated with X linked adrenal hypoplasia. Both of -  these disorders cause a deficiency of GnRH, which could presumably cause a failure of the hypothalamus to stimulate FSH release by the pituitary (Anasti. 1998), although no cases in non-syndromic POF patients have been described. POF has also been implicated in autoimmune disorders; therefore immune related genes (Table 1.4) may also be implicated in its pathogenesis. Addison's disease (adrenal  22  insufficiency) occurring with other autoimmune endocrine disorders makes up a specific condition known as autoimmune polyendocrine syndrome (APS). APS has been divided into three clinical subtypes all of which are associated with increased risk of POF. APS I is an autosomal recessive disorder caused by mutations in the AIRE gene; no genes have been identified for APS II or APS III, although risk may vary based on HLA haplotype (Goswami and Conway. 2005). POF has been reported in patients with hypothyroidism, myasthenia gravis, Crohn's diseases, vitiligo, rheumatoid arthritis, systemic lupus erythematosis, and pernicious anemia. Specifically, autoimmune destruction of cells required for steroid synthesis, including those producing the enzymes cholesterol desmolase, 17a-hydroxylase, 21-hydroxylase, and 30-hydroxysteroid dehydrogenase, have been implicated in POF patients with varying degrees of ovarian dysfunction (Davis. 1996; Goswami and Conway. 2005). It is not surprising therefore, that POF has been associated with major histocompatibility (MHC) class II markers commonly implicated in autoimmune disease (Anasti et al. 1994). In summary, although multiple single genes have been suggested to have a putative role in POF pathogenesis, for the majority of POF cases no single genetic cause is found to be responsible. As expected, genes located on the X chromosome, and genes with roles in hormone signaling necessary for follicle growth and development have been implicated in POF; however with the exception of FMRI, none of these genes have been found to be responsible for a more than a few isolated cases. In the cases of POF in association with other genetic disease, the genes identified have only very rarely been found to be responsible for isolated POF, suggesting that they are unlikely candidates for the otherwise healthy patient presenting with POF. It is possible that other non-sequence  23  related genetic factors play a role in ovarian failure, or that multiple genetic factors interact to affect age at menopause in any one individual.  1.5 Aging and Reproductive aging There is mounting evidence supporting a relationship between duration of reproductive capability and total lifespan. Both animal models and human epidemiological studies support the suggestion that longevity is associated with an increase in reproductive lifespan. Although lifespan in the normal range is highly dependent on environment, studies suggest that extreme longevity appears to be largely independent of environmental influence and is likely to be determined by genetic factors. Therefore genetic factors that are associated with general aging may be related to reproductive aging as well.  1.5.1 Evidence for a link between chronological age and reproductive age Studies in mice and flies support a link between reproductive duration and lifespan. Lines of mice selectively bred for reproductive longevity have not only significant lengthening of reproductive life but also a 17% increase in total lifespan when compared to control mice (Nagai et al. 1995). Similarly, continued selective breeding of female Drosophila that lay eggs at the oldest age results in an overall increase in fly life (Hutchinson and Rose. 1991). These experiments support the suggestion that genetic factors controlling reproductive longevity may also influence lifespan. Human population studies also support a link between reproductive longevity and lifespan. A number of studies have reported that higher total fecundity (the number of children a woman has in her lifetime) is related to longevity (Manor et al. 2000; Muller et  24  al. 2002). A high overall fecundity may represent an unusually long reproductive period; however it is an indirect and biased measure of reproductive lifespan, possibly explaining the fact that other studies failed to find an association between fecundity and longevity (Helle et al. 2002). To assess reproductive lifespan more directly, age at the time a woman has her last child can be examined. A number of studies have found a positive association between age at last reproduction and lifespan (Doblhammer. 2000; Muller et al. 2002; Smith et al. 2002). Most notably, a study of female centenarians found that women living to at least 100 are four times more likely to have had a child while in their forties than women living to age 73 (Penis et al. 1997). Furthermore, late menopause has also been associated with decreased mortality and increased post-reproductive lifespan (Cooper and Sandler. 1998; Jacobsen et al. 1999; Snowdon et al. 1989). In an ageadjusted prospective study of 5287 females, women who experienced menopause before 40 had nearly twice the risk of dying during the study period (a 6 year span) than those experiencing menopause at ages 50-54 (Snowdon et al. 1989). Population studies therefore support a link between reproductive longevity, substantiated by total fecundity, age at last reproduction, or age at menopause. There are two conflicting explanations for these trends in animal and human studies. Firstly, effective age of the ovary could directly affect longevity. This hypothesis was first suggested in studies of Caenorhabditis elegans (Hsin and Kenyon. 1999) and is supported by a mouse study in which ovaries from 2 month old mice were found to increase life expectancy by 60% when transplanted into ovariectomized 11 month old mice (Cargill et al. 2003). Alternatively, having children at a relatively late reproductive age may be an indicator of slow reproductive aging and slow aging in  25  general. It is not the act of childbearing at a late age that causes an increase in lifespan, but rather the ability to have a child at this age indicates slow rate of aging (Penis et al. 2002). Extreme longevity has not been directly selected for but has been a consequence of selection for genes that maximize a woman's reproductive years (Penis and Fretts. 2001). Thus, women experiencing POF may be considered "prematurely aged" and genetic factors that are associated with general aging may be related to reproductive aging as well. 1.5.2 Genetic factors associated with aging  Genetic factors reported to be associated with aging include chromosome features such as telomere length, skewed X chromosome inactivation, and DNA methylation, as well as common variants in the Apolipoprotein E gene, APOE. 1.5.2.1 Telomere length  Telomeres are specialized repetitive DNA sequences that cap and protect the ends of linear chromosomes. In humans telomeres consist of a terminal (TTAGGG) n DNA sequence at the end of the chromosome that repeats for 10-15 Kb (Saldanha et al. 2003). This repetitive sequence binds the proteins which are necessary to protect the ends of the chromosome. These proteins stimulate the end of the chromosome to form a lariat structure, known as a t loop in which chromosome ends are not exposed and are therefore prevented from being recognized as double-strand DNA breaks and being targeted for repair. Additional telomeric secondary structure also exists to pack telomere DNA into stable compact chromosome elements. Other postulated functions for telomere binding proteins include determination of chromosomal location in the nucleus and prevention of illegitimate telomere recombination. 26  Telomere length is heritable, highly variable, and negatively correlated with age. Telomeres have a critical role in the regulation of cellular replicative capacity by acting as a "mitotic clock" for cellular DNA (Harley et al. 1992). In every round of DNA replication, the final repeats on the end of each telomere fail to be replicated by DNA polymerase; therefore telomeres shorten with each cellular division and limit the number of replications. Telomere length in peripheral blood cells decreases with age at a rate of approximately 31 by per year (Slagboom et al. 1994). Based on twin and family studies, variation in telomere length is estimated to be at least 78% heritable (Slagboom et al. 1994) and may have a paternal mode of inheritance (Nordtjall et al. 2005). Though overall there is a decrease in telomere length with increased age of the population examined, there is also a high level of interindividual variability at any given age (Hastie et al. 1990). The variability in telomere length may be due to differing telomere lengths at conception, variation in telomerase activity during development, and differences in the rate of cell division between individuals. Furthermore, a number of factors including paternal age, life stress, and oxidative stress also have been found to influence telomere length (Epel et al. 2004; Unryn et al. 2005). Telomere length can be extended by telomerase, an enzyme that is able to add additional TTAGGG repeats to the ends of telomere. Telomerase is a ribonucleoprotein composed of telomerase RNA which is used as a template for telomere elongation by telomerase reverse transcriptase (TERT) (Cohen et al. 2007). Telomerase is only expressed in embryonic stem cells in early development and in germ line and cancer cells in the adult. Ovarian telomerase activity has been shown to decline with age and lower telomerase activity than expected for age has been reported in the ovaries of women  27  experiencing POF due to ovarian depletion (Kinugawa et al. 2000). A recombination based method of telomere lengthening not requiring telomerase has recently been discovered and is known as alternative lengthening of telomeres (ALT), however epigenetic modifications of telomeric and subtelomeric DNA prevent normal cells from undergoing telomere elongation by this manner (Muntoni and Reddel. 2005).  1.5.2.2 Skewed X chromosome inactivation X chromosome inactivation (XCI) is a process in which one of the X chromosomes present in every female cell is inactivated during early embryogenesis. XCI achieves dosage compensation so that there is approximately equal expression of X chromosome genes in the cells of males who have a single X chromosome, and females who carry two X chromosomes (Lyon. 1961). The choice of which X to inactivate is a random process, such that both the maternal and paternal X chromosomes have an equal chance of being inactivated in any given cell. Following the establishment of XCI, the inactivation status of a particular X chromosome is stably maintained through out all subsequent cell divisions. A result of the XCI process is that all women are essentially mosaic, composed of two populations of cells, one of which has the paternally inherited X chromosome active, and one which has the maternally inherited X chromosome active. The process of X inactivation is commonly broken down into three stages: marking of which X will be inactivated, inactivation of the marked X (Xi), and maintenance of inactivation. The initial marking of which X will be inactivated is not yet entirely understood, although genes in the X chromosome inactivation center (XIC) are believed to control the process (Willard. 1996). A primary player in the XIC is the X inactivation specific transcript, XIST, an untranslated functional RNA which is expressed -  28  from and coats the Xi (Brown et al. 1991). In early embryonic development there is low level expression of XIST from both X chromosomes (Daniels et al. 1997). XIST expression from the X chromosome which will remain active (Xa) is then down regulated, possibly by the actions of its antisense gene, TS/X; although this is suggested by studies in mice and has not been conclusively proven in humans (Lee et al. 1999). Included in the marking process there is necessarily a method of counting, as only a single X remains active regardless of the number of X chromosomes present. Following marking, inactivation will progress by the spread of the XIST transcripts along the Xi. Spreading of XIST is thought to progress with the help of LINE repetitive elements along the X chromosome that act as "way stations" to allow enhancement of the inactivation signal (Lyon. 1998). Following establishment of the Xi, inactivation of most Xi genes is maintained by the formation of heterochromatin, a repressive chromatin state characterized by DNA methylation at gene promoters and specific histone modifications (reviewed in Brown. 2001). Although the establishment of inactivation is essentially random, most females have a detectable bias in the proportion of cells with either parental X chromosome inactivated. A deviation from random XCI is termed skewed XCI, and is defined variably in the literature using arbitrary cut offs of 75%, 80% or 90% bias in skewing. XCI is established early in embryonic development at the 14-16 cell stage (Tonon et al. 1998) and as a result skewed XCI would be infrequently expected to occur by chance due to the limited number of cells present when XCI is established. Skewed XCI could also theoretically result from primary causes such as imprinting effects or XIST ITSIX mutations that result in a bias in which X is selected for inactivation. However it may  29  also result from secondary causes such as selection for cells which have a particular X chromosome active. Secondary causes of XCI skewing include selection against cells in which the Xa carries a deletion, or gene mutation, or where the Xi has an autosomal translocation. Furthermore, any developmental event causing a reduction in the cell precursor population, such as monozygotic twinning or trisomic rescue could result in skewed XCI (Brown and Robinson. 2000). The degree of XCI skewing, at least in peripheral blood cells, is known to increase with age. In particular, the frequency of skewed XCI >90% is 2-3% in neonates, but increases to 20-30% in women past 65 years of age (Busque et al. 1996). The agerelated increase in skewing could be due to either slight selective differences in growth between populations of cells inactivating one X or the other that become more pronounced over time, or a result of hematopoietic stem cell senescence that causes circulating blood cells to be derived from a diminishing stem cell pool (Brown and Robinson. 2000). However, there is evidence to contradict both of these suggested mechanisms. The age-related increase in skewed X chromosome inactivation occurs in a nonlinear manner, with an acceleration of the rate of increase beginning at some point between the ages of 30 and 60 years (Hatakeyama et al. 2004; Kristiansen et al. 2005). This argues against a cell selection model which would predict that skewed XCI would increase at a slower rate with time (Hatakeyama et al. 2004; Sandovici et al. 2004). Studies examining XCI skewing in twin pairs have not revealed a increase in correlation between twins with age which would be expected with a cell selection model, or a decrease in correlation with age which would be expected if stochastic events resulting from reduced follicular pool were responsible (Christensen et al. 2000; Kristiansen et al.  30  2005). It seems likely that combined effects of both of these mechanisms play a role in increasing skewed XCI with age (Kristiansen et al. 2005).  1.5.2.3 DNA methylation DNA methylation is a form of epigenetic DNA modification comprised of the addition of a methyl group to the cytosine residue, typically occurring where the DNA sequence consists of a cytosine followed by a guanine (CpG). Most unmethylated CpG sites are found in CpG islands in the promoter regions of genes, where methylation of these regions is classically associated with silencing of gene expression (Richardson. 2003). Methylation mediated gene silencing is achieved by preventing the binding of transcription factors to gene promoter regions and inducing the formation of heterochromatin, a repressive chromatin conformation (Eng et al. 2000). However, 7080% of CpG sites in the remainder of the genome are maintained in a methylated state (Richardson. 2003). DNA methylation is instituted by the work of DNA methyltransferases including Dnmtl, Dnmt3a and Dnmt3b which mediate transfer of methyl groups to cytosines (reviewed in Richardson. 2003). Dnmtl is responsible for maintenance of DNA methylation, primarily methylating newly synthesized DNA according to the methylation pattern on the parent strand. Dnmt3a and Dnmt3b on the other hand are involved in de novo DNA methylation established during cellular differentiation during development. Cytosine methylation is removed primarily by passive demethylation when Dnmtl is no longer able to methylate newly replicated DNA, but may also be achieved by proteins actively functioning as DNA demethylases. DNA methylation patterns are known to change with age. Assays of total methylated cytosine content have shown that overall there is global demethylation of the  31  human genome with age (Golbus et al. 1990). However these assays do not determine whether these methylation changes occur in regions of the genome relevant to gene transcription so the significance of these findings remains unclear (Richardson. 2003). How DNA methylation of CpG islands changes with age is less clear and may be gene specific. CpG islands of a number of tumor suppressor genes have been found to become hypermethylated with age (Issa et al. 1994), reducing their expression and possibly providing an explanation for increased incidence of malignancy with age. In contrast, demethylation with age has been reported for retrotransposons and endogenous retroviruses normally repressed by hypermethylation (Barbot et al. 2002). DNA methylation changes with age may have endogenous origins such as decreased expression of DNA methyltransferases. Expression of both Dnmtl and Dnmt3a has been reported to decrease with age (Zhang et al. 2002), which would result in a passive demethylation of the genome. In contrast, there is also evidence that age related demethylation is the result of environmental or dietary effects. DNA hypomethylation can result from a diet deficient in folate, choline, or methionine (Cooney. 2001). A number of common environmental toxins have been reported to alter chromosomal epigenetic modifications. Oocyte maturation and fertilization involves a number of chromatin reprogramming events, and if the cellular processes involved in DNA methylation were affected by a woman's age, then these changes might also affect reproductive senescence.  1.5.2.4 Gene variants Human lifespan likely has a genetic component. Heritability of life expectancy has been estimated at 20-30% in a study examining lifespan for individuals living to  32  average life expectancy (McGue et al. 1993) although it may be much higher for those with extreme longevity (Perls et al. 2002). For those individuals that are able to survive to very old age mortality actually begins to decline as individuals in poor health begin to drop out of the population. This leaves behind those with the genetic composition that facilitates remaining healthy into old age (Perls et al. 2002). Extreme longevity tends to run in families, with siblings, parents and children of centenarians having a much higher chance of living to old age than those in the general population. There may be a limited number of genes contributing to extreme longevity, whereas lifespan for most individuals is a complex trait involving many genes and environmental interactions. The apolipoprotein E gene, APOE, is associated with longevity. APOE encodes a plasma protein involved in cholesterol metabolism and transport, and has three common alleles encoding different protein isoforms, s2 s3 and s4. The s4 allele has been associated with increased risk for Alzheimer and cardiovascular disease (Christensen et al. 2006). Notably this allele has also been associated with increased risk of trisomic pregnancy (Avramopoulos et al. 1996; Nagy et al. 2000). Trisomy is the presence of an additional chromosome, and risk of concieving a child with a trisomy increases greatly with maternal age (Hassold et al. 1996). The association of APOE genotype and trisomic pregnancy may therefore suggest that its role in longevity reflects a parallel role in reproductive aging. 1.6 Research objectives The purpose of this study is to examine a number of candidate genetic factors for association with POF. Specifically, I hypothesize that genetic variants associated with endocrine function or with aging contribute to POF pathogenesis. The genetic factors  33  included fall into four broad categories which make up the chapters of this thesis: 1) the Fragile X mental retardation 1 (FMR1) gene, 2) endocrine related gene polymorphisms, 3) skewed X chromosome inactivation and 4) age-related chromosome factors and gene variants. The objectives are therefore four-fold: 1) To examine the role of the FMR1 gene in POF, by assessing FMR1 allele distribution, promoter methylation, and gene expression in a population of POF patients. 2) To assess a population of POF patients and control women for genetic association with a number of endocrine related gene variants, including those in the FSH receptor, Androgen receptor, Estrogen receptor a, Estrogen receptor A and Sex hormone binding globulin genes. 3) To examine the incidence of skewed X chromosome inactivation among POF patients and the presence of copy number abnormalities in individuals that exhibit skewing, in order to determine whether cryptic X chromosome abnormalities play a significant role in ovarian failure. 4) To determine if POF is associated with indicators of cellular aging including telomere length, XCI skewing, DNA methylation at the AR locus, and APOE genotype. Determining the genetic factors associated with POF is an important step in understanding the mechanisms involved in determining age at onset of menopause. This knowledge is critical for the development of treatment options to restore fertility in women with POF as well as predictive testing for those at risk.  34  Table 1.1 Stages of follicle development Stage^Description^  Time to next Diameter^stage  Primordial^Basal lamina separates follicle from ovarian tissue ^ Squamous granulosa cells surround oocyte Immature dormant oocyte  40 gm^<50 years  Primary  Granulosa cells become cuboidal^ Oocyte begins to mature transcription of oocyte genes begins zona pellucida begins to form around oocyte  100 gm^<120 days  Secondary  Thecal cells are recruited surround the basal lamina differentiate into theca interna and theca externa capillary network forms between interna and externa Additional layers of granulosa cells are formed Full grown oocyte surrounded by zona pellucida  200-400 gm^—71 days  Tertiary/Antral  Thecal and granulosa cells continue to divide as the follicle increases in size ^2-5 mm^—14 days Granulosa cells differentiate in response to FSH Thecal cells produce androgens in response to LH The antrum, a fluid filled cavity, forms around the oocyte  Graafian  A single follicle becomes dominant, all others undergo atresia Continued growth in size No further cell differentiation Formation of the stigma, through which the oocyte will be excreted  10-20 mm  Table 1.2 X chromosome genes implicated in POF  ^ Key references Gene^Locus Implicated by association 16-20% of premutation carriers experience POF; 7% of sporadic and 21% of familial POF cases are FMR1^Xq27.3 carriers of premutation alleles (reviewed in Sherman 2000) Small repeat sizes found at increased frequency in POF patients (Murray et al. 1999; Murray et al. 1998) FMR2^Xq28 Implicated by translocation Disrupted in a family in which POF segregates with a X;12 balanced translocation (Bione et al. 1998) DIAPH2^Xq22 Drosophila null mutatnts have a defective oogenesis and sterility (Castrillon and Wasserman 1994) Disrupted in at least 2 X;autosome translocation carriers with POF (Prueitt et al. 2000, Mumm et al. 2001) XPNPEP2^Xq25 POF1B^Xq21 Disrupted in a X;1 balanced translocation carrier with POF; no significant association with any gene variants found in 223 POF patients (Bione et al. 2004) DACH2^Xq21.3 Disrupted in a X;7 translocation carrier with POF (Prueitt et al. 2002) Association with 5 coding region variants found in 257 POF patients (Bione et al. 2004) ^ Xq21.2 CHM Disrupted in a X;4 balanced translocation carrier with choroideremia and POF (Lorda-Sanchez et al. 2000) Implicated by function BMP15/GDF9B Xp11.2 Growth differentiation factor expressed in oocytes (Otsuka et al. 2000) BMP15 mutation causes infertility in sheep due to a block in folliculogenesis (Galloway et al. 2000) Mutation in sisters with ovarian dysgenesis and primary amenorrhea (Di Pasquale et al. 2004) Population screens show 9 out of 203 (Laissue et al. 2006) and 7 out of 166 (Di Pasquale et al. 2006) POF patients carry gene variants not seen in controls. Haplotype analysis of common variants in 133 POF patients reveals association (Dixit et al. 2006) No mutations found in screens of 38 (Chand et al. 2006) or 15 POF patients (Takebayashi et al. 2000) DFFRX/USP9X Xp 11.4 Deletion in Y chromosome homologue, DFFRY , associated with azoospermia in males (Brown et al. 1998) Escapes X inactivation; however single copy appears sufficient for ovarian function (James et al. 1998) ZFX Xp22.3Knockout mice have reduced follicular pool, shortened reproductive lifespan (Luoh et al. 1997) p21.3 No functional mutations found in 52 POF patients (Davison et al. 1999) FSHPRHJ Xq22 Expressed in developing ovary before FSH receptor in response to FSH (Roberts et al. 1996) XIST Xq13.2 Responsible for X chromosome inactivation (Brown et al. 1991); mutations could cause skewing AGTR2 Xq22-q23 High expression in granulosa cells of atretic follicles in rats (Tanaka et al. 1995) No mutations found in 2 cases of familial POF (Katsuya et al. 1997) SOX3 Xq26.3 Possible role in ovary development; homology to SRY, critical for testis development (Gubbay et al. 1990) No mutations found in 164 POF patients (Goswami and Conway 2005)  Table 1.3 Autosomal genes implicated in POF  Gene  Locus^Key references 2p21-p16^Missense mutations reported in six Finnish families with severe POF (Aittomaki et al. 1995) Case reports (Beau et al. 1998; Touraine et al. 1999; Doherty et al. 2002; Allen et al. 2003; Meduri et al. 2003) No previously identified mutations found in 35 POF patients (Layman et al. 1998), 49 UK patients (Conway et al. 1999), 15 Japanese patients (Takakura et al. 2001), or 20 Argentine POF patients (Sundblad et al. 2004) No linkage with age at menopause found in 126 sibling pairs (Kok et al. 2004) 1 1p13^Homozygous mutation in a woman with primary amenorrhea (Matthews et al. 1993) No mutations found in 18 POF patients (Layman et al. 1993) 2q33-q36^Missense mutation G769A, found in 3 of 43 POF patients (Shelling et al. 2000) G769A more common in 157 POF patients than in controls (Marozzi et al. 2002) but not found in 80 Korean POF patients (Jeong et al. 2004); A257T mutation found in 9 of 80 Indian POF patients (Dixit et al. 2004) Promoter haplotype associated with POF (Harris et al. 2005) No association with promoter haplotype; G769A mutation more frequent in controls (Sundblad et al. 2006) 7p15-p13^No functional mutations found in 80 POF patients (Dixit et al. 2004) or 43 POF patients (Shelling et al. 2000) 2cen-q13^No functional mutations found in 80 POF patients (Dixit et al. 2004) or 43 POF patients (Shelling et al. 2000) 2p21^Inactivating mutations cause menstrual abnormalities and ovulation failure (Arnhold et al. 1999) 19q13.32^Polymorphism associated with POF in 245 Japanese patients (Takahashi et al. 1999) Mutation causing altered LH response found in women with ovulatory disorders (Takahashi et al. 2004) 12q13^No functional AMHR2 mutations found in 16 Japanese POF patients (Wang et al. 2002) -482G SNP associated with age at menopause in 248 postmenopausal Dutch women (Kevenaar et al. 2007) -482G associated with menopausal age in interaction with parity in 2381 Dutch women (Kevenaar et al. 2007)  FSHR  FSTIfi INHa  INH/3A INHP LHCGR LHfl AMHR2  Steroid enzymes: CYP1 1 A CYP17A1 CYP19A1 NOBOX  FOXO3A  ^  15g23-24 10q24.3 15q21.2 7q35  6q21  Deficiency in cholesterol desmolase can cause amenorrhea (Anasti 1998) Case report of 17a-hydroxylase compound heterozygotes with primary amenorrhea (Perez et al. 2004) Case report of siblings with hypergonadotropic amenorrhea and aromatase mutation (Morishima et al. 1995) Deficiency blocks folliculogenesis and causes ovarian failure in mouse (Rajkovic et al. 2004) Expressed in human ovary, specifically in the oocyte (Huntriss et al. 2006) Rare variants found at increased frequency in 96 POF patients (Qin et al. 2007) Knockout mouse has global follicle activation causing premature follicle depletion (Castrillon et al. 2003) Two potentially causal mutations found in 90 POF patients (Watkins et al. 2006) Haplotype not associated with fertility or fecundity in 701 women > 85 years old (Kuningas et al. 2007)  Gene^Locus^Key references FOXOJA^13q14.1^Regulator of granulosa cell growth during development (Cunningham et al. 2004) One potentially causal mutation found in 90 POF patients (Watkins et al. 2006) Haplotype not associated with fertility or fecundity in 701 women > 85 years old (Kuningas et al. 2007) FIGLA^2p12^A transcriptional regulator expressed in the ovary in early development (Choi and Rajkovic 2006) Mouse knockout has rapid loss of oocytes after birth, no primordial follicle development (Soyal et al. 2000) 3 variants found in 4 of 100 Chinese POF patients, 2 of these were not found in 304 controls (Zhao et al. 2007) FOXE]^9q22^Polyalanine tract repeat length is associated with POF in 250 POF patients (Watkins et al. 2006) GDF9^5q31.1^Infertility in GDF9 deficient mice due to blockage at primary follicle stage (Dong et al. 1996) No mutation found in 15 Japanese POF patients (Takebayashi et al. 2000) Missense mutation found in one of 203 (Laissue et al. 2006) and one of 60 (Kovanci et al. 2007) POF patients Two rare mutations and a haplotype associated with POF in 127 patients (Dixit et al. 2005) No mutations or association found in 38 New Zealand POF patients (Chand et al. 2006)  Table 1.4 Genetic diseases associated with POF  Gene^Locus^Disease and key references ^ 9p13^Galactosemia GALT POF develops in 60-80% of patients (Goswami and Conway 2005) ^ FOXL2 3q22-q23 Blepharophimosis-ptosis-epicanthus inversus (BPES) BPES type I is clinically defined by the presence of ovarian dysfunction & infertility (Zlotogora et al. 1983) Expressed in adult ovarian follicles (Crisponi et al 2001) Specific mutations predict presence of POF in addition to BPES (De Baere et al. 2003) Knockout mice have BPES and ovarian failure due to absence of follicle development (Uda et al. 2004) Two missense mutations found in 70 patients with isolated POF (Harris et al. 2002) No functional mutations found in studies of 30 cases of isolated POF (De Baere et al. 2001), 70 POF patients (De ^ Baere et al. 2002) or 120 POF patients (Bodega et al. 2004) EIF2B genes: Childhood ataxia with central nervous system hypomyelination/vanishing white-matter leukodystrophy ^ EIF2B2 ^ 14q24^Patients are at increased risk of ovarian failure (Fogli et al. 2003) EIF2B4^2p23.3^No mutations found in any EIF2B genes in 93 patients with isolated POF (Fogli et al. 2004) EIF2B5 ^3q27 15q25^Progressive external opthalmoplegia POLG Associated with POF (Luoma et al. 2004; Pagnamenta et al. 2006) ^ 17q22^Proximal symphalagism (SYM1) NOG Case report of a SYM1 patient with POF (Kosaki et al. 2004) ^ Expressed in the ovary and an antagonist for ovarian bone morphogenetic proteins (Goswami and Conway 2005) PPM genes: Carbohydrate-deficient glycoprotein syndrome type 1 (CDG1) ^ PMM1 ^22q13^Defiency of phosphomannomutase causes defective glycosylation and a cluster of phenotypic features including PMM2 ^ 16p13^hypogonadism (Laml et al. 2002) KAL Xp22.3^Kallman's syndrome GnRH deficiency could presumably cause a failure to stimulate pituitary FSH release (Anasti 1998) ^ DAX-1 Xp21.3^X linked adrenal hypoplasia GnRH deficiency could presumably cause a failure to stimulate pituitary FSH release (Anasti 1998) ^ AIRE 21q22.3 Autoimmune polyendocrine syndrome I (APSI) <60% of patients have ovarian failure (Perheentupa 1996) ^ 3-10% of APS type II patients have POF (Betterle et al. 2004) HLA II genes 6p21 Autoimmune susceptibility HLA-DGB1*301 and -DQB1*603 associated with 313HSD autoimmunity in POF (Arif et al. 1999) No association between POF and HLA-DR3 and -DR4 was found (Anasti et al. 1994)  hypothalamus GnRH  estrogen progesterone  pituitary LH  ^  FSH  estrogen ^inhibin  ovary Figure 1.1 Hormone regulation of oocyte maturation. Pulsatile release of gonadotropin releasing hormone (GnRH) from the hypothalamus stimulates pituitary secretion of follicle stimulating hormone (FSH) and luteinizing hormone (LH) which promote follicle maturation. The dominant follicle releases estrogen and inhibin which down regulate GnRH and FSH and result in an LH surge that initiates ovulation. The emptied follicle develops into the corpus luteum and release progesterone and estrogen. Absence of oocyte fertilization results in declining levels of ovarian hormones allowing GnRH and FSH levels to rise again.  40  1.7 References Aittomaki K, Herva R, Stenman UH, Juntunen K, Ylostalo P, Hovatta 0 and de la Chapelle A. 1996. Clinical features of primary ovarian failure caused by a point mutation in the follicle-stimulating hormone receptor gene. J. Clin. Endocrinol. Metab. 81:3722-3726. Aittomaki K, Lucena JL, Pakarinen P, Sistonen P, Tapanainen J, Gromoll J, Kaskikari R, Sankila EM, Lehvaslaiho H, Engel AR, et al. 1995. Mutation in the folliclestimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell. 82:959-968. Allen EG, Sullivan AK, Marcus M, Small C, Dominguez C, Epstein MP, Charen K, He W, Taylor KC and Sherman SL. 2007. Examination of reproductive aging milestones among women who carry the FMR1 premutation. Hum. Reprod. Allen LA, Achermann JC, Pakarinen P, Kotlar TJ, Huhtaniemi IT, Jameson JL, Cheetham TD and Ball SG. 2003. A novel loss of function mutation in exon 10 of the FSH receptor gene causing hypergonadotrophic hypogonadism: clinical and molecular characteristics. Hum. Reprod. 18:251-256. Allingham-Hawkins DJ, Babul-Hirji R, Chitayat D, Holden JJ, Yang KT, Lee C, Hudson R, Gorwill H, Nolin SL, Glicksman A, et al. 1999. Fragile X premutation is a significant risk factor for premature ovarian failure: the International Collaborative POF in Fragile X study--preliminary data. Am. J. Med. Genet. 83:322-325. Anasti JN. 1998. Premature ovarian failure: an update. Fertil. Steril. 70:1-15. Anasti JN, Adams S, Kimzey LM, Defensor RA, Zachary AA and Nelson LM. 1994. Karyotypically normal spontaneous premature ovarian failure: evaluation of association with the class II major histocompatibility complex. J. Clin. Endocrinol. Metab. 78:722-723. Arif S, Underhill JA, Donaldson P, Conway GS and Peakman M. 1999. Human leukocyte antigen-DQB 1 * genotypes encoding aspartate at position 57 are associated with 3beta-hydroxysteroid dehydrogenase autoimmunity in premature ovarian failure. J. Clin. Endocrinol. Metab. 84:1056-1060. Arnhold IJ, Latronico AC, Batista MC and Mendonca BB. 1999. Menstrual disorders and infertility caused by inactivating mutations of the luteinizing hormone receptor gene. Fertil. Steril. 71:597-601. Avramopoulos D, Mikkelsen M, Vassilopoulos D, Grigoriadou M and Petersen MB. 1996. Apolipoprotein E allele distribution in parents of Down's syndrome children. Lancet. 347:862-865.  41  Baker TG, Challoner S and Burgoyne PS. 1980. The number of oocytes and the rate of atresia in unilaterally ovariectomized mice up to 8 months after surgery. J. Reprod. Fertil. 60:449-456. Barbot W, Dupressoir A, Lazar V and Heidmann T. 2002. Epigenetic regulation of an IAP retrotransposon in the aging mouse: progressive demethylation and de-silencing of the element by its repetitive induction. Nucleic Acids Res. 30:2365-2373. Beau I, Touraine P, Meduri G, Gougeon A, Desroches A, Matuchansky C, Milgrom E, Kuttenn F and Misrahi M. 1998. A novel phenotype related to partial loss of function mutations of the follicle stimulating hormone receptor. J. Clin. Invest. 102:13521359. Betterle C, Lazzarotto F and Presotto F. 2004. Autoimmune polyglandular syndrome Type 2: the tip of an iceberg? Clin. Exp. Immunol. 137:225-233. Bione S, Rizzolio F, Sala C, Ricotti R, Goegan M, Manzini MC, Battaglia R, Marozzi A, Vegetti W, Dalpra L, et al. 2004. Mutation analysis of two candidate genes for premature ovarian failure, DACH2 and POF1B. Hum. Reprod. 19:2759-2766. Bione S, Sala C, Manzini C, Arrigo G, Zuffardi 0, 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. Bodega B, Porta C, Crosignani PG, Ginelli E and Marozzi A. 2004. Mutations in the coding region of the FOXL2 gene are not a major cause of idiopathic premature ovarian failure. Mol. Hum. Reprod. 10:555-557. Brown CJ. 2001. Equality of the sexes: mammalian dosage compensation. Semin. Reprod. Med. 19:125-132. Brown CJ, Ballabio A, Rupert JL, Lafreniere RG, Grompe M, Tonlorenzi R and Willard HF. 1991. A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature. 349:38-44. Brown CJ and Robinson WP. 2000. The causes and consequences of random and nonrandom X chromosome inactivation in humans. Clin. Genet. 58:353-363. Brown GM, Furlong RA, Sargent CA, Erickson RP, Longepied G, Mitchell M, Jones MH, Hargreave TB, Cooke HJ and Affara NA. 1998. Characterisation of the coding sequence and fine mapping of the human DFFRY gene and comparative expression analysis and mapping to the Sxrb interval of the mouse Y chromosome of the Dffry gene. Hum. Mol. Genet. 7:97-107.  42  Busque L, Mio R, Mattioli J, Brais E, Blais N, Lalonde Y, Maragh M and Gilliland DG. 1996. Nonrandom X-inactivation patterns in normal females: lyonization ratios vary with age. Blood. 88:59-65. Cargill SL, Carey JR, Muller HG and Anderson G. 2003. Age of ovary determines remaining life expectancy in old ovariectomized mice. Aging Cell. 2:185-190. Castrillon DH, Miao L, Kollipara R, Homer JW and DePinho RA. 2003. Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a. Science. 301:215-218. Castrillon DH and Wasserman SA. 1994. Diaphanous is required for cytokinesis in Drosophila and shares domains of similarity with the products of the limb deformity gene. Development. 120:3367-3377. Chand AL, Ponnampalam AP, Harris SE, Winship IM and Shelling AN. 2006. Mutational analysis of BMP15 and GDF9 as candidate genes for premature ovarian failure. Fertil. Steril. 86:1009-1012. Choi Y and Rajkovic A. 2006. Genetics of early mammalian folliculogenesis. Cell Mol. Life Sci. 63:579-590. Christensen K, Johnson TE and Vaupel JW. 2006. The quest for genetic determinants of human longevity: challenges and insights. Nat. Rev. Genet. 7:436-448. Cohen SB, Graham ME, Lovrecz GO, Bache N, Robinson PJ and Reddel RR. 2007. Protein composition of catalytically active human telomerase from immortal cells. Science. 315:1850-1853. Conway GS, Conway E, Walker C, Hoppner W, Gromoll J and Simoni M. 1999. Mutation screening and isoform prevalence of the follicle stimulating hormone receptor gene in women with premature ovarian failure, resistant ovary syndrome and polycystic ovary syndrome. Clin. Endocrinol. (Oxf). 51:97-99. Conway GS, Kaltsas G, Patel A, Davies MC and Jacobs HS. 1996. Characterization of idiopathic premature ovarian failure. Fertil. Steril. 65:337-341. Cooney CA. 2001. Dietary selenium and arsenic affect DNA methylation. J. Nutr. 131:1871-1872. Cooper GS and Sandler DP. 1998. Age at natural menopause and mortality. Ann. Epidemiol. 8:229-23 5. Coulam CB, Adamson SC and Annegers JF. 1986. Incidence of premature ovarian failure. Obstet. Gynecol. 67:604-606. Cramer DW and Xu H. 1996. Predicting age at menopause. Maturitas. 23:319-326.  43  Cramer DW, Xu H and Harlow BL. 1995a. Does "incessant" ovulation increase risk for early menopause? Am. J. Obstet. Gynecol. 172:568-573. Cramer DW, Xu H and Harlow BL. 1995b. Family history as a predictor of early menopause. Fertil. Steril. 64:740-745. Crisponi L, Deiana M, Loi A, Chiappe F, Uda M, Amati P, Bisceglia L, Zelante L, Nagaraja R, Porcu S, et al. 2001. The putative forkhead transcription factor FOXL2 is mutated in blepharophimosis/ptosis/epicanthus inversus syndrome. Nat. Genet. 27:159-166. Cunningham MA, Zhu Q and Hammond JM. 2004. FoxOla can alter cell cycle progression by regulating the nuclear localization of p27kip in granulosa cells. Mol. Endocrinol. 18:1756-1767. Daniels R, Zuccotti M, Kinis T, Serhal P and Monk M. 1997. XIST expression in human oocytes and preimplantation embryos. Am. J. Hum. Genet. 61:33-39. Davis CJ, Davison RM, Payne NN, Rodeck CH and Conway GS. 2000. Female sex preponderance for idiopathic familial premature ovarian failure suggests an X chromosome defect: opinion. Hum. Reprod. 15:2418-2422. Davis SR. 1996. Premature ovarian failure. Maturitas. 23:1-8. Davison RM, Davis CJ and Conway GS. 1999. The X chromosome and ovarian failure. Clin. Endocrinol. (Oxf). 51:673-679. De Baere E, Beysen D, Oley C, Lorenz B, Cocquet J, De Sutter P, Devriendt K, Dixon M, Fellous M, Fryns JP, et al. 2003. FOXL2 and BPES: mutational hotspots, phenotypic variability, and revision of the genotype-phenotype correlation. Am. J. Hum. Genet. 72:478-487. De Baere E, Dixon MJ, Small KW, Jabs EW, Leroy BP, Devriendt K, Gillerot Y, Mortier G, Meire F, Van Maldergem L, et al. 2001. Spectrum of FOXL2 gene mutations in blepharophimosis-ptosis-epicanthus inversus (BPES) families demonstrates a genotype--phenotype correlation. Hum. Mol. Genet. 10:1591-1600. De Baere E, Lemercier B, Christin-Maitre S, Durval D, Messiaen L, Fellous M and Veitia R. 2002. FOXL2 mutation screening in a large panel of POF patients and XX males. J. Med. Genet. 39:e43. de Bruin JP, Bovenhuis H, van Noord PA, Pearson PL, van Arendonk JA, to Velde ER, Kuurman WW and Dorland M. 2001. The role of genetic factors in age at natural menopause. Hum. Reprod. 16:2014-2018. Devi A and Benn PA. 1999. X-chromosome abnormalities in women with premature ovarian failure. J. Reprod. Med. 44:321-324.  44  Devi AS, Metzger DA, Luciano AA and Benn PA. 1998. 45,X/46,XX mosaicism in patients with idiopathic premature ovarian failure. Fertil. Steril. 70:89-93. Di Pasquale E, Beck-Peccoz P and Persani L. 2004. Hypergonadotropic ovarian failure associated with an inherited mutation of human bone morphogenetic protein-15 (BMP15) gene. Am. J. Hum. Genet. 75:106-111. Di Pasquale E, Rossetti R, Marozzi A, Bodega B, Borgato S, Cavallo L, Einaudi S, Radetti G, Russo G, Sacco M, et al. 2006. Identification of new variants of human BMP 15 gene in a large cohort of women with premature ovarian failure. J. Clin. Endocrinol. Metab. 91:1976-1979. Dixit H, Deendayal M and Singh L. 2004. Mutational analysis of the mature peptide region of inhibin genes in Indian women with ovarian failure. Hum. Reprod. 19:1760-1764. Dixit H, Rao LK, Padmalatha V, Kanakavalli M, Deenadayal M, Gupta N, Chakravarty B and Singh L. 2005. Mutational screening of the coding region of growth differentiation factor 9 gene in Indian women with ovarian failure. Menopause. 12:749-754. Dixit H, Rao LK, Padmalatha VV, Kanakavalli M, Deenadayal M, Gupta N, Chakrabarty B and Singh L. 2006. Missense mutations in the BMP 15 gene are associated with ovarian failure. Hum. Genet. 119:408-415. Doblhammer G. 2000. Reproductive history and mortality later in life: a comparative study of England and Wales and Austria. Popul. Stud. (Camb). 54:169-176. Doherty E, Pakarinen P, Tiitinen A, Kiilavuori A, Huhtaniemi I, Forrest S and Aittomaki K. 2002. A Novel mutation in the FSH receptor inhibiting signal transduction and causing primary ovarian failure. J. Clin. Endocrinol. Metab. 87:1151-1155. Dong J, Albertini DF, Nishimori K, Kumar TR, Lu N and Matzuk MM. 1996. Growth differentiation factor-9 is required during early ovarian folliculogenesis. Nature. 383:531-535. Durlinger AL, Kramer P, Karels B, de Jong FH, Uilenbroek JT, Grootegoed JA and Themmen AP. 1999. Control of primordial follicle recruitment by anti-Mullerian hormone in the mouse ovary. Endocrinology. 140:5789-5796. Elvin JA, Yan C, Wang P, Nishimori K and Matzuk MM. 1999. Molecular characterization of the follicle defects in the growth differentiation factor 9-deficient ovary. Mol. Endocrinol. 13:1018-1034. Eng C, Herman JG and Baylin SB. 2000. A bird's eye view of global methylation. Nat. Genet. 24:101-102.  45  Epel ES, Blackburn EH, Lin J, Dhabhar FS, Adler NE, Morrow JD and Cawthon RM. 2004. Accelerated telomere shortening in response to life stress. Proc. Natl. Acad. Sci. U. S. A. 101:17312-17315. Faddy MJ. 2000. Follicle dynamics during ovarian ageing. Mol. Cell. Endocrinol. 163 :43-48. Faddy MJ, Gosden RG, Gougeon A, Richardson SJ and Nelson JF. 1992. Accelerated disappearance of ovarian follicles in mid-life: implications for forecasting menopause. Hum. Reprod. 7:1342-1346. Fogli A, Gauthier-Barichard F, Schiffmann R, Vanderhoof VH, Bakalov VK, Nelson LM and Boespflug-Tanguy 0. 2004. Screening for known mutations in EIF2B genes in a large panel of patients with premature ovarian failure. BMC Womens Health. 4:8. Fogli A, Rodriguez D, Eymard-Pierre E, Bouhour F, Labauge P, Meaney BF, Zeesman S, Kaneski CR, Schiffmann R and Boespflug-Tanguy 0. 2003. Ovarian failure related to eukaryotic initiation factor 2B mutations. Am. J. Hum. Genet. 72:1544-1550. Forabosco A, Sforza C, De Pol A, Vizzotto L, Marzona L and Ferrario VF. 1991. Morphometric study of the human neonatal ovary. Anat. Rec. 231:201-208. Foster WG and Holloway AC. 2003. Do environmental contaminants adversely affect human reproductive physiology? J. Obstet. Gynaecol. Can. 25:33-44. Galloway SM, McNatty KP, Cambridge LM, Laitinen MP, Juengel JL, Jokiranta TS, McLaren RJ, Luiro K, Dodds KG, Montgomery GW, et al. 2000. Mutations in an oocyte-derived growth factor gene (BMP 15) cause increased ovulation rate and infertility in a dosage-sensitive manner Nat. Genet. 25:279-283. Golbus J, Palella TD and Richardson BC. 1990. Quantitative changes in T cell DNA methylation occur during differentiation and ageing. Eur. J. Immunol. 20:1869-1872. Gosden RG, Treloar SA, Martin NG, Cherkas LF, Spector TD, Faddy MJ and Silber SJ. 2007. Prevalence of premature ovarian failure in monozygotic and dizygotic twins. Hum. Reprod. 22:610-615. Goswami D and Conway GS. 2005. Premature ovarian failure. Hum. Reprod. Update. 11:391-410. Goswami R, Goswami D, Kabra M, Gupta N, Dubey S and Dadhwal V. 2003. Prevalence of the triple X syndrome in phenotypically normal women with premature ovarian failure and its association with autoimmune thyroid disorders. Fertil. Steril. 80:10521054. Gubbay J, Collignon J, Koopman P, Capel B, Economou A, Munsterberg A, Vivian N, Goodfellow P and Lovell-Badge R. 1990. A gene mapping to the sex-determining  46  region of the mouse Y chromosome is a member of a novel family of embryonically expressed genes. Nature. 346:245-250. Harley CB, Vaziri H, Counter CM and Allsopp RC. 1992. The telomere hypothesis of cellular aging. Exp. Gerontol. 27:375-382. Harris SE, Chand AL, Winship IM, Gersak K, Aittomaki K and Shelling AN. 2002. Identification of novel mutations in FOXL2 associated with premature ovarian failure. Mol. Hum. Reprod. 8:729-733. Hassold T, Abruzzo M, Adkins K, Griffin D, Merrill M, Millie E, Saker D, Shen J and Zaragoza M. 1996. Human aneuploidy: incidence, origin, and etiology. Environ. Mol. Mutagen. 28:167-175. Hastie ND, Dempster M, Dunlop MG, Thompson AM, Green DK and Alishire RC. 1990. Telomere reduction in human colorectal carcinoma and with ageing. Nature. 346:866-868. Helle S, Kaar P, Jokela J. 2002. Human longevity and early reproduction in pre-industrial Sami populations. J. Evol. Biol. 15:803-807. Holland CM. 2001. 47,XXX in an adolescent with premature ovarian failure and autoimmune disease. J. Pediatr. Adolesc. Gynecol. 14:77-80. Hsin H and Kenyon C. 1999. Signals from the reproductive system regulate the lifespan of C. elegans. Nature. 399:362-366. Hundscheid RD, Braat DD, Kiemeney LA, Thomas CMG and Smits AP. 2001. Increased serum FSH in female fragile X premutation carriers with either regular menstrual cycles or on oral contraceptives. Hum. Reprod. 16:457-462. Huntriss J, Hinkins M and Picton HM. 2006. cDNA cloning and expression of the human NOBOX gene in oocytes and ovarian follicles. Mol. Hum. Reprod. 12:283-289. Hutchinson EW and Rose MR. 1991. Quantitative genetics of postponed aging in Drosophila melanogaster. I. Analysis of outbred populations. Genetics. 127:719-727. Issa JP, Ottaviano YL, Celano P, Hamilton SR, Davidson NE and Baylin SB. 1994. Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat. Genet. 7:536-540. Itu M, Neelam T, Ammini AC and Kucheria K. 1990. Primary amenorrhoea in a triple X female. Aust. N. Z. J. Obstet. Gynaecol. 30:386-388. Jacobsen BK, Knutsen SF and Fraser GE. 1999. Age at natural menopause and total mortality and mortality from ischemic heart disease: the Adventist Health Study. J. Clin. Epidemiol. 52:303-307.  47  James RS, Coppin B, Dalton P, Dennis NR, Mitchell C, Sharp AJ, Skuse DH, Thomas NS and Jacobs PA. 1998. A study of females with deletions of the short arm of the X chromosome. Hum. Genet. 102:507-516. Johnson J, Canning J, Kaneko T, Pru JK and Tilly JL. 2004. Germline stem cells and follicular renewal in the postnatal mammalian ovary. Nature. 428:145-150. Kalantaridou SN and Chrousos GP. 2000. Molecular defects causing ovarian dysfunction. Ann. N. Y. Acad. Sci. 900:40-45. Kalantaridou SN and Nelson LM. 2000. Premature ovarian failure is not premature menopause. Ann. N. Y. Acad. Sci. 900:393-402. Katsuya T, Horiuchi M, Minami S, Koike G, Santoro NF, Hsueh AJ and Dzau VJ. 1997. Genomic organization and polymorphism of human angiotensin II type 2 receptor: no evidence for its gene mutation in two families of human premature ovarian failure syndrome. Mol. Cell. Endocrinol. 127:221-228. Kevenaar ME, Themmen AP, Rivadeneira F, Uitterlinden AG, Laven JS, van Schoor NM, Lips P, Pols HA and Visser JA. 2007. A polymorphism in the AMH type II receptor gene is associated with age at menopause in interaction with parity. Hum. Reprod. 22:2382-2388. Kinney A, Kline J, Kelly A, Reuss ML and Levin B. 2007. Smoking, alcohol and caffeine in relation to ovarian age during the reproductive years. Hum. Reprod. 22:1175-1185. Kinugawa C, Murakami T, Okamura K and Yajima A. 2000. Telomerase activity in normal ovaries and premature ovarian failure. Tohoku J. Exp. Med. 190:231-238. Kosaki K, Sato S, Hasegawa T, Matsuo N, Suzuki T and Ogata T. 2004. Premature ovarian failure in a female with proximal symphalangism and Noggin mutation. Fertil. Steril. 81:1137-1139. Kovanci E, Rohozinski J, Simpson JL, Heard MJ, Bishop CE and Carson SA. 2007. Growth differentiating factor-9 mutations may be associated with premature ovarian failure. Fertil. Steril. 87:143-146. Kuningas M, Magi R, Westendorp RG, Slagboom PE, Remm M and van Heemst D. 2007. Haplotypes in the human Foxola and Foxo3a genes; impact on disease and mortality at old age. Eur. J. Hum. Genet. 15:294-301. Laissue P, Christin-Maitre S, Touraine P, Kuttenn F, Ritvos 0, Aittomaki K, Bourcigaux N, Jacquesson L, Bouchard P, Frydman R, et al. 2006. Mutations and sequence variants in GDF9 and BMP 15 in patients with premature ovarian failure. Eur. J. Endocrinol. 154:739-744.  48  Laml T, Preyer 0, Umek W, Hengstschlager M and Hanzal H. 2002. Genetic disorders in premature ovarian failure. Hum. Reprod. Update. 8:483-491. Layman LC, Amde S, Cohen DP, Jin M and Xie J. 1998. The Finnish follicle-stimulating hormone receptor gene mutation is rare in North American women with 46,XX ovarian failure. Fertil. Steril. 69:300-302. Layman LC, Shelley ME, Huey LO, Wall SW, Tho SP and McDonough PG. 1993. Follicle-stimulating hormone beta gene structure in premature ovarian failure. Fertil. Steril. 60:852-857. Lee JT, Davidow LS and Warshawsky D. 1999. Tsix, a gene antisense to Xist at the Xinactivation centre. Nat. Genet. 21:400-404. Lorda-Sanchez IJ, Ibanez AJ, Sanz RJ, Trujillo MJ, Anabitarte ME, Querejeta ME, Rodriguez de Alba M, Gimenez A, Infantes F, Ramos C, et al. 2000. Choroideremia, sensorineural deafness, and primary ovarian failure in a woman with a balanced X-4 translocation. Ophthalmic Genet. 21:185-189. Luborsky JL, Meyer P, Sowers MF, Gold EB and Santoro N. 2003. Premature menopause in a multi-ethnic population study of the menopause transition. Hum. Reprod. 18:199-206. Luoh SW, Bain PA, Polakiewicz RD, Goodheart ML, Gardner H, Jaenisch R and Page DC. 1997. za mutation results in small animal size and reduced germ cell number in male and female mice. Development. 124:2275-2284. Luoma P, Melberg A, Rinne JO, Kaukonen JA, Nupponen NN, Chalmers RM, Oldfors A, Rautakorpi I, Peltonen L, Majamaa K, et al. 2004. Parkinsonism, premature menopause, and mitochondrial DNA polymerase gamma mutations: clinical and molecular genetic study. Lancet. 364:875-882. Lyon MF. 1998. X-chromosome inactivation spreads itself: effects in autosomes. Am. J. Hum. Genet. 63:17-19. Lyon MF. 1961. Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature. 190:372-373. Manor 0, Eisenbach Z, Israeli A and Friedlander Y. 2000. Mortality differentials among women: the Israel Longitudinal Mortality Study. Soc. Sci. Med. 51:1175-1188. Marozzi A, Manfredini E, Tibiletti MG, Furlan D, Villa N, Vegetti W, Crosignani PG, Ginelli E, Meneveri R and Dalpra L. 2000. Molecular definition of Xq commondeleted region in patients affected by premature ovarian failure. Hum. Genet. 107:304-311.  49  Marozzi A, Porta C, Vegetti W, Crosignani PG, Tibiletti MG, Dalpra L and Ginelli E. 2002. Mutation analysis of the inhibin alpha gene in a cohort of Italian women affected by ovarian failure. Hum. Reprod. 17:1741-1745. Matthews CH, Borgato S, Beck-Peccoz P, Adams M, Tone Y, Gambino G, Casagrande S, Tedeschini G, Benedetti A and Chatterjee VK. 1993. Primary amenorrhoea and infertility due to a mutation in the beta-subunit of follicle-stimulating hormone. Nat. Genet. 5:83-86. McGee EA and Hsueh AJ. 2000. Initial and cyclic recruitment of ovarian follicles. Endocr. Rev. 21:200-214. McGue M, Vaupel JW, Holm N and Harvald B. 1993. Longevity is moderately heritable in a sample of Danish twins born 1870-1880. J. Gerontol. 48:B237-44. Meduri G, Touraine P, Beau I, Lahuna 0, Desroches A, Vacher-Lavenu MC, Kuttenn F and Misrahi M. 2003. Delayed puberty and primary amenorrhea associated with a novel mutation of the human follicle-stimulating hormone receptor: clinical, histological, and molecular studies. J. Clin. Endocrinol. Metab. 88:3491-3498. Menon V, Edwards RL, Butt WR, Bluck M and Lynch SS. 1984. Review of 59 patients with hypergonadotrophic amenorrhoea. Br. J. Obstet. Gynaecol. 91:63-66. Morishima A, Grumbach MM, Simpson ER, Fisher C and Qin K. 1995. Aromatase deficiency in male and female siblings caused by a novel mutation and the physiological role of estrogens. J. Clin. Endocrinol. Metab. 80:3689-3698. Muller HG, Chiou JM, Carey JR and Wang JL. 2002. Fertility and life span: late children enhance female longevity. J. Gerontol. A Biol. Sci. Med. Sci. 57:B202-6. Mumm S, Herrera L, Waeltz PW, Scardovi A, Nagaraja R, Esposito T, Schlessinger D, Rocchi M and Forabosco A. 2001. X/autosomal translocations in the Xq critical region associated with premature ovarian failure fall within and outside genes. Genomics. 76:30-36. Muntoni A and Reddel RR. 2005. The first molecular details of ALT in human tumor cells. Hum. Mol. Genet. 14 Spec No. 2:R191-6. Murray A, Ennis S, MacSwiney F, Webb J and Morton NE. 2000. Reproductive and menstrual history of females with fragile X expansions. Eur. J. Hum. Genet. 8:247252. Murray A, Webb J, Dennis N, Conway G and Morton N. 1999. Microdeletions in FMR2 may be a significant cause of premature ovarian failure. J. Med. Genet. 36:767-770. Murray A, Webb J, Grimley S, Conway G and Jacobs P. 1998. Studies of FRAXA and FRAXE in women with premature ovarian failure. J. Med. Genet. 35:637-640.  50  Nagai J, Lin CY and Sabour MP. 1995. Lines of mice selected for reproductive longevity. Growth Dev. Aging. 59:79-91. Nagy B, Ban Z, Toth-Pal E, Papp C, Fintor L and Papp Z. 2000. Apolipoprotein E allele distribution in trisomy 13, 18, and 21 conceptuses in a Hungarian population. Am. J. Clin. Pathol. 113:535-538. Nelson LM, Anasti JN, Kimzey LM, Defensor RA, Lipetz KJ, White BJ, Shawker TH and Merino MJ. 1994. Development of luteinized Graafian follicles in patients with karyotypically normal spontaneous premature ovarian failure. J. Clin. Endocrinol. Metab. 79:1470-1475. Nelson LM, Covington SN and Rebar RW. 2005. An update: spontaneous premature ovarian failure is not an early menopause. Fertil. Steril. 83:1327-1332. Nikolaou D and Templeton A. 2004. Early ovarian ageing. Eur. J. Obstet. Gynecol. Reprod. Biol. 113:126-133. Nippita TA and Baber RJ. 2007. Premature ovarian failure: a review. Climacteric. 10:1122. Nordfjall K, Larefalk A, Lindgren P, Holmberg D and Roos G. 2005. Telomere length and heredity: Indications of paternal inheritance. Proc. Natl. Acad. Sci. U. S. A. 102:16374-16378. Ogata T and Matsuo N. 1995. Turner syndrome and female sex chromosome aberrations: deduction of the principal factors involved in the development of clinical features. Hum. Genet. 95:607-629. Otsuka F, Yao Z, Lee T, Yamamoto S, Erickson GF and Shimasaki S. 2000. Bone morphogenetic protein-15. Identification of target cells and biological functions. J. Biol. Chem. 275:39523-39528. Pagnamenta AT, Taanman JW, Wilson CJ, Anderson NE, Marotta R, Duncan AJ, BitnerGlindzicz M, Taylor RW, Laskowski A, Thorburn DR, et al. 2006. Dominant inheritance of premature ovarian failure associated with mutant mitochondrial DNA polymerase gamma. Hum. Reprod. 21:2467-2473. Partington MW, Moore DY and Turner GM. 1996. Confirmation of early menopause in fragile X carriers. Am. J. Med. Genet. 64:370-372. Perez MS, Benencia H, Frechtel GD, Esteban EO, Gil MC, Targovnik HM and Marquez NB. 2004. 17alpha-Hydroxylase Deficiency : Biochemical and Molecular Findings in Two Sisters and their Family. Mol. Diagn. 8:171-178. Perheentupa J. 1996. Autoimmune polyendocrinopathy--candidiasis--ectodermal dystrophy (APECED). Horm. Metab. Res. 28:353-356.  51  Pens T, Levenson R, Regan M and Puca A. 2002. What does it take to live to 100? Mech. Ageing Dev. 123:231-242. Perls TT, Alpert L and Fretts RC. 1997. Middle-aged mothers live longer. Nature. 389:133. Pens TT and Fretts RC. 2001. The evolution of menopause and human life span. Ann. Hum. Biol. 28:237-245. Practice Committee of the American Society for Reproductive Medicine. 2006. Current evaluation of amenorrhea. Fertil. Steril. 86:S148-55. Prueitt RL, Ross JL and Zinn AR. 2000. Physical mapping of nine Xq translocation breakpoints and identification of XPNPEP2 as a premature ovarian failure candidate gene. Cytogenet. Cell Genet. 89:44-50. Prueitt RL, Chen H, Barnes RI and Zinn AR. 2002. Most X;autosome translocations associated with premature ovarian failure do not interrupt X-linked genes. Cytogenet. Genome Res. 97:32-38. Qin Y, Choi Y, Zhao H, Simpson JL, Chen Z and Rajkovic A. 2007. NOBOX homeobox mutation causes premature ovarian failure. Am. J. Hum. Genet. 81:000-000. Rajkovic A, Pangas SA, Ballow D, Suzumori N and Matzuk MM. 2004. NOBOX deficiency disrupts early folliculogenesis and oocyte-specific gene expression. Science. 305:1157-1159. Rebar RW and Connolly HV. 1990. Clinical features of young women with hypergonadotropic amenorrhea. Fertil. Steril. 53:804-810. Richardson B. 2003. Impact of aging on DNA methylation. Ageing Res. Rev. 2:245-261. Rizzolio F, Bione S, Sala C, Goegan M, Gentile M, Gregato G, Rossi E, Pramparo T, Zuffardi 0 and Toniolo D. 2006. Chromosomal rearrangements in Xq and premature ovarian failure: mapping of 25 new cases and review of the literature. Hum. Reprod. 21:1477-1483. Roberts RG, Kendall E, Vetrie D and Bobrow M. 1996. Sequence and chromosomal location of a human homologue of LRPR1, an FSH primary response gene. Genomics. 37:122-124. Saldanha SN, Andrews LG and Tollefsbol TO. 2003. Assessment of telomere length and factors that contribute to its stability. Eur. J. Biochem. 270:389-403. Schlessinger D, Herrera L, Crisponi L, Mumm S, Percesepe A, Pellegrini M, Pilia G and Forabosco A. 2002. Genes and translocations involved in POF. Am. J. Med. Genet. 111:328-333.  52  Schwartz CE, Dean J, Howard-Peebles PN, Bugge M, Mikkelsen M, Tommerup N, Hull C, Hagerman R, Holden JJ and Stevenson RE. 1994. Obstetrical and gynecological complications in fragile X carriers: a multicenter study. Am. J. Med. Genet. 51:400402. Sharara FI, Seifer DB and Flaws JA. 1998. Environmental toxicants and female reproduction. Fertil. Steril. 70:613-622. Shelling AN, Burton KA, Chand AL, van Ee CC, France JT, Farquhar CM, Milsom SR, Love DR, Gersak K, Aittomaki K, et al. 2000. Inhibin: a candidate gene for premature ovarian failure. Hum. Reprod. 15:2644-2649. Sherman SL. 2000. Premature ovarian failure in the fragile X syndrome. Am. J. Med. Genet. 97:189-194. Simpson JL and Rajkovic A. 1999. Ovarian differentiation and gonadal failure. Am. J. Med. Genet. 89:186-200. Slagboom PE, Droog S and Boomsma DI. 1994. Genetic determination of telomere size in humans: a twin study of three age groups. Am. J. Hum. Genet. 55:876-882. Smith KR, Mineau GP and Bean LL. 2002. Fertility and post-reproductive longevity. Soc. Biol. 49:185-205. Snieder H, MacGregor AJ and Spector TD. 1998. Genes control the cessation of a woman's reproductive life: a twin study of hysterectomy and age at menopause. J. Clin. Endocrinol. Metab. 83:1875-1880. Snowdon DA, Kane RL, Beeson WL, Burke GL, Sprafka JM, Potter J, Iso H, Jacobs DR,Jr and Phillips RL. 1989. Is early natural menopause a biologic marker of health and aging? Am. J. Public Health. 79:709-714. Soyal SM, Amleh A and Dean J. 2000. FIGalpha, a germ cell-specific transcription factor required for ovarian follicle formation. Development. 127:4645-4654. Sundblad V, Chiauzzi VA, Escobar ME, Damn L and Charreau EH. 2004. Screening of FSH receptor gene in Argentine women with premature ovarian failure (POF). Mol. Cell. Endocrinol. 222:53-59. Takahashi K, Karino K, Kanasaki H and Miyazaki K. 2004. Altered kinetics of pituitary response to gonadotropin-releasing hormone in women with variant luteinizing hormone: correlation with ovulatory disorders. Horm. Res. 61:27-32. Takahashi K, Ozaki T, Okada M, Kurioka H, Kanasaki H and Miyazaki K. 1999. Increased prevalence of luteinizing hormone beta-subunit variant in patients with premature ovarian failure. Fertil. Steril. 71:96-101.  53  Takakura K, Takebayashi K, Wang HQ, Kimura F, Kasahara K and Noda Y. 2001. Follicle-stimulating hormone receptor gene mutations are rare in Japanese women with premature ovarian failure and polycystic ovary syndrome. Fertil. Steril. 75:207209. Takebayashi K, Takakura K, Wang H, Kimura F, Kasahara K and Noda Y. 2000. Mutation analysis of the growth differentiation factor-9 and -9B genes in patients with premature ovarian failure and polycystic ovary syndrome. Fertil. Steril. 74:976979. Tanaka M, Ohnishi J, Ozawa Y, Sugimoto M, Usuki S, Naruse M, Murakami K and Miyazaki H. 1995. Characterization of angiotensin II receptor type 2 during differentiation and apoptosis of rat ovarian cultured granulosa cells. Biochem. Biophys. Res. Commun. 207:593-598. te Velde ER and Pearson PL. 2002. The variability of female reproductive ageing. Hum. Reprod. Update. 8:141-154. Toniolo D. 2006. X-linked premature ovarian failure: a complex disease. Curr. Opin. Genet. Dev. 16:293-300. Toniolo D and Rizzolio F. 2007. X chromosome and ovarian failure. Semin. Reprod. Med. 25:264-271. Tonon L, Bergamaschi G, Dellavecchia C, Rosti V, Lucotti C, Malabarba L, Novella A, Vercesi E, Frassoni F and Cazzola M. 1998. Unbalanced X-chromosome inactivation in haemopoietic cells from normal women. Br. J. Haematol. 102:996-1003. Touraine P, Beau I, Gougeon A, Meduri G, Desroches A, Pichard C, Detoeuf M, Paniel B, Prieur M, Zorn JR, et al. 1999. New natural inactivating mutations of the folliclestimulating hormone receptor: correlations between receptor function and phenotype. Mol. Endocrinol. 13:1844-1854. Uda M, Ottolenghi C, Crisponi L, Garcia JE, Deiana M, Kimber W, Forabosco A, Cao A, Schlessinger D and Pilia G. 2004. Fox12 disruption causes mouse ovarian failure by pervasive blockage of follicle development. Hum. Mol. Genet. 13:1171-1181. Unryn BM, Cook LS and Riabowol KT. 2005. Paternal age is positively linked to telomere length of children. Aging Cell. 4:97-101. Uzumcu M and Zachow R. 2007. Developmental exposure to environmental endocrine disruptors: consequences within the ovary and on female reproductive function. Reprod. Toxicol. 23:337-352. van Asselt KM, Kok HS, Putter H, Wijmenga C, Peeters PH, van der Schouw YT, Grobbee DE, te Velde ER, Mosselman S and Pearson PL. 2004. Linkage analysis of extremely discordant and concordant sibling pairs identifies quantitative trait loci influencing variation in human menopausal age. Am. J. Hum. Genet. 74:444-453.  54  van Kasteren YM, Hundscheid RD, Smits AP, Cremers FP, van Zonneveld P and Braat DD. 1999. Familial idiopathic premature ovarian failure: an overrated and underestimated genetic disease? Hum. Reprod. 14:2455-2459. Vegetti W, Grazia Tibiletti M, Testa G, de Lauretis Y, Alagna F, Castoldi E, Taborelli M, Motta T, Bolis PF, Dalpra L, et al. 1998. Inheritance in idiopathic premature ovarian failure: analysis of 71 cases. Hum. Reprod. 13:1796-1800. Vegetti W, Marozzi A, Manfredini E, Testa G, Alagna F, Nicolosi A, Caliari I, Taborelli M, Tibiletti MG, Dalpra L, et al. 2000. Premature ovarian failure. Mol. Cell. Endocrinol. 161:53-57. Vital-Reyes V, Chhieng D, Rodriguez-Burford C, Tellez-Velasco S, Grizzle W, Chavarria-Olarte ME and Reyes-Fuentes A. 2006. Ovarian biopsy in infertile patients with ovarian dysfunction. Int. J. Gynecol. Pathol. 25:90-94. Wang HQ, 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. 78:13291330. Watkins WJ, Harris SE, Craven MJ, Vincent AL, Winship IM, Gersak K and Shelling AN. 2006a. An investigation into FOXE1 polyalanine tract length in premature ovarian failure. Mol. Hum. Reprod. 12:145-149. Watkins WJ, Umbers AJ, Woad KJ, Harris SE, Winship IM, Gersak K and Shelling AN. 2006b. Mutational screening of FOXO3A and FOXO1A in women with premature ovarian failure. Fertil. Steril. 86:1518-1521. Willard HF. 1996. X chromosome inactivation, XIST, and pursuit of the X-inactivation center. Cell. 86:5-7. Zhang Z, Deng C, Lu Q and Richardson B. 2002. Age-dependent DNA methylation changes in the ITGAL (CD1 l a) promoter. Mech. Ageing Dev. 123:1257-1268. Zhao H, Chen ZJ, Qin Y, Wang S, Simpson JL and Rajkovic A 2007. FIGLA mutations cause premature ovarian failure in a subset of Chinese women with POF. Presented at the annual meeting of The American Society of Human Genetics, October 26, 2007, San Diego, California Available from http://www.ashg.orgigenetics/ashg07s/index.shtml Zlotogora J, Sagi M and Cohen T. 1983. The blepharophimosis, ptosis, and epicanthus inversus syndrome: delineation of two types. Am. J. Hum. Genet. 35:1020-1027.  55  Chapter 2: FMR1 and premature ovarian failure' 2.1 Introduction  The most significant single gene association with premature ovarian failure is with the Fragile X mental retardation 1 gene, FMR1, at Xq27.3. Fragile X mental retardation is the most common cause of mental retardation in males and is caused by the expansion of a polymorphic CGG repeat in the 5'untranslated region of the FMRI gene (Oberle et al. 1991). FMRI allele sizes of greater than 200 CGG repeats are considered full mutations and are subject to methylation-induced silencing of the FMRI gene. Since males carry only a single X chromosome, male carriers of a full mutation have an absence of the FMRI protein, FMRP. They present with the mild to severe mental retardation and specific constellation of morphological features of Fragile X syndrome. Females have two copies of the X chromosome; the expression of genes on these is influenced by the degree and direction of X chromosome inactivation skewing. Therefore, female carriers of a full mutation present with Fragile X syndrome at a much lower frequency and severity than males. FMRI allele sizes of —55-200 repeats are prone to expansion in subsequent  meiosis and mitosis and are known as premutation alleles. In the last 20 years specific phenotypes associated with premutation alleles have been reported. In particular, elderly male carriers of premutation alleles are at high risk of specific neurological condition known as Fragile X Tremor Ataxia Syndrome (FXTAS) (Hagerman and Hagerman. 'A version of this chapter has been published: Bretherick KL, Fluker MR, Robinson WP (2005). FMRI repeat sizes in the grey zone and high end of the normal range are associated with premature ovarian failure. Hum Genet 117(4):276-282. Springer-Verlag 2005, used with the kind permission of Springer Science+Business Media. This chapter also includes expression and methylation results that were not included in this publication.  56  2002). In addition, female carriers of premutation size alleles are at risk of POF. Studies examining POF incidence in families ascertained on the basis of Fragile X syndrome (Table 2.1), report that 12-19% of premutation carriers experience POF (reviewed in Sherman et al. 2007). Overall there is a shift towards an earlier menopause in premutation carriers (Murray et al. 2000; Partington et al. 1996; Sullivan et al. 2005), and an increase in markers of early ovarian aging such as abnormal hormone regulation, altered cycle lengths and increased twinning (Allen et al. 2007; Hundscheid et al. 2001; Murray et al. 1999; Vianna-Morgante. 1999; Welt et al. 2004). As would be expected, premutation size alleles are also found at increased frequency in patients ascertained on the basis of POF (Table 2.2). Specifically, premutations are found in 1-6% of individuals with sporadic POF, and 5-20% of familial POF cases (Sherman et al. 2007), but have a prevalence of only 0.1-0.3% in control women (Crawford et al. 1999; Rousseau et al. 1995). FMR1 premutations therefore account for a small but significant proportion of POF cases. The mechanism by which FMRI premutation size alleles cause POF is unknown. Several population studies have shown that both male and female carriers of premutation size alleles have an increase in FMR1 transcript level compared to individuals with allele sizes in the normal range (Kenneson et al. 2001; Tassone et al. 2000a; Tassone et al. 2000b). This increase is accompanied by a decrease in FMRP level (Kenneson et al. 2001; Tassone et al. 2000c) due to decreased translation efficiency of the longer mRNA transcript (Chen et al. 2003; Primerano et al. 2002). It is unlikely that this decrease in FMRP is responsible for POF in premutation carriers, since full mutation carriers who have an absence of FMRP from their expanded allele, are not at increased risk for POF.  57  Therefore the increase in FMRI mRNA level seen in association with premutation alleles may be the pathogenic factor in POF. Between the classic normal (6-45 repeats) and premutation ranges (-55-200 repeats), lies an overlapping and poorly defined class of intermediate or "grey zone" alleles of —45-60 repeats in which the repeat is less stable and the risk for expansion remains unclear (Nolin et al. 2003). The risk of POF for women in this range has not been established and analysis of the available literature is complicated by the everchanging definition of the lower boundary of the premutation class. In addition, most studies have examined the prevalence of POF among FMRI premutation carriers rather than FMRI repeat size in women with POF, therefore the risk of POF for women with grey zone alleles is not known. Notably, alleles within this grey zone may be associated with increased levels of FMRI mRNA. A linear increase in FMRI transcript level has been reported for male carriers of grey zone alleles (Loesch et al. 2007) although the relationship is less clear in female carriers as a result of X chromosome inactivation skewing (Garcia-Alegria et al. 2007). When the effect of repeat length on mRNA and protein levels is examined at a molecular level, it is apparent that even allele sizes as small 43 repeats can cause an increase in mRNA levels (Chen et al. 2003). It is possible that the increase in transcript level occurs in alleles within the grey zone. Generally, DNA methylation at CpG sites in the promoter region of a gene is correlated with gene silencing, and conversely, a lack of DNA methylation in this region is associated with gene expression. Specifically, methylation at CpG sites in the FMRI promoter region (Figure 2.2) is associated with an absence of gene expression (Pieretti et al. 1991). Therefore, an alternative means of FMRI involvement in POF may be a  58  decrease in either global or local DNA methylation on the inactive X-chromosome resulting in increased FMRI mRNA levels. If FMK/ alleles at the high end of the normal repeat range are associated with increased FMRI transcript levels, carriers of these alleles may also be at increased risk of POF. Furthermore, if female carriers of normal FMRI repeat lengths have DNA hypomethylation at the promoter region of FMRI on the inactive X-chromosome, they may also have elevated FMRI mRNA levels and be at increased risk for POF. In this study I examined FMRI repeat length frequencies in women ascertained with POF and in controls. I assessed not only FMRI allele size, but also FMRI biallelic mean, and FMRI repeat size adjusted for X-chromosome inactivation (XCI) ratio, as this measure may more accurately reflect physiologic mRNA levels. I also looked at transcript level both indirectly, by comparing DNA methylation in POF patients and controls at two sites in the FMRI promoter, and directly, by assaying FMR1 mRNA level in peripheral blood leukocytes in a subset of POF patients and controls.  2.2 Methods 2.2.1 Samples The collection of samples for this study was approved by the University of British Columbia Clinical Research Ethics Board (CREB), approval number CO1-0460 (Appendix 1). Women with idiopathic POF presenting with secondary amenorrhea (N=57) were recruited from the POF clinic at BC Women's Hospital. POF diagnosis was made based on the absence of menses for at least three months and two serum follicle stimulating hormone (FSH) results of > 40mIU/mL obtained more than one month apart, prior to age 40. Genomic DNA was extracted from 5-7mL of EDTA anti-coagulated 59  blood using standard salt extraction. One woman was found to have extremely biased amplification of alleles at the FMR1 and AR loci, as well as other X chromosome markers tested. Karyotype analysis confirmed high levels of 45,X mosaicism and this sample was therefore excluded from the analysis as her POF is likely attributed to mosaic Turner syndrome. All of the other 56 women were found to be karyotypically normal on routine diagnostic analysis. Two control groups were used in this study. Control group 1 consisted of anonymous healthy donors (n = 24) and the unaffected female spouses of families with an autosomal dominant genetic disorder (n = 140). Although the reproductive status of these women is not known, this group should be unbiased in terms of pregnancy history and should represent a cross section of the population that the POF patients were drawn from. Control Group 2 (n = 50) consisted of women who had had a healthy pregnancy after the age of 37 years and had not experienced a pregnancy loss. Genomic DNA was extracted from mouthwash rather than EDTA peripheral blood, for two of the samples from Control Group 2. As mouthwash samples may contain different proportions and types of cells and different cell types would be expected to have different epigenetic and expression patterns, these samples were excluded from analysis for FMR1 methylation and expression. 2.2.2 FMR1 repeat length FMRI repeat length was determined in patients and controls by PCR  amplification (Hecimovic et al. 1997) and fragment analysis on an ABI 310 genetic analyzer. Approximately 50 ng of genomic DNA was amplified in a 10 pit PCR reaction containing: lx Buffer 1 (Expand Long Template PCR System, Roche), 10% DMSO, 350 60  liM dATP, 350 JAM dCTP, 350 p,M dTTP, 1001.1M dGTP, 250 RM 7-deaza-dGTP, 1pM of each primer (FMR1-F 5'-6-FAM-GCTCAGCTCCGTTTCGGTTTCACTTCCGGT-3' and FMR1-R 5'-AGCCCCGCACTTCCACCACCAGCTCCTCCA-3') and 0.35 Units Taq (Expand Long Template PCR System, Roche). Cycling conditions were 97°C for 30 seconds 62°C for 30 seconds and 68°C for 2 minutes, for a total of 40 cycles. 11.IL of each PCR reaction was run with 0.1 µL ROX 500 size standard on the ABI 310 Prism genetic analyzer. Two homozygous control samples were sequenced to determine exact FMR1 repeat number and these were run after every 10 samples on the ABI 310 to ensure an accurate conversion of amplicon size to repeat number. Specifically, the known repeat length for each control was multiplied by three (as this is a triplet repeat) and subtracted from the average amplicon size of that control in each PCR batch. An average of the values for each of the controls was used as a "correction factor" to convert amplicon size to repeat number for the rest of the samples in the batch. FMR1 repeat sizes larger than 80 repeats could not be amplified using our PCR protocol, therefore all samples showing amplification of only one allele by PCR were examined by Southern blot to test for the presence of larger premutation and full mutation size alleles. Southern blot was performed using the Chemiluminescent Southern Blot protocol described by Gold et al (Gold et al. 2000). For each sample amplifying only a single band on PCR, 4 lig genomic DNA was digested by overnight incubation at 37°C with 50 U EcoRI and 50 U EagI in a 20uL reaction. Digested DNA was run on a 0.8% agarose gel, at low voltage overnight until a 2.3 kb marker had migrated —16cm. DNA in agarose gels was then blotted onto a Hybond N+ membrane by overnight capillary transfer and UV crosslinking. Blots were probed with the StB12.3 probe that was DIG-  61  labeled by PCR, and developed using reagents and protocol from the DIG DNA Detection System (Roche). Developed blots were visualized and photographed using either a Lumi-Imager and Quantity One software (Roche) or a Chemigenius 2 and GeneSnap software (Syngene). An example of Southern blot results obtained by this method is shown in Figure 2.3. One sample in control group 1 could not be tested by Southern blot analysis because of insufficient quantity of DNA, this sample had an allele of 30 repeats detected by PCR and was excluded from our analysis because the presence or absence of a premutation or full mutation allele could not be verified.  2.2.3 X chromosome inactivation skewing X chromosome inactivation ratio was determined by methylation-sensitive restriction enzyme digestion of genomic DNA followed by PCR amplification and product quantification on an ABI 310 Genetic analyzer (Allen et al. 1992). The protocol was carried out as described previously (Beever et al. 2003). Briefly, 150 ng of genomic DNA was digested with 3 units of HpaII (which will only cut unmethylated sites) and 1 unit of RsaI (a secondary cutter, shown to improve skewing reproducibility) at 37°C for at least 16 hrs. An undigested sample containing identical components but lacking the HpaII cutter was prepared and run in parallel with the digested sample. Completeness of digestion was determined by PCR amplification of the 5' region of the MIC2 gene, a gene on the X chromosome which escapes X inactivation and is therefore never methylated at the 5' region. MIC2 PCR products were separated on a 1% agarose gel containing ethidium bromide, by electrophoresis at 125V for 45 minutes. Gels were visualized by UV light and absence of any visible PCR product was taken as evidence of complete digestion. Digested and undigested samples were then PCR amplified at the FMR] locus  62  using the PCR conditions described above, or at the Androgen Receptor (AR) locus. PCR conditions for the AR locus were as follows: 101.1,L reactions were carried out with lx PCR buffer (ROSE Scientific), 40011M each dNTP, 500 nM each primer (AR-F 5'-HEXGCT GTG AAG GTT GCT GTT CCT CAT-3' and AR-R 5'-TCC ACA ATC TGT TCC AGA GCG TGC-3'), 0.15 U Taq (ROSE Scientific) amplified with an initial denaturation at 95 for 5 minutes, followed by 30 cycles of 95 for 45 seconds, 60 for 30 seconds. Products were run on an ABI 310 Prism genetic analyzer and the peak area for each allele was analyzed by the use of Gene Scan software. The degree of X inactivation skewing was calculated at both the AR and FMR1 loci by normalization to the undigested sample, as described previously (Beever et al. 2003). To correlate X inactivation with relative activity of the two alleles, skewing ratios are expressed for the AR and FMR1 loci in relation to the allele with the higher repeat number. Therefore an FMR1 X chromosome inactivation skewing ratio of <50% suggests a relative activity of the larger FMR1 repeat that is <50%, indicating that the smaller repeat allele was predominately active in that individual.  2.2.4 FMR1 genotype repeat size Since the choice of which X chromosome to be inactivated in any given cell is random, the ratio of inactivation of the two X chromosomes will usually approximate 50:50. Significant variation from this ratio will affect the relative activity of the two X chromosomes and in cases where the two chromosomes differ in FMR1 repeat size, the relative expression of the two FMR1 alleles. Therefore FMR1 allele size adjusted for Xinactivation ratio may more accurately reflect physiologic mRNA levels which may be the mechanism of the FMR1 association to POF.  63  Using the X chromosome inactivation ratios, a "genotype repeat size" (Allen et al. 2004) was determined for each individual. Because of a weak correlation between skewing values measured at AR and FMRI, genotype repeat size was calculated using both AR and FMRI skewing values separately. Although the AR locus is likely a more reliable assay of true XCI skewing, the skewing values assayed at the FMRI locus may better reflect the relative activity of the two FMR1 alleles. The X chromosome inactivation ratios measured at the FMRI locus were used to determine which FMRI allele was inactivated more often and the AR skewing values were converted to reflect this. Thus all X chromosome inactivation ratios (measured at AR and FMRI) were expressed in terms of the percent inactivation of the smaller FMRI allele. Genotype repeat size reflecting the relative activity of both alleles was therefore calculated as follows: Genotype repeat size = (Larger FMRI allele*XCI ratio) + (Smaller FMRI allele*(1-XCI ratio)), rounded to the nearest whole number. For samples that were uninformative at the FMRI locus, since both alleles were approximately the same size, genotype repeat size was calculated with an XCI ratio of 50% for both the AR and FMRI genotype repeat size analysis. Samples that were uninformative at the AR locus, but informative at FMRI were omitted from the analysis of genotype repeat size calculated with AR. Since XCI skewing is age dependant (Hatakeyama et al. 2004), skewing and genotype repeat size was only assessed in those women in control group 1 that were within reproductive age (17-45yrs) (N=105), and was not assessed for women in control group 2.  64  2.2.5 DNA methylation  Degree of DNA methylation was assessed at two CpG sites upstream of the FMR1 gene using a methylation sensitive Single Nucleotide Primer Extension (ms-  SNuPE) assay (Gonzalgo and Jones. 2002) (Figure 2.1). CpG residues C5 and C6 (Figure 2.2) were chosen for analysis since sequence surrounding these CpGs allowed for optimal primer design and methylation of these sites is strongly correlated with methylation across the FMR1 CpG island (Stoger et al. 1997). For each sample, 100ng of genomic DNA in 20uL volume was bisulfite converted with the EZ DNA methylation-GOLD kit (Zymo Research). Bisulfite converted DNA was then PCR amplified for the region of the FMRI gene surrounding the C5 and C6 CpG sites. PCR reactions consisted of lx PCR GOLD buffer (Applied Biosystems), 0.2mM each dNTP, 2mM MgC12, 0.4uM each of m13 labeled primers (Boyd et al. 2006) FMR1-MS F ( 5'-TGT AAA ACG ACG GCC AGT TGA GTG TAT TTT TGT AGA AAT GGG-3') and FMR1-MS R (5'-GCA GGA ACC AGC TAT GAC CTC TCT CTC TTC AAA TAA CCT AAA AAC-3'), 1 U AmpliTaq GOLD (Applied Biosystems) and 0.5uL of bisulfite converted DNA, amplified in a final volume of 20uL. Cycling conditions were an initial 5 minute denaturation at 95°C, 5 cycles of 95°C for 30 seconds, 60°C for 2 minutes, and 72°C for 3 minutes followed by 30 cycles of 95°C for 30 seconds, 65°C for 1 minute, and 72°C for 3 minutes and finished with a 60°C extension for 1 hour and 25 minutes. To ensure complete removal of unlabelled dNTPs and primers, PCR products were digested with 1.3 U Exol and 3.3 U shrimp alkaline phosphatase at 37 °C for 1 hour, and then cleaned with the DNA Clean and Concentrator5 kit (Zymo Research) and eluted in lOuL of water.  65  Since the primers did not easily multiplex, separate SNuPE assays were performed for CpG sites C5 and C6. For each reaction 0.4uM degenerate primer C5 (5'GAG GTA GTG C/TGA TTT GTT AT-3') or C6 (5'-GAT TTG TTA AT ch-GTT TTT TAG TTT TTT-3'), and 5uL of SNAPshot Multiplex Ready Reaction Mix (Applied Biosystems) were combined with 3uL of cleaned PCR product in a lOuL reaction volume. Extension was accomplished with 25 cycles of 96°C for 10 seconds, 50°C for 5 seconds and 60°C for 30 seconds. To completely remove labeled dNTPs that had not been incorporated, SNAPshot products were incubated at 37°C for 1 hour with lx dephosphorylation buffer and 1U calf intestinal alkaline phosphatase (Invitrogen) in a final reaction volume of 15uL. Following this post extension treatment, 1 uL of each SNuPE product was diluted in lOuL HiDi formamide (Applied Biosystems) and run without a size standard on an ABI 310 Prism genetic analyzer for 15 minutes with POP-4 polymer. The amount of unmethylated, converted PCR product and methylated, unconverted PCR product can be visualized by the area of the peaks adding either a dROX labeled thymine or a dTAMRA labeled cytosine, respectively. Peak areas were quantified by Gene Scan software (Applied Biosystems). Percent methylation at C5 and C6 was calculated by dividing the peak area of the methylated PCR product (dTAMRA peak) over the total peak area (dTAMRA + dROX). For 9 of the POF patient samples, there was insufficient DNA remaining to assay methylation and these samples were excluded from this analysis.  2.2.6 FMR1 mRNA quantification For a subset of samples, RNA was extracted from a lmL aliquot of peripheral blood from the same vacuum tubes that were drawn for DNA extraction. Total RNA was  66  extracted with the RiboPure-Blood Kit (Ambion) and complete removal of genomic DNA was ensured by 30 minute 37°C incubation with 8 units of DNAse I per 100 uL eluted RNA. Approximately 10Ong of total RNA from each sample was converted to cDNA using the luL of RT Primer Mix, lx Quantiscript RT buffer, and luL Quantiscript reverse transcriptase from the Quantitect Reverse Transcription Kit (Qiagen) in a final reaction volume of 20uL. Reactions were incubated at 42°C for 30 minutes followed by enzyme inactivation at 95°C for 3 minutes. Level of mRNA in peripheral blood was determined by real time quantitative PCR (qPCR) using a TaqMan (Applied Biosystems) protocol that has been previously described (Tassone et al. 2000b) and a relative standard curve method of analysis. The cDNA primers used span the junction between FMR1 exons 3 and 4 and therefore eliminate amplification of genomic contamination. qPCR reactions were performed in 96 well plates with 20uL reactions containing lx TaqMan Universal Master Mix (Applied Biosystems) 0.luM each FMR1 mRNA F (5'-GCA GAT TCC ATT TCA TGA TGT CA3'), FMR1 mRNA R (5'-ACC ACC AAC AGC AAG GCT CT-3') and FMR1 mRNA probe (5'-(FAM)-TGA TGA AGT TGA GGT GTA TTC CAG AGC AAA TGA(TAMRA)-3'), and luL cDNA. Cycling conditions were as follows 95°C for 10 minutes followed by 50 cycles of 95°C for 15 seconds and 60°C for 1 minute. FMR1 mRNA quantity was corrected for quantity of input RNA by normalization to quantity of f3glucoronidase (GUS) a house-keeping gene that normally has an expression level in peripheral blood in the same range as FMR1. GUS amplifications were performed in parallel to FMR1 on each plate using identical reaction conditions and GUS mRNA F (5'CTC ATT TGG AAT TTT GCC GAT T-3' and GUS mRNA R (5'-CCG AGT GAA  67  GAT CCC CTT TTT A-3') primers and GUS mRNA probe (5'-(FAM)- TGA ACA GTC ACC GAC GAG AGT GCT GG-(TAMRA)-3'). FMRI and GUS standard curves were run on each plate using the same sample that was converted to cDNA with each batch of RNA and then serial diluted to concentrations of 500ng, 10Ong, 2Ong, and 4ng input RNA. Both FMRI and GUS qPCR amplifications were performed in triplicate on plates containing both POF patients and controls. FMRI expression for each sample was calculated by determining the average FMR1 amplification of each triplicate according to the standard curve and dividing by the average GUS amplification of each triplicate according to the standard curve. Examination of the rate of FMRI and reference gene, GUS, degradation over time (Figure 2.4) revealed that degradation rates of these two mRNA species in peripheral blood was not comparable, particularly after a time span of 4 days. Therefore all samples for which RNA was not extracted within 4 days of blood draw were excluded from analysis. To ensure precise quantification of FMRI mRNA the qPCR assay was validated and parameters for acceptability were set. To validate the use of quantitative real time PCR analysis, the dynamic range (the range of Ct values over which quantification is linear) of the test transcript (FMR1) and reference gene (GUS) were compared and found to be nearly identical (Figure 2.5a). The linearity of reverse transcription was also assessed using input concentrations of 25, 50 and 10Ong RNA as well as 1 and 2 uL cDNA in each qPCR reaction. Transcription was found to be linear across the range of input concentrations and relative quantity FMRI mRNA was relative to input volume (Figure 2.5b). To control for quality of qPCR runs, the following minimum standards  68  were used: 1) PCR efficiency must be between 90 and 110% (standard curve slope between 3.1 and 3.6) 2) Standard curve R 2 value must be >0.98 and 3) Coefficient of variation between the triplicates must be < 1%. Each RNA sample was converted to cDNA at least twice and quantified at least twice on separate plates and an average of the values was used. 2.2.7 Statistical analysis  For comparison of FMR1 allele size data Fisher's exact test was used because of its sensitivity with small sample sizes. Given our hypothesis that intermediate alleles and alleles at the high end of the normal range would be more prevalent in our patient population, one tailed p values were reported unless otherwise indicated. Odds ratios (OR) with 95% confidence intervals (CI) were also reported for the comparison of the patient group to control group 1. For comparison of DNA methylation and RNA expression level between the POF patients and control group 2, one-tailed t-test was used. For relationships between FMR1 repeat length, FMR1 expression, and FMR1 promoter methylation the significance of the correlation was determined based on the t-distribution. 2.3 Results 2.3.1 FMR1 allele size The FMR1 allele frequencies for the POF patients, control group 1, and control  group 2 are shown in Table 2.3. Premutation size alleles were defined in this study as > 55 CGG repeats and < 200 repeats. There was one allele of 52 repeats in control group 2 which may influence comparison of these results to studies using premutation definitions of 50 — 200 repeats. There was one individual carrying a premutation allele (-70 repeats)  69  in control group 1, and two (62 and —80 repeats) in the POF group. A full mutation allele was detected in one individual in control group 1. FMR1 repeat lengths of >35 repeats were more common in the POF patient population than controls (Table 2.4). FMR1 alleles >35 repeats were found in 18 of 112 (16.1%) alleles in the POF patient population but only 24 out of 326 (7.4%) alleles from control group 1 (OR= 2.4; 95% CI=1.3-4.6; p=0.008, one-tailed Fisher's exact test). Although control group 2 has similar frequency of alleles >35 CGG repeats (8 of 100 alleles or 8.0%), the increased skewing in POF patients is not significant when compared to control group 2 (p = 0.06 vs. control group 2, one-tailed Fisher's exact test), as a result of the smaller sample size in this group. However, when comparing to combined control groups there is a significant increase in frequency of alleles >35 CGG repeats in the POF patient group (p=0.007 one-tailed Fisher's exact test). The prevalence of alleles between 35-54 repeats was also found to be greater in POF patients than controls (p=0.01 vs. control group 1; p=0.02 vs. combined control groups, one-tailed Fisher's exact tests). The two POF patients carrying premutation alleles also carried alleles of >35 CGG on their remaining X chromosome (genotypes were 42,62 and 39,79). However even upon removal of these individuals from the patient population (that is, if we consider that POF in these individuals is attributed to their premutation allele and is not therefore "idiopathic" per se) the POF population still has significantly more alleles between 35-54 CGG repeats than controls (p=0.04 vs. control group 1, p= 0.04 vs. combined control groups, one-tailed Fisher's exact tests).  70  2.3.2 FMR1 biallelic mean In order to assess whether the average size of an individual's FMR1 alleles is a better predictor of risk for POF we examined the biallelic mean of the two FMR1 CGG repeats for the individuals in the patient and control populations. In the POF patient population 8 of 56 patients (14.3%) had an FMR1 biallelic mean >35 CGG repeats, whereas in control group 1 only 8 of 163 (4.9%) did, see table 2.4. (p=0.03 vs. control group 1, p=0.05 vs. combined control groups, one-tailed Fisher's exact tests).  2.3.3 X chromosome inactivation and genotype repeat size X chromosome inactivation skewing ratios were assessed at the AR and FMR1 loci in the patient and control populations. At the AR locus, 12.0% of POF patients and 11.4% of women between 17-45 years of age in control group 1 were homozygous, and therefore uninformative for skewing. At the FMR1 locus, 35.7% of POF patients and 32.4% of women between 17-45 years of age in control group 1 were homozygous or had alleles that were too close together to be separated, and therefore were uninformative for skewing. Using our standard measure of X chromosome inactivation, which assesses skewing primarily at AR, and uses FMR1 only in cases where AR is uninformative, there is no difference in the number of women with extremely skewed X chromosome inactivation (defined as >90% inactivation of one allele) in the POF patient population and control groups (p=1.0 vs. control group 1, two tailed Fisher's exact test). For samples that were informative at both the AR and FMR1 loci, there was a significant, but weak correlation between skewing values measured at the two loci, in both POF patients (N=30, R2 =0.49, p<0.0001, one-tailed test for significance of the correlation based on the  71  t-distribution) and Control group 1 (N=63, R 2 =0.40, p<0.0001, one-tailed test for significance of the correlation based on the t-distribution; Figure 2.6). Genotype repeat size was calculated using both FMR1 and AR skewing values. Although AR may more accurately reflect the true X inactivation, methylation at FMR1 may be a more precise measure of the relative activity of the two FMR1 alleles. More individuals with genotype repeat size >35 were found in the POF population than in the control groups (Table 2.4). Using X chromosome inactivation values calculated at the FMR1 locus, 11 out of 56 (19.6%) POF patients had a genotype repeat number >35, whereas 9 out of 105 (8.6%) samples in control group 1 (p = 0.04, one tailed Fisher's exact test). Results were similar using X chromosome inactivation values calculated at the AR locus (p = 0.03, one tailed Fisher's exact test).  2.3.4 Methylation at the FMR1 promoter To indirectly assess whether FMR1 expression is altered in POF patients with normal allele sizes, DNA methylation at 2 CpG sites (Figure 2.2) in the promoter region of FMR1 was assayed in POF patients (N=48) and women from control group 2 (N=48). Methylation at site C6 was consistently higher than that at site C5, however there was a highly significant correlation between methylation at the two sites in both POF patients and control women (POF patients R 2 =0.15, p=0.003, Control group 2 R 2 =0.37, p<0.0001, one-tailed test for significance of the correlation based on the t-distribution; Figure 2.7). FMRI promoter methylation was similar in POF patients and controls. There was no difference in mean degree of methylation at either CpG site C5 or C6, between the POF patient group (mean methylation C5=40.1± 4.7%; C6=46.7±3.5%) and control group 2 (mean methylation C5=41.0± 4.6%; C6=46.9±4.9%) (p=0.17 and p=0.40 for 72  sites C5 and C6 sites respectively, one-tailed t-test). Furthermore, the range in methylation in POF patients and control group 2 was similar at both sites (see Figure 2.8). Control group 2 is a stringent control for POF patients, since these women have proven reproductive capacity at an advanced reproductive age. Since there was no trend suggesting a difference between FMR1 promoter DNA methylation between these two groups, DNA from women in control group 1 was not assessed for methylation status. Global DNA methylation may decrease with age (Golbus et al. 1990) and methylation at specific gene promoters may also change with age, either as a reflection of global methylation changes or as a result of age-related changes in gene expression (Issa et al. 1994). There was no evidence for age-related changes in FMR1 promoter methylation. Specifically there was no correlation between age and percent methylation at C5 and C6 in either the POF patient group, Control group 2 or both groups combined (Table 2.5, Figure 2.9). Since allele sizes at the higher end of the normal range are associated with an increased risk of POF and the mechanism of POF pathogenesis in FMR1 premutation carriers may be related to elevated levels of RNA, we examined whether promoter methylation is correlated with FMR1 repeat length. There was no relationship between methylation at CpG site C5 or C6 and repeat length of the longer FMR1 allele or FMR1 biallelic mean in either POF patients, Control group 2, or both groups combined (Table 2.5, Figure 2.10). Furthermore, there was also no correlation between methylation and long allele size when considering only samples with FMR1 allele sizes between 35 and 55 repeats (R 2 =0.063 p= 0.18, vs methylation at C5 in combined controls, N=15, onetailed test for significance of the correlation based on the t-distribution).  73  2.33 FMR1 transcript level  FMRI transcript level was assayed in total RNA extracted from peripheral blood of 14 POF patients and 34 members of Control group 2. This assay is moderately reproducible, as shown in Figure 2.11 (R 2 =0.43, slope=0.71, p<0.0001, for combined samples N=48, one-tailed test for significance of the correlation based on the tdistribution), therefore samples were assayed twice and an average of the two values was used. There was no difference in mean relative FMRI expression level between the POF patient group and Control group 2 (p=0.14, two-tailed t-test). Moreover, the POF patients have a trend towards lower relative transcript level than control women (0.55±0.13 in POF patients vs. 0.64±0.20 in Control group 2). FMRI transcript level was not related to FMRI repeat size in this study. Specifically, there was no correlation between relative quantity of FMRI mRNA and repeat length of the longer FMRI allele in either the POF patient group or Control group 2 or both groups combined (Table 2.6, Figure 2.13). Likewise, there was no correlation between relative quantity of FMRI mRNA and FMRI biallelic mean either the POF patient group or Control group 2 or both groups combined (Table 2.6, Figure 2.13). Methylation at CpG sites in the FMR1 promoter region is inversely correlated with FMRI transcript level, at least in a qualitative manner in males with methylated full mutation alleles (Pieretti et al. 1991). We therefore examined whether FMRI transcript level is related to methylation at FMRI promoter sites C5 and C6 (Figure 2.14). There was a weak inverse correlation between FMRI transcript level and DNA methylation that was significant only at promoter site C5 for correlations with Control group 2 alone or combined controls (p=0.01 and 0.02 for comparisons to Control group 2 and both groups  74  combined, respectively, one-tailed tests for significance of the correlation based on the tdistribution).  2.4 Discussion Currently many clinics are screening women experiencing POF for the presence of FMR1 premutation size alleles on a routine basis. However our present results suggest that FMR1 allele between 35 and 55 repeats are also associated with POF. If these results are found to be reproducible in larger populations, diagnostic testing and counselling of POF patients may need to be reassessed. These observations also suggest that the mechanism of FMR1 involvement in POF is distinct from that of Fragile X syndrome.  2.4.1 FMR1 CGG repeat length and POF We report a statistically significant association between POF and FMR1 repeat sizes between 35-54 repeats. These findings are supported by a study that also found an increased incidence of grey zone alleles in a group of 190 Italian POF patients (Bodega et al. 2006). Although the category of alleles found at increased frequency in this group was defined as 41-58 repeats, it nonetheless highlights the fact that the range of FMR1 alleles affecting POF risk may be larger than previously thought. This will have important affects on public health as it delineates a much larger group of women that may be at risk of POF. The FMR1 repeat sizes found in the control population in this study are comparable to those in other studies of FMR1 allele frequencies in the general population (Dawson et al. 1995; Larsen et al. 1997; Patsalis et al. 1999b; Snow et al. 1993). All studies find that alleles of 35-54 repeats are relatively rare in normal control populations  75  and make up only —5 (Dawson et al. 1995; Larsen et al. 1997) to —9% (Patsalis et al. 1999b; Snow et al. 1993) of alleles. Therefore the 14% of alleles that fall into this category in the POF population reported here is indeed elevated over controls. A recent study examining the prevalence of POF and early menopause in women with a variety of sizes of FMR] alleles found an odds ratio for intermediate size alleles (defined as 41-58 repeats) that is nearly identical to the 2.3 we observed for alleles between 35-54 repeats (Sullivan et al. 2005). Both biallelic mean and genotype repeat size calculated using AR skewing values, were also associated with incidence of POF. However it remains to be determined whether either of these values are better predictors of POF risk than FMR1 repeat size alone. Allen et al. report that FMR1 genotype repeat size (measured using X chromosome inactivation ratios calculated at AR, FMR1 or DXS6673E) only explains —9% of the variation seen in the natural log of the FMR1 transcript level (2004). If it is indeed FMR1 transcript level that is critical in POF risk and expression of FMR1 doesn't increase linearly with repeat size, examining biallelic mean or genotype repeat size may not accurately reflect transcript levels. Although FMR1 transcript level and repeat size are linearly related, this linear effect is primarily in the premutation range (Allen et al. 2004) and is diluted in females due to X chromosome inactivation skewing (GarciaAlegria et al. 2007). Furthermore, X chromosome inactivation skewing values measured in blood may not correlate well with X inactivation ratios in the tissue or developmental stage where the FMR1 transcript effects are critical in ovarian development.  76  2.4.2 No association between FMR1 expression and POF  In this study we found no evidence to suggest that increased expression of FMR1 may be involved in POF pathogenesis for women without FMR1 premutation alleles, at least as measured in blood. There was no difference in either FMR1 transcript level or promoter methylation between the POF patient group and controls. It appears that the pathogenic role of FMR1 in POF is restricted to those with expanded repeat lengths. This is not surprising given that there are a large number of genes with suspected involvement in POF, and that POF may be a part of the spectrum of age at menopause and can therefore be considered a quantitative trait for which many genetic factors would be expected to play a role. We found no correlation between FMR1 genotype and repeat length or transcript level for samples (primarily normal and grey zone alleles) in this study. Given the small sample sizes, the range of transcript levels found in individuals with normal allele sizes, and the inherent variability in the qPCR assay, this study may not have been sensitive enough to detect any correlation that does exist within this range. A correlation between repeat length for grey zone alleles and transcript level has been reported for males (Loesch et al. 2007); however, this effect is likely diluted in females due to X chromosome inactivation skewing. Regardless, there were too few RNA samples from individuals with grey zone alleles in this study to warrant any meaningful analysis of whether transcript level may be correlated with repeat length for this group. Since the FMR1 gene is on the X chromosome and half of the X chromosomes in a woman are inactivated and exhibit methylation-induced silencing of the FMR1 gene, it was expected that the level of methylation in control women would be approximately 77  50% on average. The mean degree of methylation in both POF patients and controls was however, lower than the expected 50%. Whether this reflects the true state of methylation at these sites or is a result of inherent technical aspects of the assay that result in a bias towards hypomethylation is unclear. If this assay does reflect the state of methylation at the FMRI promoter, it suggests that a moderate loss of methylation and expression of FMRI from the inactive X chromosome may occur. Slight differences in the extent of this "leaky" expression could explain the range of transcript levels found in normal individuals. However, it should be emphasized that this assay assesses methylation only at two CpG sites on the X chromosome, and there may be loss of methylation at specific CpG sites even if the promoter in general remains mostly methylated. 2.4.3 Possible mechanisms for FMR1 in POF  It is possible that the mechanism of the association between FMRI and POF is the pathogenic effect of the elevated transcript levels that are associated with increased FMRI repeat size. Allen et al. (2004) report a positive linear relationship between FMRI  CGG repeat size and transcript level in both males and females, however they found this relationship primarily in the premutation range (defined in their study as 61-200 repeats). When CGG repeats of various sizes in the context of the FMRI promoter and a luciferase reporter were transfected into human cell lines, increasing CGG repeat size was found to cause elevated mRNA expression and decreased protein translation (Chen et al. 2003) an effect that has also been observed in vivo for repeat sizes in the grey zone range (Loesch et al. 2007). Elevated mRNA transcript levels could be pathogenic in a number of ways. The expression of expanded CGG repeats causes intranuclear inclusions of a RNA-  78  protein complex in the brains of mice (Willemsen et al. 2003) and Drosophila (Jin et al. 2003) resulting in neuronal degeneration. Inclusions have also been observed in males with FXTAS (Greco et al. 2002) and provide a possible explanation for pathology associated with FMR1 premutations and not full mutation alleles. A similar mechanism could be at play in the case of POF, with accumulation of RNA-protein clusters causing premature degradation of ovarian follicles. In contrast, elevated FMR1 mRNA levels could also acquire excess RNA binding proteins on their expanded repeat tract. Excess FMR1 transcripts of large repeat alleles could act as a binding protein sink and sequester  RNA binding proteins necessary for the translation of other proteins that contain similar CGG repeat tracts and are necessary for proper ovarian function. An analogous model has been suggested in the pathogenesis of myotonic dystrophy type 1 (DM1). In DM1 a CUG expansion in the 3'untranslated region of the DMPK gene binds and depletes transcription factors from active chromatin, resulting in decreased expression of a number of genes causing the characteristic pleiotropic symptoms of the disorder (Ebralidze et al. 2004). Alternatively, the pathogenic factor increasing POF risk in women with FMR1 expansions may be related to the nature rather than the quantity of FMR1 transcript. Firstly, the CGG repeat region at FMR1 is transcribed, but not translated, and therefore women with longer FMR1 repeats will have not only elevated quantity of transcript but also transcripts with more CGG repeats than normal. Transcripts containing long CGG repeat tracts may have a direct effect by forming toxic secondary structures or an indirect effect by sequestering more of a CGG-specific binding protein. Secondly, in normal individuals the FMR1 CGG repeat tract is commonly interrupted by one or two AGG  79  repeats. Notably, carriers of FMRI grey zone alleles who experience POF were found to have uninterrupted CGG repeat tracts, while those without POF had the normal AGGinterrupted repeats (Bodega et al. 2006). Uninterrupted CGG repeat tracts are predicted to be prone to form more complex secondary structures than those with interruptions (Bodega et al. 2006). Thirdly, FMRI has at least three known transcription start sites (Figure 2.2), the use of which will result in subtle increases in mRNA length at the 5' end (Beilina et al. 2004). The use of these start sites varies with CGG repeat length, with the most commonly used start site (Site I, used for —70% of transcripts in individuals with normal repeat sizes) being used relatively infrequently with increased repeat length (used for <20% of transcripts in carriers of 160 CGG repeats) (Beilina et al. 2004). The mechanism of FMRI involvement in POF may therefore be through an indirect effect on transcription start site usage. Finally, very recent reports of novel mRNA transcripts that appear to be influenced by FMRI CGG repeat length (Khalil et al. 2007; Ladd et al. 2007) highlight the possibility that it may be an entirely different transcript that is critical in POF. The FMRI gene presents an intriguing trinucleotide repeat in which the phenotype associated with the premutation, an increased incidence of POF and FXTAS, is distinct from that of the full mutation, Fragile X syndrome. We report here that FMRI repeat sizes in the intermediate (grey zone) and high end of the normal ranges are associated with an increased risk for POF. These preliminary results are clinically relevant, and bring to light the necessity for a separate understanding of the FMRI gene repeat in the context of Fragile X syndrome and POF. Our results draw attention to the need for further study in this area, in particular the examination of FMRI repeat sizes  80  among larger POF patient populations, rather than merely incidence of POF in FMR1 premutation carriers. It is also necessary to determine whether the best predictor of POF risk is the size of an individual's largest FMR1 repeat, the biallelic mean of their genotype repeats, or an XCI ratio adjusted repeat number such as genotype repeat size. In addition, we found no evidence that the FMR1 gene plays a role in POF pathology in patients without expanded repeats. There is a need for further study to not only determine the mechanism by which FMR1 causes POF, but also to discover other genetic factors that interact with FMR1 to influence risk of POF.  81  Table 2.1 Studies examining incidence of POF in Fragile X Syndrome families Incidence of POF in carriers of Reference^ Cronister et al. 1991^ Schwartz et al. 1994^ Partington et al. 1996^  Definition of premutationl^Normal^Premutation^Full mutation n.s. 2^  8/61 (13.1%)  50-200^2/34 (5.9%)^12/49 (24.5%)^3/8 (37.5%) n.s. 2^  14/49 (28.6%)  Allingham-Hawkins et al. 1999^n.s.2^1/237 (0.4%)^63/395 (15.9%)^0/128 (0.0%) Uzielli et al. 1999^  >54^0/118 (0.0%)^32/170 (18.8%)^0/63 (0.0%)  Vianna-Morgante 1999^  n.s. 2^0/48 (0.0%)^11/88 (12.5%)^0/29 (0.0%)  Hundscheid et al. 2000^ Murray et al. 2000^  50-200^  24/109 (22.0%)  >51^3/205 (1.5%)^10/116 (8.6%)^2/31 (6.5%)  Vianna-Morgante & Costa 2000^n.s. 2^0/50 (0.0%)^14/59 (23.7%) Mallolas et al. 2001^  >52^  Hundscheid et al. 2003^  >50^  12/98 (12.2%)^0/6 (0.0%) 22/103 (21.4%)  Machado-Ferreira et al. 2004^ n.s. 2^0/19 (0.0%)^11/33 (33.3%)^0/6 (0.0%) Sullivan et al. 2005^ 'range of CGG repeats considered premutation n.s.=not stated; premutation size not defined  2  >59^2/157 (1.2%)^12/93 (12.9%)  Table 2.2 Studies examining FMR1 repeat length in women with POF  Reference^  Frequency of premutation alleles in Definition of^ premutationl^Controls^Sporadic POF^Familial POF 0/37 (0.0%)^2/9 (22.2%)  Conway et al. 1995^  60-200^  Kenneson et al. 1997^  —60-200^1/107 (0.9%)^0/16 (0.0%)^0/17 (0.0%)  Murray et al. 1998^  61-200^0/1268 (0.0%)^2/244 (0.8%)^4/50 (8.0%)  Conway et al. 1998^  60-200^  3/106 (2.8%)^3/23 (13.0%)  Patsalis et al. 1999a^  n.s. 2^  Syrrou et al. 1999^  n.s. 2^0/100 (0.0%)^0/16 (0.0%)^1/7 (14.2%)  0/18 (0.0%) 3A  Uzielli et al. 1999^  54-200^0/332 (0.0%)^7/108 (6.5%) 3  Marozzi et al. 2000^  60-200^  2/61 (3.0%)^4/33 (12.0%)  Mallolas et al. 2001^  52-200^  1/28 (3.6%)^1/15 (6.7%)  Bussani et al. 2004^  50-200^0/28 (0.0%)^3/40 (7.5%)^0/5 (0.0%)  Gersak et al. 2003^  —53-200^  range of CGG repeats considered premutation n.s.—not stated; premutation size not defined 3 familial not distinguished from sporadic cases 4 these patients are described as primary or secondary amenorrheic, not POF I  2  4/83 (4.8%)  Table 2.3 FMR1 alleles in POF patients and Control populations # CGG repeats 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 52 62 -70 -80 Full mutation total  POF patients 0 (0.0%) 0 (0.0%) 0 (0.0%) 5 (4.5%) 2 (1.8%) 0 (0.0%) 10 (8.9%) 2 (1.8%) 0 (0.0%) 1 (0.9%) 3 (2.7%) 0 (0.0%) 36 (32.1%) 26 (23.2%) 4 (3.6%) 4 (3.6%) 1 (0.9%) 0 (0.0%) 0 (0.0%) 1 (0.9%) 4 (3.6%) 0 (0.0%) 4 (3.6%) 3 (2.7%) 1 (0.9%) 2 (1.8%) 1 (0.9%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (0.9%) 0 (0.0%) 1 (0.9%) 0 (0.0%) 112  Control group 1 2 (0.6%) 1 (0.3%) 1 (0.3%) 30 (9.2%) 4 (1.2%) 3 (0.9%) 30 (9.2%) 1 (0.3%) 2 (0.6%) 1 (0.3%) 4 (1.2%) 1 (0.3%) 72 (22.1%) 105 (32.2%) 17 (5.2%) 20 (6.1%) 6 (1.8%) 2 (0.6%) 0 (0.0%) 1 (0.3%) 1 (0.3%) 6 (1.8%) 4 (1.2%) 1 (0.3%) 2 (0.6%) 2 (0.6%) 0 (0.0%) 2 (0.6%) 1 (0.3%) 1 (0.3%) 1 (0.3%) 0 (0.0%) 0 (0.0%) 1 (0.3%) 0 (0.0%) 1 (0.3%) 326  Control group 2 0 (0.0%) 0 (0.0%) 1 (1.0%) 6 (6.0%) 0 (0.0%) 1 (1.0%) 9 (9.0%) 2 (2.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 6 (6.0%) 23 (23.0%) 31 (31.0%) 4 (4.0%) 5 (5.0%) 3 (3.0%) 1 (1.0%) 2 (2.0%) 0 (0.0%) 0 (0.0%) 2 (2.0%) 0 (0.0%) 1 (1.0%) 0 (0.0%) 0 (0.0%) 1 (0.0%) 1 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 1 (1.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 100  84  Table 2.4 FMR1 repeat size, biallelic mean and genotype repeat size  POF patients 56  Control group 1 163  Control group 2 50  P value  Odds ratio' (95% CI)  alleles >35  18/112 (16.1%)  24/326 (7.4%)  8/100 (8.0%)  0.008, 0.06, 0.007 2  2.4 (1.3-4.6)  alleles 35-54  16/112 (14.3%)  22/326 (6.7%)  8/100 (8.0%)  0.01, 0.11, 0.02 2  2.3 (1.2-4.6)  Biallelic mean >35  8/56 (143%)  8/163 (4.9%)  5/50 (10.0%)  0.03, 0.4, 0.05 2  3.2 (1.2-9.1)  GRS (AR) 3 >35  10/50 (20.0%)  7/97 (7.2%)  0.02 4  3.1 (1.1-9.1)  GRS (FMRJ) 3 >35  11/56 (19.6%)  9/105 (8.6%)  0.04 4  2.6 (1.0-6.7)  N individuals  'for the comparisons of POF patients to control group 1 one tailed Fisher's exact tests for comparisons of POF patients to control group 1, control group 2 and combined controls respectively 3 GRS (AR) and GRS (FMR1) = genotype repeat size calculated with AR or FMR1 skewing values 4 one tailed Fisher's exact test for comparisons of POF patients to control group 1 2  Table 2.5 Correlations with FMR1 methylation levels Site C5 R2^P value  Site C6 R2^P value  Methylation and age at blood draw' POF patients (N=32)  0.048  0.22  0.002  0.83  Control group 2 (N=48)  0.014  0.42  0.016  0.39  Combined samples (N=80)  0.003  0.65  0.006  0.83  POF patients (N=48)  0.015  0.20  0.019  0.18  Control group 2 (N=48)  0.012  0.23  0.004  0.33  Combined samples (N=96)  0.015  0.11  0.010  0.17  POF patients (N=48)  0.025  0.14  0.012  0.23  Control group 2 (N=48)  0.004  0.34  0.0001  0.47  Combined samples (N=96)  0.017  0.10  0.003  0.30  Methylation and FMR1 longer allele 2  Methylation and FMR1 biallelic mean t  l 2  two-tailed test for significance of the correlation based on the t-distribution one-tailed test for significance of the correlation based on the t-distribution  86  Table 2.6 Correlations with FMR1 expression levels R2  P value'  Expression and FMR1 longer allele POF patients (N=14)  0.005  0.41  Control group 2 (N=34)  0.001  0.44  <0.0001  0.48  POF patients (N=14)  0.077  0.17  Control group 2 (N=34)  0.0001  0.48  Combined samples (N=48)  0.006  0.30  POF patients (N=14)  0.023  0.30  Control group 2 (N=34)  0.149  0.01  Combined samples (N=48)  0.092  0.02  POF patients (N=14)  0.011  0.36  Control group 2 (N=34)  0.055  0.09  Combined samples (N=48)  0.020  0.16  Combined samples (N=48) Expression and FMR1 biallelic mean  Expression and methylation at site C5  Expression and methylation at site C6  i one-tailed test for significance of the correlation based on the t-distribution  87  acctgtcaccgcccttcag  Methylated auut gtuauc guuut tuag  Unmethylated  Bisulfite treatment  auut gt uauuguuut. tuag -  PCR region surrounding CpG island atttgttatcgttttttag taaacaatagcaaaaaatc  atttgttattgttttttag taaacaataacaaaaaatc  SNAPshot PCR with ms-SNuPE primer [ taaacaatagcaaaaaatc  taaacaataacaaaaaatc  ■  atttgttatC* taaacaatagcaaaaaatc  QuantifST C* and T* by ABI 310 fragment analysis:  31% methylation  42.4% methylation  Figure 2.1 Methylation quantification by single nucleotide primer extension (SNuPE). Genomic DNA is bisulfate treated, converting unmethylated cytosines (c) to uracil (u) (red) while methylated cytosines remain unconverted (black). Bisulfite converted DNA is PCR amplified and a SNuPE reaction, which adds only a single fluorescently labeled dNTP, is performed. Extension products are quantified by ABI 310 fragment analysis.  Accession no. X61378 H.sapiens fragile X DNA -996 ttaaaaaata tatagtcaag tgaaagtatg aaaatgagtt gaggaaaggC gagtaCgtgg -916 gtcaaagctg ggtctgagga aaggctcaca ttttgagatc cCgactcaat ccatgtccct -856 taaagggcac agggtgtctc cacagggcCg cccaaaatct ggtgagagag ggCgtagaCg -796 cctcaccttc tgcctctaCg ggtcacaaaa gcctgggtca ccctggttgc cactgttcct -736 agttcaaagt cttcttctgt ctaatccttc acccctattc tCgccttcca ctccacctcc -676 Cgctcagtca gactgCgcta ctttgaacCg gaccaaacca aaccaaacca aaccaaacca 616 aaccagacca gacaccccct ccCgCggaat cccagagagg cCgaactggg ataacCggat FlAR1-MS F -556 gcatttgatt tcccaCgcca citlagtgcac ctctgcagaa atgggCgttc tggccctCgC  ^CS^C6 496FAcicagtqC Wctcrtcac_Igcccttcag ccttcclgcc ctccaccaag ccCgCgcaCg Eagl  436 ccCggccCgC gCgtctgtct ttCgaccCgg cacccCggcC ggttcccagc agCgCgAll FMRI-MS R 376 CgCgCgctcc caggccactt claaqaclacraq'ggggcCg aggggctgag ccCgCggggg  ^Site-111^Site4I —■^—■ -316 gagggaaCag Cgttgatcac gtgaCgtggt ttcagtgttt acaccCgcag CgggcCgggg Site 4 FIR l - F HpaII —■^ 256 gttCggccct agtcagguppc tcagctcCgt ttCqqtttca cttcCggtgg agggcCgcct  -196 ctgagCgggc ggCgggccga CggCgagCgC gggCggCggc ggtgaCggag gCgcCgctgc -136 cagggggCgt gCggcagCgC ggCggqggq,g_gCggCggCgg CggCggCggC ggCggCggCg HpaII -76 gCggCggctg ggcctCgagC gccCgcagcc cacctctCgg gggCgggctc cCggCgctag  +1 Translation start -16 cagggctgaa gagaagatgq aggagctqgt qgtggaagtg cqqq4c4cca atggcgcttt +44 ctacaaggta cttggctcta gggcaggccc catcttcgcc cttccttccc tcccttttct +104 tcttggtgtc ggcgggaggc aggcccgggg ccctcttccc gagcaccgcg cctgggtgcc  Figure 2.2 FMR1 gene promoter region. Approximately 1Kb of sequence upstream of the FMR1 translation start site is shown. Three known transcription start sites are indicated with an arrow and designated Sites I, II, and III. FMRP coding region is shown  in green. Polymorphic CGG repeat is highlighted in red. Transcription factor binding sites are shown in blue. Cytosine sites subject to methylation are bold and capitalized. Primersusedforms-SNuPE(FMR1-MSFandFMR1-MSR)andFMR/repeatlength determination and XCI skewing (FMR1-F and FMR1-R) are shown underlined in black with arrows. CpG sites C5 and C6 are highlighted in dark blue and purple and SNuPE primers used for ms-SNuPE are in corresponding colors. The EagI site used in Southern blot and HpaII sites used in XCI skewing are underlined. (adapted from Beilina et al. 2004 and Stoger et at 1997)  89  L12  ^  4 5  ^  7 8 9 10 11 12 13 14 15 L  8.6 7.4 61 49  3.6  2.8  Figure 2.3 Example of FMR1 Southern blot results. Digestion is performed with both EcoRl, and methylation sensitive EagI. Upper bands (at —5.2kb) are the methylated inactive allele; lower bands (-2.8kb) are the unmethylated active allele. Lanes 1-15: samples amplifying a single band by FMR1 PCR; Lane 12 failed. All other samples are homozygous except for lane 11, which carries alleles of 30 and 70 repeats. Overloading in lanes 3, 7, and 15 results in dark, poorly migrating bands.  90  Time (hours) 0 12 24 48 96 192 B  28 S 18 S  B)  • FMR1 100 ng 6  ^• FMR1 50 ng  • •  • FMR1 25 ng  •  GUS 100 ng  0  GUS 50 ng ▪ GUS 25 ng  0^50^100^150^200^250 time since blood draw (hrs)  Figure 2.4 RNA degradation over time. A) Total RNA was extracted at 0, 12, 24, 48,  96, and 192 hours and separated by gel electrophoresis. The presence of 18S and 28S bands indicates that quality total RNA could be extracted even 8 days after blood draw. B) 100, 50 or 25ng of RNA extracted at different time points was quantified by real-time PCR and the number of cycles required to reach threshold (Ct) is indicated. After 96 hours the Ct for the reference gene, GUS, is higher than that of FMR1, indicating that GUS degrades more quickly.  91  A) 35.0 34.5 34.0 33.5 33.0 C) 32.5 32.0 31.5 31.0 30.5  0 FMR1 ■ GUS — Linear (FMR1) ^ Linear (GUS)  25  50^75 input RNA (ng)  100  B) 5 4 3 2  .9. g 1 rz. x a) 0 a.) .-occ! -1 •5 75 -2 -3 -4 -5  is  50  o luL cDNA ■ 2uL cDNA • relative Linear (l uL cDNA) 75^11)0 ^ Linear (2uL cDNA) - - - - Linear (relative)  input RNA (ng)  Figure 2.5. Validity of qPCR assay for FMR1 with reference gene GUS. A) the input quantity of RNA falls within the dynamic range (the range of Ct values over which quantification is linear) of the test transcript (FMR1) and reference gene (GUS). B) The linearity of reverse transcription with 25, 50 and 100 ng input RNA as well as 1 or 2 uL cDNA in each qPCR reaction.  92  100% -  0  o^ 0^  •  o  •  90% -  70% -  60% -  •  •  • •o • •• • 50% ^ 0  •  I 50%^60%^70%^80%^90%^100% Degree of XCI skewing AR locus •  Control group 1 (N=63) ^o POF patients (N=30) Linear (Control group 1 (N=63)) — — Linear (POF patients (N=30))  Figure 2.6 Correlation between XCI skewing assayed at the AR and FMRJ loci.  There was a significant correlation between skewing at the AR and FMR1 in POF patients (R2 =0.49, p<0.0001) and Control group 1 (R 2 =0.40, p<0.0001) (one-tailed p values for significance of the correlation based on the t-distribution).  93  ▪  0.60 -  • •  0.55 -  0 •  ••  •0  0^• • 0 •^°11 o 0^ # e^4 •.0.- ... 0^ • m0^--• • ..... 0 ..-^• •^o o^ ^• o o^c9 o ^• * 0 • 0^...... ....^cs .....^ ciP (2.-- • 0  v:)  1E; 0.50 -  f,  :  0.45 -  E  (2) 0  0  o•  ••  to 0.40 -  A  •  40  • 0  0  • •  0.35 -  0.30 ^ 0.25^0.30^0.35^0.40^0.45^0.50^0.55 Degree of methylation at C5 •  Control group 2 (N=48)  0 POF patients (N=48)  Figure 2.7 Correlation between FMRJ methylation assayed at sites C5 and C6.  There was a weak but significant correlation between sites C5 and C6 in POF patients (R2 =0.15, p=0.003) and Control group 2 (R 2 =0.37, p<0.0001) (one-tailed p values for significance of the correlation based on the t-distribution).  94  ^ ^  •• o^••  55% -  0  ^o^g^$ o 50% -  •  0  • § 0^  ^o  •  0  o ^ 8 o o o^  0  8  /0  1 i_ • •  •  •^•  1 • mr $ li.^ ♦^• o o • ♦ ^e ^ :^ o oo  e^ 0o  o  8 0^  ••  • $^ • •  30% 0  25%  C5^C6  C5^C6  0 POF patients (N=48)  • Control group 2 (N=48)  Figure 2.8 FMR1 promoter methylation assayed at CpG sites CS and C6. Average methylation at sites C5 and C6 was 40.1±4.7%, 46.7±3.5% for POF patients and 41.0±4.6%, 46.9±4.9%, for Control group 2 (p=0.17 and p=0.40 for sites C5 and C6 respectively, one-tailed p values for significance of the correlation based on the tdistribution).  95  •  55% -  kr)  tE; 45%  •  ‘•f • •••  •  cis t, 40% 8 to 35% a.)  •  •• $  50% -  • •  •  30% -  • • ••  • •^• •• •^• • ••••• • •^ • • ^•^ • •••• ••^ •• • $•• $ •^• •• • * • • •  •  25% ^ 20  30  40^50  60  Age at blood draw (years)  Figure 2.9 FMR1 promoter methylation at site C5 by age. There is no correlation between site CS promoter methylation and age in POF patients (N=32) and Control group 2 samples (N=50) combined (W=0.003, p=0.65, one-tailed p values for significance of the correlation based on the t-distribution). Information on age at time of blood draw was not available for all POF patients therefore sample size is reduced in this analysis. A lack of correlation was also seen for site C6.  96  A) 55% kr) 50% - ••• • • -lics' .•• • . 4 . ^• .44 45% -• • 4 .  •  ••  •  t 40% - •^••  •• :6 • ^• • • 35% -^ • •• • o  •  •  •  :  ***• • • •^•  30% -  •  25% 20  40^60  80  Repeat length of long FMR1 allele B) 55% -  • ••  U 50% - • g 45% f, 40% E  D  35%  6) 30% -  • * • • • • • • • • '0 • • •^ • 44, •• • ••• • • • •• •* * • • • No* •• • • ••• •• • • • • •  •  &  •  •  25% 20^30^40^50^60 FMR1 biallelic mean  Figure 2.10 Promoter methylation at site C5 and FMR1 repeat length. There was no correlation between methylation at site C5 and A) long FMR1 allele (R =0.015, p=0.11) or B) FMR1 biallelic mean (R 2 =0.017, p=0.10) in POF patients (N=48) and Control group 2 samples (N=48) combined (one-tailed p values for significance of the correlation based on the t-distribution). Similar results were seen for site C6.  97  0.2^0.4^0.6^0.8^1.0  ^  1.2  ^  1.4  relative FMR1 expression first run  Figure 2.11 Reproducibility of FMR1 quantitative PCR assay. All samples were reverse transcribed into cDNA and quantified in triplicate by real time PCR twice. Error bars show standard deviation of triplicate samples run on each plate. The correlation is R2 =0.43, p<0.0001 for combined samples, N=48 (one-tailed p value for significance of the correlation based on the t-distribution).  98  1.4 -  0  .— v)  1.2 -  2 1.0 0.8  w  -  0.6-  a.) 0.4 ct  0.2 0.0  IfII  NJ NJ NJ NJ NJ NJ NJ NJ (44 U4 U.) NJ ^,J (44 VD CD LO CD CD LO CD CD^vp 144 NJ 144 1NJ 144 144 -NJ 'NJ 4=. 1,4 1,4 •4 ^( 44 CD VD J CD VD VD CD CN^VD  1  (4.) NJ L4 NJ (44 NJ W NJ NJ NJ NJ NJ NJ NJ NJ O CD VD CD^VD .-, U4 00 VD U4 VD (44 (44 U.) 144 4J IJ ".p. 1,4 1,4 1,4 1,4 Iv 1.4 1,4 1..4 1.4 1.4 1.4 1,4 O CD VD CD NJ CD NJ l0 U4 00 CD VD CD NJ CD CD (.44  1  1 ^1 1 i 1 I I 1  NJ U.)^NJ NJ NJ to uJ („J t.,J NJ NJ NJ Lo NJ NJ NJ w CD 00 CD VD VD CD 00 CD CD CD CD VD CD CD CD Lo CD VD CD C) ,4 1,4 1,4 1.4 1,4 1,4 1/1 1,4 1,4 1,4 1.4 N W 1•4 144 1,4 1,4 1,4 1,4 1,4 CD^CD CA CD ," NJ C) CD CD CD VD trl 00 NJ VD VD CD 4 t44 (44  1  FMR1 repeat genotype  Figure 2.12 Relative FMRI transcript level in POF patients and Controls. Each bar represents one individual in the POF patient group (white bars, N=14) or Control group 2 (black bars, N=34). Error bars show the standard deviation of the replicates for each sample. FMRI repeat genotypes are shown on the X axis beneath each bar.  A) 1.50 = 1.25 0  E.-, Loo a, t5 ,_,.  0.75  713  ...o 0.50 71) c4 0.25 -  1  i T  • E  0.00 20 22 24 26 28 30 32 34 36 38 40 Repeat length of long FMR1 allele B) 1.50 -  0.25 0.00 20 22 24 26 28 30 32 34 36 38 40 FMR1 biallelic mean  Figure 2.13 Relative transcript level by FMR1 repeat length. There was no correlation between transcript level and A) long FMR1 allele (R 2 <0.0001, p=0.48) or B)FMR1 biallelic mean (R 2=0.006, p=0.30) in POF patients (N=14) and Control group 2 (N=34) combined (one-tailed p values for significance of the correlation based on the tdistribution). Error bars show standard deviation of the replicates for each sample. 100  • A) 1.4 1.2 1.6• 1.0 c) 0.8  -  0.6a^  ft$ 0.4 a.) r:4 0.2 .  0.0 ^ 25%^30%^35%^40% 45%^50%^ 55% Degree ofmethylation at C5 B) 1.4 1.2 •••••I C4  1.0  -  Cal  0.8  -  0.6 a)  ct 0.4 '75  as  0.2 0.0 30%^35%^40%^45%^50%^55%^60% Degree ofmethylation at C6  Figure 2.14 Relative FMRI transcript level and methylation. There was a significant  correlation between transcript level and methylation at A) site C5 (R 2 =0.092, p=0.02) but not B) site C6 (R 2 =0.020, p=0.16) in POF patients (N=14) and Control group 2 (N=34) combined (one-tailed p values for significance of the correlation based on the tdistribution). Error bars show standard deviation of the replicates.  101  2.5 References Allen EG, He W, Yadav-Shah M and Sherman SL. 2004. A study of the distributional characteristics of FMR1 transcript levels in 238 individuals. Hum. Genet. 114:439447. Allen EG, Sullivan AK, Marcus M, Small C, Dominguez C, Epstein MP, Charen K, He W, Taylor KC and Sherman SL. 2007. Examination of reproductive aging milestones among women who carry the FMR1 premutation. Hum. Reprod. 22:2142-2152. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM and Belmont JW. 1992. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgenreceptor gene correlates with X chromosome inactivation. Am. J. Hum. Genet. 51:1229-1239. Allingham-Hawkins DJ, Babul-Hirji R, Chitayat D, Holden JJ, Yang KT, Lee C, Hudson R, Gorwill H, Nolin SL, Glicksman A, et al. 1999. Fragile X premutation is a significant risk factor for premature ovarian failure: the International Collaborative POF in Fragile X study--preliminary data. Am. J. Med. Genet. 83:322-325. Beever CL, Stephenson MD, Penaherrera MS, Jiang RH, Kalousek DK, Hayden M, Field L, Brown CJ and Robinson WP. 2003. Skewed X-chromosome inactivation is associated with trisomy in women ascertained on the basis of recurrent spontaneous abortion or chromosomally abnormal pregnancies. Am. J. Hum. Genet. 72:399-407. Beilina A, Tassone F, Schwartz PH, Sahota P and Hagerman PJ. 2004. Redistribution of transcription start sites within the FMR1 promoter region with expansion of the downstream CGG-repeat element. Hum. Mol. Genet. 13:543-549. Bodega B, Bione S, Dalpra L, Toniolo D, Ornaghi F, Vegetti W, Ginelli E and Marozzi A. 2006 Influence of intermediate and uninterrupted FMR1 CGG expansions in premature ovarian failure manifestation. Hum. Reprod. 21:952-957. Boyd VL, Moody KI, Karger AE, Livak KJ, Zon G and Burns JW. 2006. Methylationdependent fragment separation: direct detection of DNA methylation by capillary electrophoresis of PCR products from bisulfite-converted genomic DNA. Anal. Biochem. 3 54:266-273. Bussani C, Papi L, Sestini R, Baldinotti F, Bucciantini S, Bruni V and Scarselli G. 2004. Premature ovarian failure and fragile X premutation: a study on 45 women. Eur. J. Obstet. Gynecol. Reprod. Biol. 112:189-191. Chen LS, Tassone F, Sahota P and Hagerman PJ. 2003. The (CGG)n repeat element within the 5' untranslated region of the FMR1 message provides both positive and negative cis effects on in vivo translation of a downstream reporter. Hum. Mol. Genet. 12:3067-3074.  102  Conway GS, Hettiarachchi S, Murray A and Jacobs PA. 1995. Fragile X premutations in familial premature ovarian failure. Lancet. 346:309-310. Conway GS, Payne NN, Webb J, Murray A and Jacobs PA. 1998. Fragile X premutation screening in women with premature ovarian failure. Hum. Reprod. 13:1184-1187. Crawford DC, Meadows KL, Newman JL, Taft LF, Pettay DL, Gold LB, Hersey SJ, Hinkle EF, Stanfield ML, Holmgreen P, et al. 1999. Prevalence and phenotype consequence of FRAXA and FRAXE alleles in a large, ethnically diverse, special education-needs population. Am. J. Hum. Genet. 64:495-507. Cronister A, Schreiner R, Wittenberger M, Amin K, Harris K and Hagerman RJ. 1991. Heterozygous fragile X female: historical, physical, cognitive, and cytogenetic features. Am. J. Med. Genet. 38:269-274. Dawson AJ, Chodirker BN and Chudley AE. 1995. Frequency of FMR1 premutations in a consecutive newborn population by PCR screening of Guthrie blood spots. Biochem. Mol. Med. 56:63-69. Ebralidze A, Wang Y, Petkova V, Ebralidse K and Junghans RP. 2004. RNA leaching of transcription factors disrupts transcription in myotonic dystrophy. Science. 303:383387. Garcia-Alegria E, Ibanez B, Minguez M, Poch M, Valiente A, Sanz-Parra A, MartinezBouzas C, Beristain E and Tejada MI. 2007. Analysis of FMR1 gene expression in female premutation carriers using robust segmented linear regression models. RNA. 13:756-762. Gersak K, Meden-Vrtovec H and Peterlin B. 2003. Fragile X premutation in women with sporadic premature ovarian failure in Slovenia. Hum. Reprod. 18:1637-1640. Golbus J, Palella TD and Richardson BC. 1990. Quantitative changes in T cell DNA methylation occur during differentiation and ageing. Eur J Immunol. 20:1869-1872. Gold B, Radu D, Balanko A and Chiang CS. 2000. Diagnosis of Fragile X syndrome by Southern blot hybridization using a chemiluminescent probe: a laboratory protocol. Mol. Diagn. 5:169-178. Gonzalgo ML and Jones PA. 2002. Quantitative methylation analysis using methylationsensitive single-nucleotide primer extension (Ms-SNuPE). Methods. 27:128-133. Greco CM, Hagerman RJ, Tassone F, Chudley AE, Del Bigio MR, Jacquemont S, Leehey M and Hagerman PJ. 2002. Neuronal intranuclear inclusions in a new cerebellar tremor/ataxia syndrome among fragile X carriers. Brain. 125:1760-1771. Hagerman RJ and Hagerman PJ. 2002. The fragile X premutation: into the phenotypic fold. Curr. Opin. Genet. Dev. 12:278-283.  103  Hatakeyama C, Anderson CL, Beever CL, Penaherrera MS, Brown CJ and Robinson WP. 2004. The dynamics of X-inactivation skewing as women age. Clin. Genet. 66:327332. Hecimovic S, Barisic I, Muller A, Petkovic I, Buie I, Ligutic I and Pavelic K. 1997. Expand Long PCR for fragile X mutation detection. Clin. Genet. 52:147-154. Hundscheid RD, Braat DD, Kiemeney LA, Thomas CMG and Smits AP. 2001. Increased serum FSH in female fragile X premutation carriers with either regular menstrual cycles or on oral contraceptives. Hum. Reprod. 16:457-462. Hundscheid RD, Sistermans EA, Thomas CM, Braat DD, Straatman H, Kiemeney LA, Oostra BA and Smits AP. 2000. Imprinting effect in premature ovarian failure confined to paternally inherited fragile X premutations. Am. J. Hum. Genet. 66:413418. Hundscheid RD, Smits AP, Thomas CM, Kiemeney LA and Braat DD. 2003. Female carriers of fragile X premutations have no increased risk for additional diseases other than premature ovarian failure. Am. J. Med. Genet. A. 117:6-9. Issa JP, Ottaviano YL, Celano P, Hamilton SR, Davidson NE and Baylin SB. 1994. Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat. Genet. 7:536-540. Jin P, Zarnescu DC, Zhang F, Pearson CE, Lucchesi JC, Moses K and Warren ST. 2003. RNA-mediated neurodegeneration caused by the fragile X premutation rCGG repeats in Drosophila. Neuron. 39:739-747. Kenneson A, Cramer DW and Warren ST. 1997. Fragile X premutations are not a major cause of early menopause. Am. J. Hum. Genet. 61:1362-1369. Kenneson A, Zhang F, Hagedorn CH and Warren ST. 2001. Reduced FMRP and increased FMR1 transcription is proportionally associated with CGG repeat number in intermediate-length and premutation carriers. Hum. Mol. Genet. 10:1449-1454. Khalil A, Faghihi M, Modarresi C and Wahlestedt C 2007. FMR4: A novel primatespecific transcript silenced in Fragile X syndrome. Am. J. Hum. Genet. Annual Meeting supplement:36. Ladd PD, Smith LE, Rabaia NA, Moore JM, Georges SA, Hansen RS, Hagerman RJ, Tassone F, Tapscott SJ and Filippova GN. 2007. An antisense transcript spanning the CGG repeat region of FMR1 is upregulated in premutation carriers but silenced in full mutation individuals. Hum. Mol. Genet. 16:3174-3187. Larsen LA, Gronskov K, Norgaard-Pedersen B, Brondum-Nielsen K, Hasholt L and Vuust J. 1997. High-throughput analysis of fragile X (CGG)n alleles in the normal and premutation range by PCR amplification and automated capillary electrophoresis. Hum. Genet. 100:564-568. 104  Loesch DZ, Bui QM, Huggins RM, Mitchell RJ, Hagerman RJ and Tassone F. 2007. Transcript levels of the intermediate size or grey zone fragile X mental retardation 1 alleles are raised, and correlate with the number of CGG repeats. J. Med. Genet. 44:200-204. Machado-Ferreira Mdo C, Costa-Lima MA, Boy RT, Esteves GS and Pimentel MM. 2004. Premature ovarian failure and FRAXA premutation: Positive correlation in a Brazilian survey. Am. J. Med. Genet. A. 126:237-240. Mallolas J, Duran M, Sanchez A, Jimenez D, Castellvi-Bel S, Rife M and Mila M. 2001. Implications of the FMR1 gene in menopause: study of 147 Spanish women. Menopause. 8:106-110. Marozzi A, Vegetti W, Manfredini E, Tibiletti MG, Testa G, Crosignani PG, Ginelli E, Meneveri R and Dalpra L. 2000. Association between idiopathic premature ovarian failure and fragile X premutation. Hum. Reprod. 15:197-202. Murray A, Ennis S, MacSwiney F, Webb J and Morton NE. 2000. Reproductive and menstrual history of females with fragile X expansions. Eur. J. Hum. Genet. 8:247252. Murray A, Webb J, Grimley S, Conway G and Jacobs P. 1998. Studies of FRAXA and FRAXE in women with premature ovarian failure. J. Med. Genet. 35:637-640. Murray A, Webb J, MacSwiney F, Shipley EL, Morton NE and Conway GS. 1999. Serum concentrations of follicle stimulating hormone may predict premature ovarian failure in FRAXA premutation women. Hum. Reprod. 14:1217-1218. Nofin SL, Brown WT, Glicksman A, Houck GE,Jr, Gargano AD, Sullivan A, Biancalana V, Brondum-Nielsen K, Hjalgrim H, Holinski-Feder E, et al. 2003. Expansion of the fragile X CGG repeat in females with premutation or intermediate alleles. Am. J. Hum. Genet. 72:454-464. Oberle I, Rousseau F, Heitz D, Kretz C, Devys D, Hanauer A, Boue J, Bertheas MF and Mandel JL. 1991. Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome. Science. 252:1097-1102. Partington MW, Moore DY and Turner GM. 1996. Confirmation of early menopause in fragile X carriers. Am. J. Med. Genet. 64:370-372. Patsalis PC, Hettinger JA and Sismani C. 1999a. FMR1 repeat analysis in patients with ovarian dysfunction or failure. Am. J. Med. Genet. 83:329-330. Patsalis PC, Sismani C, Hettinger JA, Holden JJ, Lawson JS, Chalifoux M, Wing M, Walker M and Leggo J. 1999b. Frequencies of "grey-zone" and premutation-size FMR1 CGG-repeat alleles in patients with developmental disability in Cyprus and Canada. Am J Med Genet. 84:195-197.  105  Pieretti M, Zhang FP, Fu YH, Warren ST, Oostra BA, Caskey CT, Nelson DL. 1991. Absence of expression of the FMR-1 gene in fragile X syndrome. Cell. 66:817-822. Primerano B, Tassone F, Hagerman RJ, Hagerman P, Amaldi F and Bagni C. 2002. Reduced FMR1 mRNA translation efficiency in fragile X patients with premutations. RNA. 8:1482-1488. Rousseau F, Rouillard P, Morel ML, Khandjian EW and Morgan K. 1995. Prevalence of carriers of premutation-size alleles of the FMRI gene--and implications for the population genetics of the fragile X syndrome. Am. J. Hum. Genet. 57:1006-1018. Schwartz CE, Dean J, Howard-Peebles PN, Bugge M, Mikkelsen M, Tommerup N, Hull C, Hagerman R, Holden JJ and Stevenson RE. 1994. Obstetrical and gynecological complications in fragile X carriers: a multicenter study. Am. J. Med. Genet. 51:400402. Sherman, S.L., Taylor, K. and Allen, E.G. (2007) FMR1 premutation: a leading cause of inherited ovarian dysfunction. In Arrieta, I., Penagarikano, 0. and Telez, M. (eds) Nova Science Publishers, Hauppauge NY, 299-320. Snow K, Doud LK, Hagerman R, Pergolizzi RG, Erster SH and Thibodeau SN. 1993. Analysis of a CGG sequence at the FMR-1 locus in fragile X families and in the general population. Am. J. Hum. Genet. 53:1217-1228. Stoger R, Kajimura TM, Brown WT and Laird CD. 1997. Epigenetic variation illustrated by DNA methylation patterns of the fragile-X gene FMR1. Hum. Mol. Genet. 6:1791-1801. Sullivan AK, Marcus M, Epstein MP, Allen EG, Anido AE, Paquin JJ, Yadav-Shah M and Sherman SL. 2005. Association of FMR1 repeat size with ovarian dysfunction. Hum. Reprod. 20:402-412. Syrrou M, Georgiou I, Patsalis PC, Bouba I, Adonakis G and Pagoulatos GN. 1999. Fragile X premutations and (TA)n estrogen receptor polymorphism in women with ovarian dysfunction. Am. J. Med. Genet. 84:306-308. Tassone F, Hagerman RJ, Loesch DZ, Lachiewicz A, Taylor AK and Hagerman PJ. 2000a. Fragile X males with unmethylated, full mutation trinucleotide repeat expansions have elevated levels of FMR1 messenger RNA. Am. J. Med. Genet. 94:232-236. Tassone F, Hagerman RJ, Taylor AK, Gane LW, Godfrey TE and Hagerman PJ. 2000b. Elevated levels of FMR1 mRNA in carrier males: a new mechanism of involvement in the fragile-X syndrome. Am. J. Hum. Genet. 66:6-15. Tassone F, Hagerman RJ, Taylor AK, Mills JB, Harris SW, Gane LW and Hagerman PJ. 2000c. Clinical involvement and protein expression in individuals with the FMR1 premutation. Am. J. Med. Genet. 91:144-152.  106  Uzielli ML, Guarducci S, Lapi E, Cecconi A, Ricci U, Ricotti G, Biondi C, Scarselli B, Vieri F, Scarnato P, et al. 1999. Premature ovarian failure (POF) and fragile X premutation females: from POF to to fragile X carrier identification, from fragile X carrier diagnosis to POF association data. Am. J. Med. Genet. 84:300-303. Vianna-Morgante AM. 1999. Twinning and premature ovarian failure in premutation fragile X carriers. Am. J. Med. Genet. 83:326. Vianna-Morgante AM and Costa SS. 2000. Premature ovarian failure is associated with maternally and paternally inherited premutation in Brazilian families with fragile X. Am. J. Hum. Genet. 67:254-5; author reply 256-8. Welt CK, Smith PC and Taylor AE. 2004. Evidence of early ovarian aging in fragile X premutation carriers. J. Clin. Endocrinol. Metab. 89:4569-4574. Willemsen R, Hoogeveen-Westerveld M, Reis S, Holstege J, Severijnen LA, Nieuwenhuizen IM, Schrier M, van Unen L, Tassone F, Hoogeveen AT, et al. 2003. The FMR1 CGG repeat mouse displays ubiquitin-positive intranuclear neuronal inclusions; implications for the cerebellar tremor/ataxia syndrome. Hum. Mol. Genet. 12:949-959.  107  Chapter 3: Survey of hormone-related gene variants in POF 2 3.1 Introduction Premature ovarian failure (POF [MIM 311360]) is marked by the cessation of ovulation and menstruation caused by an absence of folliculogenesis. Possible physiologic causes include: 1) poor development of the ovaries resulting in a small follicular pool size established during fetal development, 2) increased rate of follicular atresia throughout life resulting in prematurely reduced follicular pool (Schlessinger et al. 2002) and 3) abnormal hormonal regulation resulting in dysfunctional follicle maturation (Simpson and Rajkovic. 1999). The only gene that has been consistently associated with POF is the Fragile X Mental Retardation 1 gene (FMRI [MIM 309550]) at chromosome Xq27.3. However, FMRI expansions account for only a portion of familial POF cases, and there is incomplete penetrance of the POF phenotype in carriers of FMRI premutations (see Chapter 2). Therefore, there must be other genetic factors that modulate POF risk in women carrying FMRI premutation alleles and are associated with POF in women who carry normal FMRI alleles. Polymorphisms in hormone receptor and binding protein genes may affect reproductive hormone function and are therefore intriguing candidate genes for POF. Disruption or even slight attenuation of hormone systems necessary for initiating follicle development, maturation, or ovulation, could contribute to altered follicular pool dynamics resulting in early menopause. Variations in hormone related genes could  2 A version of this chapter has been published: Bretherick KL, Hanna CW, Currie LM, Fluker MR, Hammond GL, Robinson WP (2008). Estrogen receptor a gene polymorphisms are associated with idiopathic premature ovarian failure. Fertil Steril 89(2):318-324. This chapter also includes ESRI reporter gene assay results that were not included in this publication.  108  contribute to POF at both the fetal stage, by influencing the establishment of initial follicular pool size, and in adults, by influencing rate of follicle recruitment or regulation of follicle maturation. Follicle stimulating hormone (FSH) is secreted from the pituitary and has a key role in the cyclic regulation of follicular development. FSH specifically affects granulosa cells, stimulating maturation of developing follicles. It is regulated by negative feedback from rising levels of both estradiol and inhibin B, a process that is critical in ensuring maturation of a single dominant follicle. FSH exerts its effects by binding to the FSH receptor, and a reduction in the quantity or function of this receptor would therefore result in failure to initiate follicle maturation and properly regulate this process. A variety of inactivating mutations in the FSH receptor gene (FSHR [MIM 136435]), have been reported in isolated cases of POF (Aittomaki et al. 1995; Allen et al. 2003; Beau et al. 1998; Doherty et al. 2002; Meduri et al. 2003; Touraine et al. 1999). However, they do not appear to be common causes of POF in the patient population (Sundblad et al. 2004). Common polymorphisms in FSHR may have a slight effects on gene expression or receptor function and may alter the rate of follicular recruitment, a possible cause of POF. Three single nucleotide polymorphisms (SNPs) in FSHR have been described. The  FSHR -29A/G SNP [rs1394205] in the promoter region disrupts a transcription factor binding site and alters FSHR expression (Nakayama et al. 2006; Simoni et al. 2002). Two missense SNPs in exon 10, FSHR 919A/G [rs6165] and FSHR 2039A/G [rs6166], alter the coding sequence but have not been proven to affect receptor function (Simoni et al. 2002). These SNPs are in strong linkage disequilibrium and together have been  109  implicated in a number of human diseases (Table 3.1) suggesting that at least some polymorphism in this region has functional consequences. Estrogen, specifically estradio1-1713 (E2), is produced by ovarian follicles and regulates the expression of a number of genes that support the action of FSH in folliculogenesis (Britt and Findlay. 2002). Estrogen acts through estrogen receptor a (ERa) at the hypothalamus-pituitary-ovarian (HPO) axis to regulate gonadotropin release (Kolibianakis et al. 2005). ERa is a nuclear transcription factor encoded by the ESRI gene [MIM 133430] that is responsible for transmitting circulating estrogenic compounds into genomic actions. ERa knock out (ERKO) mice are acyclic and infertile due to a block in folliculogenesis (Korach. 1994). They have an obstruction in the proliferation of granulosa cells and do not develop past the antral follicle stage (Couse and Korach. 1999). ESRI has a polymorphic (TA)„ repeat in the promoter that is in linkage disequilibrium with two SNPs in intron 1, -397C/T [rs2234693] and -351A/G [rs9340799] (Becherini et al. 2000). Although the functional effects of these polymorphisms have not been confirmed, they have been associated with a number of clinical conditions (Table 3.2) suggesting that they, or other genetic variants in linkage disequilibrium with them, do have physiological significance. Specifically, the haplotype composed of short (TA)„ repeats, -397T, and -351A alleles has been associated with conditions often correlated with low estrogen, whereas the haplotype consisting of long (TA) n repeats, -397C, and -351G alleles is associated with conditions correlated with high estrogen, suggesting that the latter confers a more active promoter. Estrogen also acts through a second receptor, estrogen receptor 13 (ER-(3), in the ovary to enhance follicular development (Kolibianakis et al. 2005). ER0 knockout  110  (BERKO) mice do develop antral follicles and ovulate, although they have small ovaries and reduced litter sizes that appear to be the result of abnormal follicle differentiation (Britt and Findlay. 2002). Elevated estrogen levels, such as those that occur from exposure to synthetic estrogenic compounds, cause development of follicles containing more than one oocyte, a consequence of altered ovarian differentiation mediated through ERf3 (Jefferson et al. 2007). It appears that whereas ERa functions in enhancing proliferation of granulosa cells, ERP functions in regulating follicular differentiation. The ER(3 gene, ESR2 [MIM 601663] contains a polymorphic intronic CA repeat that has been associated with both clinical phenotypes and physiological indicators of hormonal axis dysfunction (Table 3.1). Specifically, since estogen is known to stimulate sex hormone binding globulin (SHBG) production, the finding that women with short CA repeats have higher levels of testosterone and lower levels of SHBG than women with long repeats, suggests that short repeats are associated with decreased receptor activity (Westberg et al. 2001). Although androgens are generally considered male hormones, they have an important role in follicle development and also function as precursors to estrogen formation in the ovary (Kimura et al. 2007). The ability of androgens to exert their effects relies on the expression and function of the androgen receptor, a nuclear sex-  hormone receptor encoded by the AR gene [MIM 313700]. The androgen receptor acts as a transcription factor for a number of genes, including several crucial factors for folliculogenesis (Kimura et al. 2007). AR is expressed in human thecal cells (Hone et al. 1992), and has been shown in primate studies to stimulate primordial follicle growth (Vendola et al. 1999) and increase the number of primary, secondary and early pre-antral  111  follicles (Vendola et al. 1998). Androgen receptor deficient mice have lower follicle counts, aberrant ovarian gene expression, small litter sizes, and eventually develop POF (Shiina et al. 2006). The AR gene contains a translated, highly polymorphic CAG repeat that has been inversely correlated with receptor activity (Chamberlain et al. 1994; Tut et al. 1997). The CAG repeat results in an expanded polyglutamine tract that is assumed to affect interactions between AR and its coactivators (Irvine et al. 2000). This functional polymorphism also appears to be physiologically relevant in human disease (Table 3.1). Sex hormone binding globulin (SHBG) binds circulating androgens and estrogens, transports them to target tissues, and regulates their tissue-availability (Hammond. 2002). Increased plasma SHBG reduces free sex steroid levels, effectively creating physiologic conditions of low estrogens and androgens which could alter initiation of follicle development and maturation through the mechanisms described above. SHBG is encoded by the SHBG gene [MIM 182205] and a polymorphic pentanucleotide (TAAAA) repeat in promoter region has been described. The SHBG (TAAAA) 6 allele was correlated with reduced SHBG transcriptional activity in in-vitro reporter assays (Hogeveen et al. 2001). In population studies, however, the SHBG (TAAAA)6 allele has been associated with both increased (Cousin et al. 2004; Xita et al. 2003) and decreased (Haiman et al. 2005) circulating SHBG levels (Table 3.1). This finding may be explained by competing effects of the (TAAAA)6 microsatellite which decreases transcription activity, and an exon 8 missense SNP which influences the rate of SHBG clearance (Haiman et al. 2005). In this study I examined allele frequencies for repeat polymorphisms and/or SNPs at the FSHR, ESR1, ESR2, AR, and SHBG genes in a population of women with POF and  112  control women from western Canada. Furthermore, I followed up a positive finding with functional studies to assess the consequences of the associated polymorphism. 3.2 Methods 3.2.1 Samples  Women experiencing idiopathic secondary amenorrhea (N=54) were ascertained from the POF clinic at the Women's Health Centre of British Columbia. POF diagnosis was made based on the absence of menses for at least 3 months and two serum FSH results of >40 mIU/mL obtained more than one month apart, prior to age 40. All POF patients included in this study were normal on routine diagnostic karyotype, and had no known environmental cause (radiation, chemotherapy) for ovarian failure. Since only —20% of fragile X premutation carriers develop POF and it is possible that other genetic factors such as those examined here contribute to ovarian failure in premutation carriers FMR1 premutation carriers were not excluded. Two of 54 POF patients had an FMRI premutation; sizes were approximately 60 and 80 repeats. Average age at time of blood draw was known for 37 of the POF women and was 35.4 yrs (range 21-50). Two control groups were used in this study. Control group 1 was used to determine allele frequencies in the general population; this group was not selected for reproductive history so that allele frequencies in the POF patient group can be compared to expected allele frequencies. Control group 2 was used to determine allele frequencies in a group of women at the opposite end of the spectrum of reproductive capacity from the POF patients. The use of this more stringent control group enables a more sensitive measure of whether of not the polymorphisms studied have any bearing on ovarian function. Control group 1 (N=107) consisted of locally obtained anonymous healthy 113  donors (n=23) and unaffected spouses from families with Huntington disease (n=84); the average age of women in this group at the time of blood draw was 35.0 yrs (range 17-45). Although the reproductive history of these individuals was unknown, the group should be unbiased in terms of fertility, and should represent allele frequencies in the general population. As only 1% of women are expected to experience POF, the possible inclusion of such women would have negligible effects on allele distribution in this control group. Control group 2 (n=27) consisted of women who had not experienced any pregnancy losses and had had a healthy naturally-conceived pregnancy after the age of 37. The average age of women in this group at the time of blood draw was 41.4 yrs (range 38-58). The collection of samples for this study was approved by the University of British Columbia Clinical Research Ethics Board (CREB), approval number CO1-0460 (Appendix 1), and was done with patient consent. Samples were collected from a predominantly Caucasian population from western Canada. However, there is significant Asian (primarily Chinese) admixture, and some individuals are of mixed ancestry, therefore ethnic stratification was a concern. To provide an unbiased estimate of the ethnic origin of the patient and control populations, two polymorphic microsatellite markers, EGFR and D13S317, which show substantially different allele distributions between Chinese and Caucasian populations (Table 3.3) and are expected to be unrelated to disease status, were genotyped in patients and controls. There is no significant difference in allele frequencies between POF patient and control groups, at the alleles reported in the literature to show large differences in allele frequency between Caucasian and Asian populations (Table 3.3)  114  3.2.2 Genotyping Microsatellite markers were amplified by PCR, using fluorescently labelled primers from previously published primer sequences (Table 3.4). In general, PCR reactions for microsatellite analysis were performed with lx PCR buffer (Rose Scientific), —0.2 mM each dNTP (Invitrogen), 0.3-0.6 uM each forward and reverse primer, 0.2 U Taq DNA polymerase (Rose Scientific), and 5Ong genomic DNA in a lOuL reaction. Cycling was performed on a PTC-200 Peltier Thermal Cycler (MJ Research) with the following conditions: initial denaturation at 95°C for 2 minutes, followed by 30 cycles of 95°C for 30 seconds, 55-65°C (depending on primer set) for 45 seconds, and 72°C for 90 seconds, and completed with extension at 72°C for 7 minutes. luL of each PCR product was combined with 0.1 uL ROX size standard (Applied Biosystems) and 9 uL HiDi Formamide (Applied Biosystems) and separated in POP-4 polymer on an ABI310 Genetic Analyzer (Applied Biosystems). Fragment sizes were determined using GeneScan software (Applied Biosystems, Foster City, USA). Control samples were run with every PCR experiment to ensure consistency of allele calling between PCR batches. Single nucleotide polymorphisms were genotyped by TaqMan allelic discrimination assay (Figure 3.1) using primers and probes from previously published sequences (Table 3.4). For each sample —50ng of genomic DNA was combined with lx TaqMan Universal Master Mix (Applied Biosystems), 1.0 uM each forward and reverse primer (Applied Biosystems), 0.2 uM each VIC or FAM labelled MGB probe (Applied Biosystems). Alleles were assigned using ABI7000 sequence detection software v1.2. An example of allele calling using this method is shown in Figure 3.1. Two samples  115  homozygous for each allele, two heterozygous samples, and 2 blanks were run as controls with every allelic discrimination experiment.  3.2.3 X chromosome inactivation skewing The degree of X chromosome inactivation skewing was determined by assaying methylation at the Androgen Receptor (AR) locus as described in chapter 2. X chromosome inactivation ratios were expressed in terms of the percent inactivation of the smaller AR allele. AR repeat size was weighted by the degree of XCI skewing to determine genotype repeat size (GRS) (Allen et al. 2004) which reflects relative activity of both alleles: GRS = (Larger AR allele*XCI ratio) + (Smaller AR allele*(1-XCI ratio)), rounded to the nearest whole number. For samples that were uninformative for skewing, since both alleles were approximately the same size, GRS was calculated with an XCI ratio of 50%.  3.2.4 ESR1 luciferase assay Potential regulatory effects of the polymorphic (TA) n repeat in the ESR1 promoter were assessed by luciferase reporter assay. ESR1 promoter fragments containing either a short (TA)14, or long (TA) 22 , repeat were each cloned into a luciferase reporter plasmid, transiently transfected into MCF-7 and Ishikawa cell lines expressing ESR1, and the level of luciferase expression after 48 hours was assayed. The expression of the luciferase gene reflects the influence of repeat length on ESR1 promoter activity. Fragments of the ESR1 promoter (Figure 3.2) containing either (TA)14 or (TA)22 were cloned into a promoter-less pGL2-Basic plasmid vector (Promega). Three different fragment sizes (referred to as 801bp, 1346bp, and 1512bp in reference to the fragment  116  length in alleles with 14 TA repeats) were assessed. The 801bp fragment was amplified directly from genomic DNA of control samples that were homozygous for either the 14 or 22 repeat allele using conventional PCR (Invitrogen) and primers -1297F 5'-GTT GGT GTT TGG GAT AGC A-3' and -249R 5'-CAG GGA AGA CTG GGC TTA AAA3'. 25uL each PCR product and 2uL pGL2-Basic plasmid (Promega) were digested overnight at 37°C with lx React 2 buffer, and 10U each XhoI and BglII (Invitrogen). 4uL of digested insert and 2uL of digested vector were ligated overnight at 4°C with lx T4 DNA ligase buffer and 1U T4 DNA ligase (Invitrogen) in a 20uL reaction. DH5a competent E. coli (Invitrogen) were transformed with 10uL of the ligation reaction by heat shocking the cells with 25 minutes on ice, 30 seconds at 42°C and 2 minutes on ice. Transformed cells were grown at 37°C for 1 hour in 900uL LB broth, pelleted, resuspended in 100uL LB, plated on agarose-ampicillin plates and incubated overnight at 37°C. Plasmid minipreps were performed on 6-8 colonies selected from each plate that had been grown in 2.5mL LB overnight at 37°C. Clones containing the correct 801bp fragment with either (TA) 14 or (TA) 22 were confirmed by digestion and fragment analysis on polyacrylamide gels and then grown in large batches for maxiprep plasmid extraction. PGL-2 basic vectors carrying longer promoter fragments were created by PCR amplification of genomic DNA using primers -780F 5'-CCA AGG TTT TTC TGA ATC ATC C-3', and +52R 5'-GAC CCG ACG GAG CAA GTG CAG CTC CC-3' or +215R 5'-AGG GTG CAG ACC GTG TCC-3' for 1346bp and 1512bp respectively. These fragments were then digested with EcoRl and HindIll, and ligated with T4 DNA ligase to PGL2 basic vector digested with XhoI and HindIII, and fragments of the existing 801bp plasmid digested with XhoI and EcoRl. Positive colonies carrying PGL-2 basic vectors  117  with the correct 1346bp and 1512bp promoter fragments with either (TA) 14 or (TA)22 were selected as described for the 801bp vector and grown in large batches for maxiprep extraction. Vectors containing the 801bp, 1346bp, or 1512bp fragments with either the (TA) 14 or (TA)22 repeats were transiently transfected into MCF-7 or Ishikawa cells. Both MCF-7 and Ishikawa cells were routinely maintained in MEM-a media (GIBCO) supplemented with 10% FBS, 100U/mL penicillin, and 10Oug/mL streptomycin (GIBCO). Cells were grown to 40-60% confluence in 3cm wells with 2mL media and co-transfected with 1.2ug of vector and 0.2ug of pRCMV-lacZ control plasmid with either HiPerfect Transfection Reagent (Qiagen) or LipofectAMINE reagent (Invitrogen) according to manufacturers' instructions. From the point of transfection, cells were grown in phenol-red free MEM-a media with 10% delipidated FBS (Cocalico Biologicals) with the addition of 100uM ethinyl estradiol (EE2) diluted in ethanol where indicated. 0.2ug of ERa containing plasmid was co-transfected with the described pGL2basic vectors where indicated, to ensure sufficient intracellular ESR1 expression. Experiments were run in parallel with a vector containing several estrogen response elements (EREs) as a control to ensure cells were able to respond appropriately to estrogen. After 48 hours of growth cells had reached approximately 100% confluence and were harvested and the level of luciferase and LacZ expression was assayed. Cells were washed twice with cold phosphate buffered saline (PBS) and harvested by scraping into lmL cold PBS. Cells were pelleted, PBS aspirated, and cells resuspended into 100uL of 0.25M Tris-HC1(pH=7.8), and lysed by 2 cycles of freezing for 10 minutes in an ethanol-  118  dry ice bath and thawing for 10 minutes at 37°C. Cell debris was removed by centrifugation and supernatant was collected and used to assay luciferase and 13galactosidase activity. 50uL of supernatant was combined with 100uL of luciferase assay reagent (Promega) and light units emitted were assayed on a Lumat LB 9507 luminometer (EG&G Berthold). 20uL of supernatant was diluted with 30uL of water and combined with 50uL of ortho-nitropheny1(3-galactoside (ONPG) P-galactosidase reagent. Colour was developed for 20-60 minutes at 37°C, stopped with addition of 150uL HNO3, and intensity was measured by a 0.1s absorbancy reading at 405 nm on a Victor 3V 1420 multi-label counter (Perkin-Elmer). To correct for transfection efficiency, light units from the luciferase assay were normalized by dividing the OD reading from the  13-  galactosidase assay. Each experiment was performed in triplicate and an average of three normalized values, referred to as 'relative luciferase expression', was compared for all analyses. Slight inconsistencies in time required to assay luciferase and develop 13galactosidase between different sets of experiments can affect absolute luciferase and (3galactosidase readings, therefore results are only comparable for experiments run in parallel and cannot be compared to results from experiments run on different days.  3.2.5 ESR1 mRNA quantification To confirm endogenous ESR1 expression in MCF-7 and Ishikawa cell lines, total mRNA was extracted from cultured cells and ESR1 expression was assayed by real-time quantitative PCR (qPCR). Cells at 100% confluence were harvested by trypsinization, pelleted and re-suspended in PBS. RNA was extracted with the RNeasy mini Kit (Qiagen) according to manufacturer's instructions and converted to cDNA with the Quantitect Reverse Transcription Kit (Qiagen) as described in Chapter 2. cDNA was  119  quantified by qPCR as described in Chapter 2 using pre-designed TaqMan Gene Expression Assays for ESRI and /3-actin (Applied Biosystems). ESRI and /3-actin standard curves consisting of serial dilutions of 500ng, 10Ong, 2Ong, and 4ng were run on each plate. Both ESRI and /.3-actin qPCR amplifications were performed in triplicate.  3.2.6 Statistical analysis Overall allele distributions between POF patients and Control groups were compared using two-tailed x2 tests. Alleles were also grouped by division, based on mean allele size in Control group 1, and compared using two-tailed x2 tests. To account for multiple tests performed a Bonferroni correction was applied. Since association was assessed at nine separate loci, the Bonferroni corrected p value of =0.0055 was used as an indicator of significance for comparisons of overall allele distribution and allele frequencies using short and long cut-offs only. For reporter gene assay comparisons of relative luciferase expression one-tailed ttest was used because of the a priori hypotheses that addition of exogenous estrogen would increase promoter activity and that the (TA)22 allele confers a more active promoter than the (TA)14 repeat.  3.3 Results 3.3.1 ESR2, AR, SHBG, and FSHR are not associated with POF There was no significant difference in overall allele distribution at the ESR2 (Figure 3.3), AR (Figure 3.4a), or SHBG (Figure 3.5) microsatellite polymorphisms between POF patients and either control group 1, control group 2, or combined controls. Furthermore, when using a cut off based on median repeat size in control group 1 to lump  120  AR and ESR2 alleles into categories of either short or long alleles, there was no difference in allele distribution between patient and control groups at any of these loci (Table 3.5). The AR gene is on the X chromosome and any affect of repeat size at this locus may therefore be modulated by degree of X chromosome inactivation skewing. I therefore weighted repeat size at AR by X chromosome inactivation ratio (for control group 1 only, as X chromosome inactivation ratio is influenced by age and control group 2 has an older mean age), to determine an AR genotype repeat size for each individual. There was no difference in overall distribution of AR genotype repeat size (Figure 3.4b) or frequency of high and low genotype repeat sizes (Table 3.5) between the POF patients and control group 1. At SHBG there was no difference in overall allele distribution between POF patients and combined control groups (Figure 3.5). Furthermore there was also no difference in either the frequency of (TAAAA)6 alleles or homozygosity for the (TAAAA)6 genotype between the POF population and either control group or combined controls (Table 3.5). Genotype distribution for the FSHR SNPs -29A/G, 919A/G, and 2039A/G in POF patients, and combined controls are shown in Figure 3.5. There was no difference in allele or genotype frequency for any of these SNPS between POF patients and either control group 1, control group 2, or combined controls (Table 3.5).  3.3.2 ESR1 is associated with POF The allele distribution for the ESR1 (TA) 6 repeat in POF patients and combined control groups 1 and 2 is shown in Figure 3.7. There is a difference in the overall allele distribution between the POF population and the combined control groups (p=0.05, z2 contingency test) however this is not statistically significant at the Bonferroni corrected significance level of a=0.0055. In both POF patients and controls there is a bimodal  121  distribution of ESR1 (TA)„ repeat allele frequencies. The low point of this distribution was used as a cut off to divide alleles into categories of either short (S<18 TA repeats) or long alleles (L>18 TA repeats). Short alleles were significantly less common in the POF patient population than in both control group 1, and control group 2, and combined controls (p=0.002, two-tailed x2 test vs. combined controls, Table 3.5). Genotypes consisting of two short alleles (SS) were noticeably rare in the POF patient population; found in only 3 of 54 POF patients (5.5%), but in 40 out of 105 women in control group 1(38.1%), and 8 out of 27 women in control group 2 (29.6%) (p<0.0001, two-tailed x2 test vs. combined controls, Table 3.5). The ESR1 (TA)„ repeat is tightly linked to two downstream SNPs, -397C/T [rs2234693] and -351A/G [rs9340799] (Becherini et al. 2000), which were also genotyped and found to be similarly associated with POF (Table 3.5). Parental DNA on POF patients and controls was not available, and therefore individual haplotypes could not be determined. However, individual genotypes at these three ESR1 loci were used to estimate haplotype frequencies in the patient and control groups through the use of an iterative procedure (Hill. 1974). Estimated haplotype frequencies in the POF patient and control groups are shown in Figure 3.8. The STA haplotype consisting of a short (TA)„ repeat and the -397T and -351A variants, was less common in the POF patient population than in either control group, whereas every other haplotype was present at an increased frequency. 3.3.3 Mode of inheritance for ESR1 associated POF risk  To test whether the genotype distribution in patients was compatible with a simple dominant or recessive model I followed the approach of Wittke-Thompson et al. 2005  122  (Wittke-Thompson et al. 2005). For the ESRI (TA) n genotypes in the POF population, Hardy-Weinberg is rejected (p<0.001, x2 contingency test, table 3.6) and the difference between the observed and expected allele frequencies is A = 0.056 - 0.181 = -0.125. A negative value of A is not compatible with a recessive mode of inheritance (WittkeThompson et al. 2005), I therefore assessed whether the data fit a simple dominant model. In a dominant model the relative risk for a homozygous carrier of a risk allele, y, is equal to the relative risk of a heterozygote carrier, fl. A value of y=/3=9.8 provides a good fit for the data (Table 3.6) although the confidence interval for this estimate is quite large (95% CI = 2.7 — 36.2). Therefore the risk of POF for a carrier of a susceptibility allele, in this case a long (TA)„ repeat at ESR1, is approximately ten times that of a woman homozygous for the short (TA) n repeat. Furthermore, using this dominant model and assuming a population prevalence of POF of 1.0%, the relative risk of POF in the non-susceptible genotype, a, can be determined, a=0.0015. Thus the estimated risk of POF in individuals homozygous for short (TA),, repeats is 1/667 whereas the risk of POF in carriers of a long (TA)„ repeat is 1/68.  3.3.4 Interaction between ESR1 and FMR1 The ESRI susceptibility haplotype may confer risk independently of FMRI and be responsible for POF in women who do not carry FMR1 premutations, or alternatively, it may interact with the FMRI gene to compound the risk for POF in women who carry FMRI susceptibility alleles. I therefore examined the distribution of ESRI (TA) n repeat genotypes (divided into short and long alleles) in POF patients that carry a long FMRI repeat allele (Figure 3.9). A cut off of >35 repeats was used to define long FMRI repeats, based on previous findings that FMRI allele sizes >35 repeats are increased in  123  the POF patient population as compared to controls (Bretherick et al. 2005). Assuming the simple dominant model determined above, in which ESR1 LL genotypes are at the same risk of SL genotypes, there is no difference in distribution of ESR1 risk genotypes between POF patients with or without long FMR1 alleles (p=0.1, two-tailed Fisher's exact test). POF patients with long FMR1 alleles were more likely to be homozygous for ESR1 risk alleles (LL), than those not carrying a long FMR1 allele (p=0.001, two tailed Fisher's exact test). There is no difference in ESR1 (TA). repeat genotype distribution between control women with or without a long FMRI allele (p=0.62, x2 contingency test).  3.3.5 Regulatory effects of ESR1 (TA). repeat Regulatory effects of the polymorphic (TA)„ repeat in the ESR1 promoter were assessed by luciferase reporter assay in MCF-7 and Ishikawa cell lines. Endogenous ESR1 expression in these cell lines was assessed by qPCR. Both cell lines express ESR1; MCF-7 cells express ESR1 approximately 3 times higher than Ishikawa cells. Three different sized promoter fragments were assessed (Figure 3.2). In MCF-7 cells, the 1346bp promoter fragment had nearly double the level of activity as the 801bp fragment (Figure 3.10a). However the promoter activity of both of these fragments was very low compared to that of a control containing a promoter region with several estrogen response elements (Figure 3.10b), suggesting they are unlikely to contain the critical functional regions of the ESR1 promoter. Furthermore, the 1346bp fragment showed no increase in activity with the addition of the exogenous estrogen, ethinyl estradiol (EE2) (Figure 3.10b), and therefore this region of the ESR1 promoter does not appear to be upregulated by ESR1 transcription factor binding. The 1512bp promoter fragment shows  124  higher relative luciferase expression in both MCF-7 and Ishikawa cells suggesting that it does likely contain the necessary elements required for promoter activity. Furthermore, there was an increase in relative luciferase expression of the 1512bp fragments in both MCF-7 and Ishikawa cells when exposed to exogenous EE2 (Figure 3.11). The increase is significant for both the (TA)14 repeat (p=0.003, one tailed t-test) and (TA)22 repeat (p=0.01 one tailed t-test) in Ishikawa cells, and only the (TA)22 repeat (p=0.0003, one tailed t-test) in MCF-7 cells in the data shown (Figure 3.11), but was poorly reproducible in repeat experiments. ESR1 promoter activity in MCF-7 and Ishikawa cells does not appear to be influenced by length of the (TA),, repeat. In the 801bp and 1346bp promoter fragments, the (TA)22 repeat allele had lower expression than the (TA)14 repeat (Figure 3.10a). However, this finding was not reproducible in repeat experiments and due to the very low promoter activity of these fragments, is unlikely to be of physiological significance. The 1512bp fragment carrying the (TA)22 repeat allele showed a increased relative luciferase expression compared to the (TA)14 repeat both with and without the addition of EE2 in both Ishikawa and MCF-7 cells (Figure 3.11). This finding was significant in MCF-7 cells (p=0.02 and p=0.05 one-tailed t-test for comparisons with and without EE2, respectively), but not in Ishikawa cells (p=0.2 and p=0.4 one-tailed t-test for comparisons with and without EE2, respectively). In addition, in multiple repeat experiments this trend was not significant.  3.4 Discussion In the present study I examined polymorphisms in several endocrine related genes for association to POF. Variants in the AR, ESR2, FSHR, and SHBG genes do not appear  125  to be associated with POF in this patient population. A lack of association suggests that either these genes do not have a significant role in POF pathogenesis, the polymorphisms studied are not functional or in linkage disequilibrium with a functional variant, or any alteration in POF risk due to these loci is so minor as to be undetectable in the POF and control populations studied. The (TA)„ repeat and SNPs at the ESR1 gene were associated with POF in this patient population. Short (TA)„ repeats occur significantly less often in the POF patient group than in control women; furthermore genotypes consisting of two short alleles are noticeably rare among women with POF. The (TA)„ repeat is linked to -397C/T and 351A/G SNPs in ESR1 intron 1, and long (TA)„ repeats occur most often as part of a 397C, and -351 G haplotype (LCG haplotype). The ESR1 (TA) n repeat appears to confer risk for POF in a simple dominant manner in which women carrying a long (TA)„ repeat allele have approximately 10 times the risk of POF compared to women homozygous for short ESR1 (TA)„ repeats. These results are supported by the findings of Weel et al. (1999), who report that women homozygous for the -397C allele have a 1.1 year earlier onset of menopause compared to those homozygous for -397T. However, they report an allele dose effect, in which each copy of the -397C allele corresponds to a 0.5 year earlier onset of menopause (Weel et al. 1999), whereas in this study the inheritance follows a simple dominant model, for which SL carriers are at the same risk of LL carriers. It is possible that the because of the relatively small effect of each risk allele and the small sample size an allele dosage affect of the ESR1 gene may have been obscured in the present study.  126  The present findings are contrary, however, to those of Syrrou et al. (1999) who report that familial POF patients, and FMR1 premutation carriers with POF have significantly shorter (TA) n repeat size than controls. There are several possible explanations for the discrepancies between the previous study and the results reported here. Firstly, Syrrou et al. examined difference between median (TA) n repeat length in the patient and control groups (Syrrou et al. 1999), whereas I report a difference between overall allele distribution and allele size divided into short and long repeats. As repeat size is bimodally distributed, dividing repeat size into short or long alleles is statistically stronger and may be more biologically relevant that examining median repeat length in the two groups. Secondly, the number of samples in the previous study was quite limited, with only 7 cases of familial POF, and 7 cases of FMR1 premutation carriers with POF, thus chance may account for the difference in results seen between the reports. Thirdly, the POF patient populations examined are quite different between the two studies. I report a significant decrease in short alleles in a patient population of women with largely sporadic idiopathic POF; however, Syrrou et al. saw no association in a group of 16 sporadic POF patients, but smaller median repeat size in both familial POF cases and FMR1 premutation carriers with POF (Syrrou et al. 1999). If long ESR1 alleles and FMR1 premutations are separate and independent causes of POF, it would be expected  that in carriers of FMR1 premutations in which POF is attributed to FMR1 genotype, there may be an increased frequency of short ESR1 TA repeats as this protective genotype will have no affect on a background of high FMR1 associated risk. Additional studies examining ESR1 (TA) n repeat length in large populations of both familial and sporadic POF patients are necessary to verify an association between long repeats and  127  sporadic POF, and to determine whether or not genotype at this locus interacts with other susceptibility loci, such as FMR1, to increase risk for POF. The interaction of the long ESRI (TA) n repeat alleles with long FMR1 repeat alleles is unclear. In this report POF patients with long FMRI alleles were more likely to be homozygous for ESRI risk alleles, than those not carrying a long FMR1 allele. These unexpected findings suggest that the ESRI (TA)„ may follow a recessive inheritance pattern when operating on a background of increased FMR1 risk. Due to the small sample size available in this study, there are very limited numbers of genotypes in each category, and differences in ESR1 genotype distribution observed between POF carriers of long and short alleles may be due to chance. Additional studies examining ESRI (TA). repeat length in large numbers of affected and unaffected carriers of long FMR1 repeats in women ascertained on the basis of a positive family history for Fragile X syndrome are necessary to fully understand the dynamics of ESRI (TA) n repeat associated POF risk. The finding of an association of POF with the ESRI (TA)„ repeat polymorphism and intron 1 SNPs suggests that either one of these variants or other closely linked polymorphisms in ESRI or a nearby gene have functional effects that lead to increased risk of POF. The (TA) n repeat itself could affect ESRI gene expression or function. Microsatellite repeat polymorphisms, even those situated in non-coding gene regions, may have functional roles in regulating gene expression by altering DNA secondary structure and influencing transcription factor binding (Comings. 1998). The ESRI gene promoter has a very complex genomic organization with multiple promoters and alternative splice sites resulting in expression of alternative first exons and different ERa  128  transcripts (Grandien et al. 1997; Kos et al. 2001). The (TA)„ repeat polymorphism is found in a 1.5Kb region between exons C and B (Figure 3.2). Although tissue specific promoter use remains largely unknown, exon C is expressed in endometrial and ovarian tissue (Flouriot et al. 1998), exon B usage has been observed estrogen receptor positive breast cancer cell lines (Thompson et al. 1997). (TA)„ dinucleotide repeat length may affect promoter usage resulting in unsuitable estrogen receptor a expression in certain tissues (Becherini et al. 2000). In addition, a regulatory enhancer element that may act as a steroid response element has been identified approximately 200 base pairs down stream of the (TA)„ repeat (Cohn et al. 1999). Although the role of this enhancer is not yet clear, its proximity to the polymorphic repeat region makes it a potential target for functional affects of (TA)„ repeat size. Reporter gene assessment of the ESR1 promoter region did not conclusively demonstrate a relationship between ESR1 (TA)„ repeat and gene expression. The 1512bp construct was the only fragment that showed sufficient expression of the reporter gene to be considered a promoter. The results shown (Figure 3.11 a) demonstrate a higher baseline and ligand-responsive activity for the fragment containing the (TA)22 repeat in MCF-7 (human breast adenocarcinoma) cells. However, this was not found in Ishikawa (endometrial adenocarcinoma) cells and was not reproducible in MCF-7 cells despite numerous repeat experiments. This suggests that any effect observed is not only cell line specific, but is also so minimal that it can not be significantly reproduced within the error of this assay system. That the (TA)22 repeat fragment shows higher expression than the (TA)14 repeat fragment even in the absence of EE2 suggests that it influences binding of transcription factors other than ERa itself. The finding of increased activity with the  129  (TA)22 repeat fragment, is supported by previous studies that have reported increased frequency of this allele in clinical conditions correlated with excess estrogen (Table 3.2).  ESR1 has several alternative transcription start sites, and promoter usage differs between tissue types (Kos et al. 2001). Therefore (TA)„ repeat length may have functional effects on ESR1 expression that have not been apparent in this assay but do provide the physiological explanation for the association between this gene region and POF. Alternatively, one of the SNPs in ESR1 intron 1 may be the functional variant responsible for the association with POF. The -397C allele has a potential transcription factor binding site for myb transcription factors that is abolished in the -397T variant (Herrington et al. 2002). An in vitro reporter assay has shown that the expression of the construct containing the C allele was enhanced more than 10 fold with co-transfection of a constitutively expressed B-myb vector, whereas there was no such increase in expression of a construct carrying the T allele (Herrington et al. 2002). The myb family transcription factors are important in cell cycle regulation, and the widely expressed Bmyb is critical for embryonic development in the mouse (Joaquin and Watson. 2003). Although it has not been proven that the presence or absence of -397C alters the ESR1 expression or function in vivo, it is possible that the -397C variant that is part of the LCG haplotype, is responsible for the association reported here. The up-regulation of expression seen in the -397C variant is consistent with the association of this allele with a higher prevalence of hysterectomy due to fibroids and menorrhagia in women (Weel et al. 1999), as these conditions can also result from increased estrogen. The ESR1 long (TA)„ repeats, -397C, and -351G alleles have been associated with conditions correlated with high estrogen, suggesting that this haplotype confers a more  130  active promoter. This may be due to influence of the (TA) n repeat, as is weakly supported by the ESRI reporter assay results in this paper, or the -397C allele, due to its influence on B-myb transcription factor binding, or a combination of both factors. Regardless, this study has demonstrated that the ESRI haplotype that may confer a more active ERa promoter is associated with significantly increased risk of POF. A more active ERa promoter may influence POF risk by increasing the rate of follicular atresia. Elevated estrogen responsiveness may support the recruitment of slightly larger cohorts of immature follicles in each menstrual cycle, the result being increased atresia of the excess follicles. This could cause a premature reduction of follicles below the critical level required for menstruation and ovulation. In conclusion, this study showed a highly significant association between ESRI genotype and risk for POF. This may be related to the affects of the -397C/T allele on myb-mediated transcription, affects of the (TA) n repeat on promoter usage, or affects of some other functional polymorphism. Further studies are necessary confirm these findings as well as to further clarify: 1) the mode of inheritance of the ESR1 risk alleles 2) the dynamics of the interactions between genotype at ESR1, FMR1 and other POF susceptibility genes and 3) the functional affects of the various ESRI polymorphisms to determine whether the -397C allele enhances ESRI expression in vivo, and whether there are indeed functional affects of the (TA) n repeat allele on tissue-specific promoter usage.  131  Table 3.1 Published phenotypic associations with FSHR, ESR2, AR and SHBG  Gene FSHR  Association hypertension serum estradiol serum FSH ovarian cancer polycystic ovary syndrome (PCOS) response to ovarian stimulation menstrual cycle length  Reference Nakayama et al. 2006 Nakayama et al. 2006 Sudo et al. 2002; Perez Mayorga et al. 2000; Greb et al. 2005; de Koning et al. 2006 Yang et al. 2006 Sudo et al. 2002 Sudo et al. 2002; Perez Mayorga et al. 2000 Greb et al. 2005  ERS2 bone mineral density (BMD)  Scariano et al. 2004 ; Ogawa et al. 2000b Westberg et al. 2004 Ashworth et al. 2005 Westberg et al. 2001 Geng et al. 2007 Ogawa et al. 2000a  AR^prostate cancer  Giovannucci et al. 1997 Giguere et al. 2001; Yu et al. 2000 Levine et al. 2001 Yaron et al. 2001 Hickey et al. 2002; Ibanez et al. 2003  SHBG age at menarche  Xita et al. 2005 Xita et al. 2003 Cousin et al. 2004 Ferk et al. 2007, Xita et al. 2003 Haiman et al. 2005 Eriksson et al. 2006  serum prolactin venous ulceration serum androgens and SHBG post menopausal osteoporosis systolic blood pressure  breast cancer ovarian cancer endometrial cancer PCOS and ovarian hyperandrogenism PCOS serum SHBG in hirsute women serum SHBG in women with PCOS postmenopausal serum SHBG serum androgens and BMD in men  132  Table 3.2 Published phenotypic associations with ESR1 polymorphisms Association Reference Weel et al. 1999 Sowers et al. 2006 Sundarrajan et al. 1999  Altmae et al. 2007 Georgiou et al. 1997 Gerhardt et al. 2006 Corbo et al. 2007 Molvarec et al. 2007 Kjaergaard et al. 2007 Kim et al. 2005 Georgiou et al. 1999 Hsieh et al. 2007  -397T and/or -351A and/or (TA)Short  i follicle #; loocyte #; lembryo #; .follicular size & 4, pregnancy rate in IVF patients I risk of unexplained infertility  I follicle:oocyte ratio after ovarian stimulation i risk of late fetal loss i spontaneous abortions I number of children I risk of severe pre-eclampsia I risk of breast cancer 1' risk of myocardial infarction I risk of mild endometriosis I risk of endometriosis  Weiderpass et al. 2000 Alevizaki et al. 2007 Herrington et al. 2002 Schuit et al. 2005 Becherini et al. 2000 van Meurs et al. 2003 Langdahl et al. 2000 Ioannidis et al. 2004 Khosla et al. 2004 Kitamura et al. 2007  -397C and/or -351 G and/or (TA)lon g earlier onset of menopause I advanced ovarian aging  I follicle # and oocyte # after controlled ovarian hyperstimulation  i risk of endometriosis i risk of leiomyoma 4. risk of endometrial cancer / triglycerides; i insulin; 4. # births 4, severity of coronary artery disease 4. E-selectin response to HRT  4, postmenopausal plasma E2 1, bone mineral density (BMD) i fracture risk 1 BMD; 1 bone area; 1 fracture risk I risk of osteoporotic fractures -  Riancho et al. 2006  I influence of aromatase genotype on bone mineral density (BMD)  Brandi et al. 1999 Westberg et al. 2003  I psychotic and non-conforming  4. risk of fracture 1' effect of E2 deficiency on BMD i effect of lean tissue mass on BMD  I risk of Alzheimer's disease  personality Guarducci et al. 2006  4 sperm production  133  Table 3.3 Published frequencies of EGFR and D13S317 alleles  EGFR (CA). This study This study Liu et al. 2003 Bielawski et al. 2005 Chi et al. 1992 Brandt et al. 2004 Liu et al. 2003 Liu et al. 2003  population  N alleles  Frequency of (CA) I 6 (%)  Frequency of (CA)20 (%)  Combined control groups POF patients Caucasians Polish Caucasian (CEPH) German Chinese Non-chinese Asians  242 88 366 360 142 2126 52 80  D13S317 (TATC). This study This study Budowle et al. 1999 Ross et al. 2001 Ross et al. 2001 Raczek et al. 2006 Chen et al. 2004 Chan et al. 2005 Gao et al. 2005  Combined control groups POF patients American (Caucasian) Austrian Croatian Polish Chinese (northwest China) Chinese (Hong Kong) Chinese (eastern China)  254 108 392 328 276 2308 240 650 200  40.5 48.9 43.2 42.5 42.0 66.6 5.8 25.0 Frequency of (TATC)8 (%) 9.8 15.7 9.9 14.0 14.0 13.9 22.6 31.2 22.0  19.0 20.4 21.0 17.8 26.0 34.5 65.4 61.3 Frequency of (TATC)118,12 (%) 60.0 57.0 62.8 63.1 61.2 60.4 44.2 37.4 46.5  134  Table 3.4 Primer sequences used in genotyping Primer EGFR^F  R  D13S317^F  R AR^F R ESR1 (TA).^F R ESR1 SNPs^F R -397T probe -397C probe -351A probe -351G probe ESR2 (CA).^F R SHBG^F R FSHR -29A/G^F R -29A probe -29G probe FSHR 919A/G^F R 919A probe 919G probe FSHR 2039 A/G F R 2039A probe 2039G probe I  Primer sequence (5'—>3') HEX-GTTTGAAGAATTTGAGCCAACC TTCTGCACAC'TTGGCAC HEX-ACAGAAGTCTGGGATGTGGA GCCCAAAAAGACAGACAGAA HEX-GCTGTGAAGGTTGCTGTTCCTCAT TCCACAATCTGTTCCAGAGCGTGC HEX-GACGCATGATATACTTCACC GCAGAATCAAATATCCAGATG TCCATCAGTTCATCTGAGTTCCAA TTCAGAACCATTAGAGACCAATGCT FAM-TGTCCCAGCTGTTTT VIC-CCCAGCCGTTTTA FAM-CCCAACTCTAGACCAC VIC-TCCCAACTCCAGACCA HEX-GGTAAACCATGGTCTGTACC AACAAAATGTTGAATGAGTGG HEX-GCTTGAACTCGAGAGGCAG CAGGGCCTAAACAGTCTAGCAGT AGCTTCTGAGATCTGTGGAGGTTT AATTATGCATCCATCCACCTGATT VIC-TGCAAATGCAGAAGG FAM-TGCAAATGCAGGAGG CTTCATCCAATTTGCAACAAATCTAT TGTCTTCTGCCAGAGAGGATCTC FAM-ATTATATGACTCAGACTAGG VIC-TTATATGACTCAGGCTAGG AAGGAATGGCCACTGCTCTTC GGGCTAAATGACTTAGAGGGACAA VIC-AGTCACCAATGGTTC FAM-AGAGTCACCAGTGGTT  reference Chi et al. 1992 GDB I Allen et al. 1992 Sano et al. 1995 Koch et al. 2005  Tsukamoto et al. 1998 Hogeveen et al. 2001 Ahda et al. 2005  Ahda et al. 2005  Ahda et al. 2005  The GDB Human Genome database  135  Table 33 Hormone receptor allele frequencies in POF patients and controls AR (CAG). N alleles <21 repeats GRS <21' ESR1 (TA). N alleles <18 repeats (S) SS genotype ESR1-397C/T N alleles -397T allele TT genotype ESRI-351A/G N alleles -351A allele AA genotype ESR2 (CA)„ N alleles <23 repeats SHBG (TAAAA). N alleles 6 repeats 6/6 genotype FSHR -29A/G N alleles -29A allele FSHR 919A/G N alleles 919A allele FSHR 2039A/G N alleles 2039A allele  p values  POF patients  Control group 1  Control group 2  108 47 (43.5%) 22 (40.7%)  214 107 (50.0%) 46 (43.0%)  54 26 (48.2%)  108 46 (42.6%) 3 (5.6%)  210 127 (60.5%) 40 (38.1%)  54 31 (57.4%) 8 (29.6%)  0.002, 0.08, 0.002 2 <0.0001, 0.005, <0.0001 2  108 45 (41.7%) 3 (5.6%)  212 126 (59.4%) 39 (36.8%)  54 32 (59.3%) 9 (33.3%)  0.003, 0.043, 0.002 2 <0.0001, 0.002, <0.0001 2  108 53 (49.1%) 7 (13.0%)  212 145 (68.4%) 49 (46.2%)  54 38 (70.3%) 13 (48.2%)  0.007, 0.01, 0.0003 2 <0.0001, 0.001, <0.0001 2  90 45 (50.0%)  54 32 (58.5%)  108 23 (21.3%) 3 (5.5%)  208 55 (26.4%) 3 (2.9%)  54 16 (29.6%) 2 (7.4%)  0.31, 0.24, 0.24 2 0.66, 1.0, 0.69 2  108 34 (31.5%)  214 61 (28.5%)  54 12 (22.2%)  0.58, 0.21, 0.45 2  108 69 (63.9%)  210 124 (57.9%)  54 26 (48.2%)  0.40, 0.06, 0.25 2  108 70 (64.8%)  210 124 (57.9%)  54 27 (50.0%)  0.32, 0.07, 0.17 2  0.27, 0.58, 0.28 2 0.793  0.283  = genotype repeat size, an XCI weighted AR repeat size, as described in methods two-tailed x2 tests for comparisons to control group 1, control group 2 and combined controls, respectively 3 two-tailed x2 tests for comparisons to control group 1 1 GRS 2  136  Table 3.6 Observed and Hardy-Weinberg-expected genotypes for ESR1 (TA)n repeat.  Control group 1 observed genotypes Control group 1 H-W expected genotypes  SS genotype 40 (38.1%) 38 (36.6%)  SL genotype 47 (44.8%) 50 (47.8%)  LL genotype 18 (17.1%) 16 (15.6%)  0.9'  POF patient observed genotypes POF patient H-W expected genotypes POF patient expected genotypes based on model 2  3 (5.6%) 10 (18.1%) 3 (5.4%)  40 (74.1%) 26 (48.9%) 39 (70.9%)  11 (20.4%) 18 (32.9%) 13 (23.8%)  0.0002' 0.93  p  x2 contingency test for comparison to observed frequencies, df=1 Frequencies expected based on a simple dominant model in which fl=y=9.8 3 X2 contingency test with Monte Carlo simulation for comparison to observed frequencies, df2 1  2  137  A)  7M*1-1—Mn©  victp—m—,© cagaagg  cagaagg FP  cagaagg gtcttcc  F  V  I cOmmimi©  caggagg gtcttcc ^  F  R  -40 aagg  VI  ^^(9.t.tcc ^ 1  B)  -  1  1-1  E  -P  gagg ctcc —  R  ^  2.40  2.20  1.80  1.40  • a.  80  ^  120^1.40^1.60^1.80^2^220^2.41  Allele X (FSHR  *Allele X^*Allele V^A80th^  -  29A)  NTC^ *Undetermined  Figure 3.1 TaqMan allelic discrimination. A) Differentially labeled probes specific for each allele do not fluoresce while attached to quenchers (Q). Exonuclease activity of the Taq DNA polymerase (Pol) releases quenchers and fluorescence is emitted. B) Example of results. In homozygous DNA samples, one of the probes will not bind noncomplementary sequence and will not be released from the quenchers and only a single fluorescence will be emitted. Heterozygotes will bind, release, and fluoresce both. NTC=no template control (water blank).  138  801 by  1346 by  1512 by  -2105 -2045 -1985 -1925 -1865 -1805 -1745 -1685 -1625 -1565 -1505 -1445 -1385 -1325 -1265 -1205 -1145 -1085 -1025 -965 -905 -845 -785 -725 -665 -605 -545 -485 -425 -365 -305 -245 -185 -125 -65 -5 +56 +116 +176 +236 +296 +356 +416 +476 +536 +596  atccagcgc tgtgcagtag cccagctgcg tgtctgccgg gaggggctgc caagticcct —■ gcctactggc tgcttcccga atccctgcca ttccacgcac aaacacatcc acacactctc ^ExonC ■ —■^ tctgcctagt tcacacactg agccactcgc acatgcgagc acattccttc cttccttctc actctctcgg cccttgactt ctacaagccc atggaacatt tctggaaaga cgttcttgat ccagcaglgt aggcttgttt tgatttctct ctctgtagct ttagcatttt gagaaagcaa cttacctttc tggctagtgt ctgtatccta gcagggagat gaggattgct gttctccatg gggtatgtgt gtgtctcctt tttctttcag gacttgtagg acttctttgt gccatttgca tataatttgg caggttcaca ttttttaaga gccctatgaa gtgctttttg catgtgtttt aaaaaggcat ttgaaaattg aaagtgtgat ttatggaaat taaatcatct gtaaaaaatt gctttggaaa gtaatgattg ctggccataa agggaaatat ctgcgatgca cctaatgtgt ttttaaccct ttatttgctg acaatctata gtcattaatg ctaaactcga ttttggcttc agctacattt gcatattgtc caacaatggt ctatttttgt aagaattaga taaaatgtat acttgatata aaatagtcaa aaaMact cttagtaaca gtaagcttgg catttagata gaccatgaac cacttcgtca gatactcfit tqgqtqtttq gqatagcaat taaaacaaag tattgatagt tgtatcagag tctattaggc tgcagcaaag gaagtttatt caaaagtata aactatccaa gattatagac gcatgatata cttcacctat tttttgtctc cttaatatgl III1Ftatat at gitatat atatatacac atatatgtgt gtgtgtatgt gcgtgtgcat gtttaacttt taattcagtt aaaaactttt ttctatttgt ttttcatctg gatatttgat tctgcatatc ctagcccaag tgaaccgaga agatcgagtt gtaggactaa aggatagaca tgcagaaatg cattttaaaa atctgttagc tggaccagac cgacaatgta acataattgc caaagctttg gttcbtgacc Egaggttatg tttggtatga aarTg tcaca ttttatattc agttttctga agttttggtt gcataaccaa cctgtggaag gcatgaacac ccatgtgcgc cjffaFtaaa qgtttttctg aatcatcctt cacatgagaa ttcctaatgg gaccaagtac agtactgtgg tccaacataa acacacaagt caggctgaga gaatctcaga aggttgtgga agggtctatc tactttggga gcattttgca gaggaagaaa ctgaggtcct ggcaggttgc attctcctga tggcaaaatg cagctcttcc tatatgtata ccctgaatct ccgccccctt cccctcagat gccccctgtc agttccccca gctgctaaat atagctgtct gtggctggct gcgtatgcaa ccgcacaccc cattctatct gccctatctc ggttacagtg tagtcctccc cagggtcatc ctatgtacac actacgtatt tctagccaac gaggazgggg aatcaaacag cagcaca_Ap, aaagagagac aaacagagat atatcggagt ctggcacggg gcac a ExonB gagaaagccg gcccctggat ccgtctttcg cgtttatttt aagcccagtc ttcc"ggc cacctttagc agatcctcgt gcgcccccgc cccctggccg tgaaactcag cctctatcca gcagcgacga caagtaagt aaagttcagg gaagctgctc tttgggatgc tcaaatcgag ttgtgcctgg agtgatgttt aa g e41111t cagggcaagg caacagtccc tggccgtcct ccagcacctt tgtaatgcat atgagctcgg gagaccagta cttaaagttg gaggcccg agcCh Caggag ctggcggagg gcgttcgtcc tqggaqctqc acttqctccq tcgqc[CcIgC C' ExonA ggcttcaccg gaccgcaggc tcccggggca gggccggggc cagagctcgc gtgtcggcgg -  -  ,  -  gacatgcgct ttctgagcct accatgaccc gagctggagc gtgtacctgg ttcaacgccg cccgggtctg agcgtgtctc cagccccacg  v  gcgtcgcctc taacctcggg ctgtgctctt tt_tcca^gcccgcc 41 1+21DR tctgccctgc ggqqacacqq tctqcaccctl gcccg^c acggacC4 3g tccacaccaa agcatccggg atggccctac tgcat^t ccaagggaac ccctgaaccg tccgcagctc aagatccccc tggagO1Pc cctgggcgag acagcagcaa gcccgccgtg tacaactacc ccgagggcgc cgcctacgag cggccgccgc caacgcgcag gtctacggtc agaccggcct cccctacggc aggctgcggc gttcggctcc aacggcctgg ggggtttccc cccactcaac cgagcccgct gatgctactg cacccgccgc cgcagctgtc gcctttcctg gccagcaggt gccctactac ctggagaacg agcccagcgg ctacacggtg -  Exon 1  Figure 3.2 EMU gene promoter region. Alternative first exons (in yellow) are spliced onto a common splice acceptor site (X). The coding region in exon 1 (green) is indicated in bold. The TA repeat region is highlighted in red. Putative ERE sequences are shown in pink, and CAAT and TATA boxes are shown in blue. Bars on the right indicate regions cloned into luciferase promoter construct by PCR primers indicated. (promoter as described by Becherini et al. 2000, using exon nomenclature suggested by Kos et al. 2001)  139  35%  ■ Control group 1 (N alleles = 54) ^ POF patients (N alleles = 90)  30% 25% 0 0 0-1  20% 15% 10% 5% 0%  s--1 t , on 1 1 [^U [ [ i mil [  ,  i:  15 16 17 18 19 20 21 22 23 24 25 ESR2 repeat length  Figure 3.3 ESR2 allele distribution in POF patients and control group 1.  140  A) ■ Combined control groups (N alleles=268) ^ POF patients (N alleles=108)  20% -  15% -  5% -  0% <1313 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29>29 AR repeat length B) 30% ■ Control group 1 (N=107) ^ POF patients (N=54)  25% -  5% 0%  n. i  16 17 18 19 20 21 22 23 24 25 26 27 28 29 AR genotype repeat size  Figure 3.4 AR allele (A) and genotype repeat size distribution (B).  141  50% -  ■ Combined control groups (N alleles = 262) 0 POF patients (N alleles = 108)  40% -  10% 0%  r  6^7^8^9^10^11 SHBG repeat length  Figure 3.5 SHBG allele distribution in POF patients and combined controls.  142  A) 80% -  ■ Combined control groups (N = 134) OPOF patients (N=54)  0% GG  ^  GA^AA  FSHR -29A/G genotype  B) ■ Combined control groups (N = 132) 0 POF patients (N=54)  GG  ^  GA  ^  AA  FSHR 919A/G genotype  C) 80% -  ■ Combined control groups (N = 132) ^ POF patients (N=54)  0% GG  1 In  ^  GA  ^  AA  FSHR 2039A/G genotype  Figure 3.6 Genotype distribution of FSHR SNPs. A) -29 B) 919 and C) 2039 SNPs  143  40% 35% 30% -  ■ Combined control groups (N alleles=264) ^ POF patients (N alleles=108)  25% 20% 15% 10% 5% 0%  —rn —n  11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 >26  ESR1 repeat length  Figure 3.7 ESR1 (TA)„ repeat allele distribution.  144  60% 50% -  ■ Control group 1 (N=208) 0 Control group 2 (N=54) 0 POF patients (N=108)  10% 0% SCA LCA  In SCG LCG STA LTA STG LTG  ESR1 haplotype (TA repeat, -397 SNP, -351 SNP)  Figure 3.8 Estimated ESR1 haplotype frequencies. In POF patients and control  groups 1 and 2. Haplotypes are described as (TA) n repeat, S<18 repeats or L?18 repeats; -397 SNP, C or T variant; and -351 SNP, A or G variant.  145  100% 75% ^ ESR1 LL (N=11) ^ ESR1 SL (N=40) ■ ESR1 SS (N=3) 25% -  0% <35 (N=38)^>1=35 (N=16) long FMR1 repeat allele  Figure 3.9 ESR1 genotype distribution and FMR1 repeat size in POF patients.  146  A) .  5000 00 4500 4000 3500 3000 2500 o) 2000 c a) 1500  -4=1 ct:t  2  1000  500 0 801bp (TA)14  801bp^1346bp (TA)22^(TA)14  1346bp (TA)22  ESR1 promoter fragment B)  25000 20000  -  2  0 no ligand 0 EE2  cvcv 15000 (1)  5 10000 .4=  5000  I  0 1346bp (TA)14^1346bp (TA)22  ^  ERE  Figure 3.10 Relative expression of 801bp and 1346bp reporter constructs. Luciferase reporter gene contructs were co-transfected with LacZ constructs into MCF-7 cell lines and luciferase expression is presented relative to (3-galactosidase expression. Error bars indicate the standard deviation of the 3 replicates. A) 801bp or 1346bp ESR1 fragment with (TA)14 or (TA)22 repeats. B) 1346 by (TA)14, 1346 by (TA)22 construct or ERE-containing construct with or without the addition of ethinyl estradiol (EE2). Only the ERE promoter showed a significant response to EE2. Note that samples are only comparable to those run in parallel and not to other experiments.  147  A) 35000 30000 9.,4 2'sa,4 25000 et tu^20000 or' .,..t4 15000 c.)^ F 10000 731-4^ 5000 -  0 no ligand 0 EE2 p=0.02  r^l r^1 p=0.05  --I--  0 1512bp (TA)14^1512bp (TA)22^ERE B) 35000 0o 30000 v, f:61.IL^25000 (1.1 4>^20000 cti .r.)0'^15000 '" o 10000 73 "^5000 -  0 no ligand 0 EE2  0 1512bp (TA)14^1512bp (TA)22^ERE Figure 3.11 Relative expression of 1512bp reporter with (TA)14 or (TA)22 repeats. Luciferase reporter gene contructs were co-transfected with LacZ constructs and  luciferase expression is presented relative to f3-galactosidase expression. Error bars indicate the standard deviation of the 3 replicates. A) MCF-7 and B) Ishikawa cell lines with and without the addition of ethinyl estradiol (EE2). Cells were grown in steroid depleted media from the time of transfection and in B) co-transfected with ERa to ensure sufficient endogenous ESR1 expression. Note that samples are only comparable to those run in parallel and not to other experiments. 148  3.5 References  Online Sources: Online Mendelian Inheritance in Man (OMIM), http://www.ncbi.nlm.nih.gov/Omim dbSNP Homepage, http://www.ncbi.nlm.nih.gov/SNP/ Ahda Y, Gromoll J, Wunsch A, Asatiani K, Zitzmann M, Nieschlag E and Simoni M. 2005. Follicle-stimulating hormone receptor gene haplotype distribution in normozoospermic and azoospermic men. J. Androl. 26:494-499. Aittomaki K, Lucena JL, Pakarinen P, Sistonen P, Tapanainen J, Gromoll J, Kaskikari R, Sankila EM, Lehvaslaiho H, Engel AR, et al. 1995. Mutation in the folliclestimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell. 82:959-968. Alevizaki M, Saltiki K, Cimponeriu A, Kanakakis I, Xita N, Alevizaki CC, Georgiou I and Sarika HL. 2007. Severity of cardiovascular disease in postmenopausal women: associations with common estrogen receptor alpha polymorphic variants. Eur. J. Endocrinol. 156:489-496. Allen EG, He W, Yadav-Shah M and Sherman SL. 2004. A study of the distributional characteristics of FMR1 transcript levels in 238 individuals. Hum. Genet. 114:439447. Allen LA, Achermann JC, Pakarinen P, Kotlar TJ, Huhtaniemi IT, Jameson JL, Cheetham TD and Ball SG. 2003. A novel loss of function mutation in exon 10 of the FSH receptor gene causing hypergonadotrophic hypogonadism: clinical and molecular characteristics. Hum. Reprod. 18:251-256. Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM and Belmont JW. 1992. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgenreceptor gene correlates with X chromosome inactivation. Am. J. Hum. Genet. 51:1229-1239. Altmae S, Haller K, Peters M, Hovatta 0, Stavreus-Evers A, Karro H, Metspalu A and Salumets A. 2007. Allelic estrogen receptor 1 (ESR1) gene variants predict the outcome of ovarian stimulation in in vitro fertilization. Mol. Hum. Reprod. 13:521526. Ashworth JJ, Smyth JV, Pendleton N, Horan M, Payton A, Worthington J, 011ier WE and Ashcroft GS. 2005. The dinucleotide (CA) repeat polymorphism of estrogen receptor beta but not the dinucleotide (TA) repeat polymorphism of estrogen receptor alpha is associated with venous ulceration. J. Steroid Biochem. Mol. Biol. 97:266-270. Beau I, Touraine P, Meduri G, Gougeon A, Desroches A, Matuchansky C, Milgrom E, Kuttenn F and Misrahi M. 1998. A novel phenotype related to partial loss of function mutations of the follicle stimulating hormone receptor. J. Clin. Invest. 102:13521359. 149  Becherini L, Gennari L, Masi L, Mansani R, Massart F, Morelli A, Falchetti A, Gonnelli S, Fiorelli G, Tanini A, et al. 2000. Evidence of a linkage disequilibrium between polymorphisms in the human estrogen receptor alpha gene and their relationship to bone mass variation in postmenopausal Italian women. Hum. Mol. Genet. 9:20432050. Bielawski KP, Klos P, Welnicka-Jaskiewicz M, Tidow M, Brandt B, Falkiewicz B and Zaczek A. 2005. An epidermal growth factor receptor intron 1 polymorphism in healthy women in Poland. Int. J. Biol. Markers. 20:184-188. Brandi ML, Becherini L, Gennari L, Racchi M, Bianchetti A, Nacmias B, Sorbi S, Mecocci P, Senin U and Govoni S. 1999. Association of the estrogen receptor alpha gene polymorphisms with sporadic Alzheimer's disease. Biochem. Biophys. Res. Commun. 265:335-338. Brandt B, Hermann S, Straif K, Tidow N, Buerger H and Chang-Claude J. 2004. Modification of breast cancer risk in young women by a polymorphic sequence in the egfr gene. Cancer Res. 64:7-12. Bretherick KL, Fluker MR and Robinson WP. 2005. FMR1 repeat sizes in the gray zone and high end of the normal range are associated with premature ovarian failure. Hum. Genet. 117:376-382. Britt KL and Findlay JK. 2002. Estrogen actions in the ovary revisited. J. Endocrinol. 175:269-276. Budowle B, Moretti TR, Baumstark AL, Defenbaugh DA and Keys KM. 1999. Population data on the thirteen CODIS core short tandem repeat loci in African Americans, U.S. Caucasians, Hispanics, Bahamians, Jamaicans, and Trinidadians. J. Forensic Sci. 44:1277-1286. Chamberlain NL, Driver ED and Miesfeld RL. 1994. The length and location of CAG trinucleotide repeats in the androgen receptor N-terminal domain affect transactivation function. Nucleic Acids Res. 22:3181-3186. Chan KM, Chiu CT, Tsui P, Wong DM and Fung WK. 2005. Population data for the Identifier 15 STR loci in Hong Kong Chinese. Forensic Sci. Int. 152:307-309. Chen T and Li S. 2004. Genetic polymorphism at nine STR loci in two ethnic groups of China. Forensic Sci. Int. 146:53-55. Chi DD, Hing AV, Helms C, Steinbrueck T, Mishra SK and Donis-Keller H. 1992. Two chromosome 7 dinucleotide repeat polymorphisms at gene loci epidermal growth factor receptor (EGFR) and pro alpha 2 (I) collagen (COL1A2). Hum. Mol. Genet. 1:135.  150  Cohn CS, Sullivan JA, Kiefer T and Hill SM. 1999. Identification of an enhancer element in the estrogen receptor upstream region: implications for regulation of ER transcription in breast cancer. Mol. Cell. Endocrinol. 158:25-36. Comings DE. 1998. Polygenic inheritance and micro/minisatellites. Mol. Psychiatry. 3:21-31. Corbo RM, Ulizzi L, Piombo L, Martinez-Labarga C, De Stefano GF and Scacchi R. 2007. Estrogen receptor alpha polymorphisms and fertility in populations with different reproductive patterns. Mol. Hum. Reprod. 13:537-540. Couse JF and Korach KS. 1999. Estrogen receptor null mice: what have we learned and where will they lead us? Endocr. Rev. 20:358-417. Cousin P, Calemard-Michel L, Lejeune H, Raverot G, Yessaad N, Emptoz-Bonneton A, Morel Y and Pugeat M. 2004. Influence of SHBG gene pentanucleotide TAAAA repeat and D327N polymorphism on serum sex hormone-binding globulin concentration in hirsute women. J. Clin. Endocrinol. Metab. 89:917-924. de Koning CH, Benjamins T, Harms P, Homburg R, van Montfrans JM, Gromoll J, Simoni M and Lambalk CB. 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. Doherty E, Pakarinen P, Tiitinen A, Kiilavuori A, Huhtaniemi I, Forrest S and Aittomaki K. 2002. A Novel mutation in the FSH receptor inhibiting signal transduction and causing primary ovarian failure. J. Clin. Endocrinol. Metab. 87:1151-1155. Eriksson AL, Lorentzon M, Mellstrom D, Vandenput L, Swanson C, Andersson N, Hammond GL, Jakobsson J, Rane A, Orwoll ES, et al. 2006. SHBG gene promoter polymorphisms in men are associated with serum sex hormone-binding globulin, androgen and androgen metabolite levels, and hip bone mineral density. J. Clin. Endocrinol. Metab. 91:5029-5037. Ferk P, Teran N and Gersak K. 2007. The (TAAAA)n microsatellite polymorphism in the SHBG gene influences serum SHBG levels in women with polycystic ovary syndrome. Hum. Reprod. 22:1031-1036. Flouriot G, Griffin C, Kenealy M, Sonntag-Buck V and Gannon F. 1998. Differentially expressed messenger RNA isoforms of the human estrogen receptor-alpha gene are generated by alternative splicing and promoter usage. Mol. Endocrinol. 12:19391954. Gao Y, Zhang Z, Wang Z and Bian S. 2005. Genetic data of 15 STR forensic loci in eastern Chinese population. Forensic Sci. Int. 154:78-80.  151  Geng L, Yao Z, Yang H, Luo J, Han L and Lu Q. 2007. Association of CA repeat polymorphism in estrogen receptor beta gene with postmenopausal osteoporosis in Chinese. J. Genet. Genomics. 34:868-876. Georgiou I, Konstantelli M, Syrrou M, Messinis IE and Lolis DE. 1997. Oestrogen receptor gene polymorphisms and ovarian stimulation for in-vitro fertilization. Hum Reprod. 12:1430-143. Georgiou I, Syrrou M, Bouba I, Dalkalitsis N, Paschopoulos M, Navrozoglou I and Lolis D. 1999. Association of estrogen receptor gene polymorphisms with endometriosis. Fertil. Steril. 72:164-166. Gerhardt A, Scharf RE, Mikat-Drozdzynski B, Krussel JS, Bender HG and Zotz RB. 2006. Maternal IVS 1-401 T allele of the estrogen receptor alpha is an independent predictor of late fetal loss. Fertil. Steril. 86:448-453. Giguere Y, Dewailly E, Brisson J, Ayotte P, Laflamme N, Demers A, Forest VI, Dodin S, Robert J and Rousseau F. 2001. Short polyglutamine tracts in the androgen receptor are protective against breast cancer in the general population. Cancer Res. 61:58695874. Giovannucci E, Stampfer MJ, Krithivas K, Brown M, Dahl D, Brufsky A, Talcott J, Hennekens CH and Kantoff PW. 1997. The CAG repeat within the androgen receptor gene and its relationship to prostate cancer. Proc. Natl. Acad. Sci. U. S. A. 94:3320-3323. Grandien K, Berkenstam A and Gustafsson JA. 1997. The estrogen receptor gene: promoter organization and expression. Int. J. Biochem. Cell Biol. 29:1343-1369. Greb RR, Grieshaber K, Gromoll J, Sonntag B, Nieschlag E, Kiesel L and Simoni M. 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. Guarducci E, Nuti F, Becherini L, Rotondi M, Balercia G, Forti G and Krausz C. 2006. Estrogen receptor alpha promoter polymorphism: stronger estrogen action is coupled with lower sperm count. Hum. Reprod. 21:994-1001. Haiman CA, Riley SE, Freedman ML, Setiawan VW, Conti DV and Le Marchand L. 2005. Common genetic variation in the sex steroid hormone-binding globulin (SHBG) gene and circulating shbg levels among postmenopausal women: the Multiethnic Cohort. J. Clin. Endocrinol. Metab. 90:2198-2204. Hammond GL. 2002. Access of reproductive steroids to target tissues. Obstet. Gynecol. Clin. North Am. 29:411-423. Herrington DM, Howard TD, Brosnihan KB, McDonnell DP, Li X, Hawkins GA, Reboussin DM, Xu J, Zheng SL, Meyers DA, et al. 2002. Common estrogen receptor 152  polymorphism augments effects of hormone replacement therapy on E-selectin but not C-reactive protein. Circulation. 105:1879-1882. Hickey T, Chandy A and Norman RJ. 2002. The androgen receptor CAG repeat polymorphism and X-chromosome inactivation in Australian Caucasian women with infertility related to polycystic ovary syndrome. J. Clin. Endocrinol. Metab. 87:161165. Hill WG. 1974. Estimation of linkage disequilibrium in randomly mating populations. Heredity. 33:229-239. Hogeveen KN, Talikka M and Hammond GL. 2001. Human sex hormone-binding globulin promoter activity is influenced by a (TAAAA)n repeat element within an Alu sequence. J. Biol. Chem. 276:36383-36390. Horie K, Takakura K, Fugiwara H, Suginami H, Liao S and Mori T. 1992. Immunohistochemical localization of androgen receptor in the human ovary throughout the menstrual cycle in relation to oestrogen and progesterone receptor expression. Hum. Reprod. 7:184-190. Hsieh YY, Wang YK, Chang CC and Lin CS. 2007. Estrogen receptor alpha-351 XbaI*G and -397 PvuII*C-related genotypes and alleles are associated with higher susceptibilities of endometriosis and leiomyoma. Mol. Hum. Reprod. 13:117-122. Ibanez L, Ong KK, Mongan N, Jaaskelainen J, Marcos MV, Hughes IA, De Zegher F and Dunger DB. 2003. Androgen receptor gene CAG repeat polymorphism in the development of ovarian hyperandrogenism. J. Clin. Endocrinol. Metab. 88:33333338. Ioannidis JP, Ralston SH, Bennett ST, Brandi ML, Grinberg D, Karassa FB, Langdahl B, van Meurs JB, Mosekilde L, Scollen S, et al. 2004. Differential genetic effects of ESR1 gene polymorphisms on osteoporosis outcomes. JAMA. 292:2105-2114. Irvine RA, Ma H, Yu MC, Ross RK, Stallcup MR and Coetzee GA. 2000. Inhibition of p160-mediated coactivation with increasing androgen receptor polyglutamine length. Hum. Mol. Genet. 9:267-274. Jefferson WN, Padilla-Banks E and Newbold RR. 2007. Disruption of the female reproductive system by the phytoestrogen genistein. Reprod. Toxicol. 23:308-316. Joaquin M and Watson RJ. 2003. Cell cycle regulation by the B-Myb transcription factor. Cell Mol. Life Sci. 60:2389-2401. Khosla S, Riggs BL, Atkinson EJ, Oberg AL, Mavilia C, Del Monte F, Melton LJ,3rd and Brandi ML. 2004. Relationship of estrogen receptor genotypes to bone mineral density and to rates of bone loss in men. J. Clin. Endocrinol. Metab. 89:1808-1816.  153  Kim SH, Choi YM, Jun JK, Kim SH, Kim JG and Moon SY. 2005. Estrogen receptor dinucleotide repeat polymorphism is associated with minimal or mild endometriosis. Fertil. Steril. 84:774-777. Kimura S, Matsumoto T, Matsuyama R, Shiina H, Sato T, Takeyama K and Kato S. 2007. Androgen receptor function in folliculogenesis and its clinical implication in premature ovarian failure. Trends Endocrinol. Metab. 18:183-189. Kitamura I, Ando F, Koda M, Okura T and Shimokata H. 2007. Effects of the interaction between lean tissue mass and estrogen receptor alpha gene polymorphism on bone mineral density in middle-aged and elderly Japanese. Bone. 40:1623-1629. Kjaergaard AD, Ellervik C, Tybjaerg-Hansen A, Axelsson CK, Gronholdt ML, Grande P, Jensen GB and Nordestgaard BG. 2007. Estrogen receptor alpha polymorphism and risk of cardiovascular disease, cancer, and hip fracture: cross-sectional, cohort, and case-control studies and a meta-analysis. Circulation. 115:861-871. Koch W, Hoppmann P, Pfeufer A, Mueller JC, Schomig A and Kastrati A. 2005. No replication of association between estrogen receptor alpha gene polymorphisms and susceptibility to myocardial infarction in a large sample of patients of European descent. Circulation. 112:2138-2142. Kolibianakis EM, Papanikolaou EG, Fatemi HM and Devroey P. 2005. Estrogen and folliculogenesis: is one necessary for the other? Curr. Opin. Obstet. Gynecol. 17:249253. Korach KS. 1994. Insights from the study of animals lacking functional estrogen receptor. Science. 266:1524-1527. Kos M, Reid G, Denger S and Gannon F. 2001. Minireview: genomic organization of the human ERalpha gene promoter region. Mol. Endocrinol. 15:2057-2063. Langdahl BL, Lokke E, Carstens M, Stenkjaer LL and Eriksen EF. 2000. A TA repeat polymorphism in the estrogen receptor gene is associated with osteoporotic fractures but polymorphisms in the first exon and intron are not. J. Bone Miner. Res. 15:22222230. Levine DA and Boyd J. 2001. The androgen receptor and genetic susceptibility to ovarian cancer: results from a case series. Cancer Res. 61:908-911. Liu W, Innocenti F, Chen P, Das S, Cook EH,Jr and Ratain MJ. 2003. Interethnic difference in the allelic distribution of human epidermal growth factor receptor intron 1 polymorphism. Clin. Cancer Res. 9:1009-1012. Meduri G, Touraine P, Beau I, Lahuna 0, Desroches A, Vacher-Lavenu MC, Kuttenn F and Misrahi M. 2003. Delayed puberty and primary amenorrhea associated with a novel mutation of the human follicle-stimulating hormone receptor: clinical, histological, and molecular studies. J. Clin. Endocrinol. Metab. 88:3491-3498.  154  Molvarec A, Ver A, Fekete A, Rosta K, Derzbach L, Derzsy Z, Karadi I and Rigo J,Jr. 2007. Association between estrogen receptor alpha (ESR1) gene polymorphisms and severe preeclampsia. Hypertens. Res. 30:205-211. Nakayama T, Kuroi N, Sano M, Tabara Y, Katsuya T, Ogihara T, Makita Y, Hata A, Yamada M, Takahashi N, et al. 2006. Mutation of the follicle-stimulating hormone receptor gene 5'-untranslated region associated with female hypertension. Hypertension. 48:512-518. Ogawa S, Emi M, Shiraki M, Hosoi T, Ouchi Y and Inoue S. 2000a. Association of estrogen receptor beta (ESR2) gene polymorphism with blood pressure. J. Hum. Genet. 45:327-330. Ogawa S, Hosoi T, Shiraki M, Orimo H, Emi M, Muramatsu M, Ouchi Y and Inoue S. 2000b. Association of estrogen receptor beta gene polymorphism with bone mineral density. Biochem. Biophys. Res. Commun. 269:537-541. Perez Mayorga M, Gromoll J, Behre HM, Gassner C, Nieschlag E and Simoni M. 2000. Ovarian response to follicle-stimulating hormone (FSH) stimulation depends on the FSH receptor genotype. J. Clin. Endocrinol. Metab. 85:3365-3369. Raczek E. 2006. Population data on the 11 STR loci in the Upper Silesia (Poland). Forensic Sci. Int. 168:68-72. Riancho JA, Zarrabeitia MT, Valero C, Sanudo C, Mijares V and Gonzalez-Macias J. 2006. A gene-to-gene interaction between aromatase and estrogen receptors influences bone mineral density. Eur. J. Endocrinol. 155:53-59. Ross J, Parson W, Furac I, Kubat M and Holland M. 2001. Multiplex PCR amplification of eight STR loci in Austrian and Croatian Caucasian populations. Int. J. Legal Med. 115:57-60. Sano M, Inoue S, Hosoi T, Ouchi Y, Emi M, Shiraki M and Orimo H. 1995. Association of estrogen receptor dinucleotide repeat polymorphism with osteoporosis. Biochem. Biophys. Res. Commun. 217:378-383. Scariano JK, Simplicio SG, Montoya GD, Garry PJ and Baumgartner RN. 2004. Estrogen receptor beta dinucleotide (CA) repeat polymorphism is significantly associated with bone mineral density in postmenopausal women. Calcif. Tissue Int. 74:501-508. Schlessinger D, Herrera L, Crisponi L, Mumm S, Percesepe A, Pellegrini M, Pilia G and Forabosco A. 2002. Genes and translocations involved in POF. Am. J. Med. Genet. 111:328-333. Schuit SC, de Jong FH, Stolk L, Koek WN, van Meurs JB, Schoofs MW, Zillikens MC, Hofman A, van Leeuwen JP, Pols HA, et al. 2005. Estrogen receptor alpha gene polymorphisms are associated with estradiol levels in postmenopausal women. Eur. J. Endocrinol. 153:327-334. 155  Shiina H, Matsumoto T, Sato T, Igarashi K, Miyamoto J, Takemasa S, Sakari M, Takada I, Nakamura T, Metzger D, et al. 2006. Premature ovarian failure in androgen receptor-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 103:224-229. Simoni M, Nieschlag E and Gromoll J. 2002. Isoforms and single nucleotide polymorphisms of the FSH receptor gene: implications for human reproduction. Hum. Reprod. Update. 8:413-421. Simpson JL and Rajkovic A. 1999. Ovarian differentiation and gonadal failure. Am. J. Med. Genet. 89:186-200. Sowers MR, Jannausch ML, McConnell DS, Kardia SR and Randolph JF,Jr. 2006. Menstrual cycle markers of ovarian aging and sex steroid hormone genotypes. Am. J. Med. 119:S31-43. Sudo S, Kudo M, Wada S, Sato 0, Hsueh AJ and Fujimoto S. 2002. Genetic and functional analyses of polymorphisms in the human FSH receptor gene. Mol. Hum. Reprod. 8:893-899. Sundarrajan C, Liao W, Roy AC and Ng SC. 1999. Association of oestrogen receptor gene polymorphisms with outcome of ovarian stimulation in patients undergoing IVF. Mol. Hum. Reprod. 5:797-802. Sundblad V, Chiauzzi VA, Escobar ME, DaM L and Charreau EH. 2004. Screening of FSH receptor gene in Argentine women with premature ovarian failure (POF). Mol. Cell. Endocrinol. 222:53-59. Syrrou M, Georgiou I, Patsalis PC, Bouba I, Adonakis G and Pagoulatos GN. 1999. Fragile X premutations and (TA)n estrogen receptor polymorphism in women with ovarian dysfunction. Am. J. Med. Genet. 84:306-308. Thompson DA, McPherson LA, Carmeci C, deConinck EC and Weigel RJ. 1997. Identification of two estrogen receptor transcripts with novel 5' exons isolated from a MCF7 cDNA library. J. Steroid Biochem. Mol. Biol. 62:143-153. Touraine P, Beau I, Gougeon A, Meduri G, Desroches A, Pichard C, Detoeuf M, Paniel B, Prieur M, Zorn JR, et al. 1999. New natural inactivating mutations of the folliclestimulating hormone receptor: correlations between receptor function and phenotype. Mol. Endocrinol. 13:1844-1854. Tsukamoto K, Inoue S, Hosoi T, Orimo H and Emi M. 1998. Isolation and radiation hybrid mapping of dinucleotide repeat polymorphism at the human estrogen receptor beta locus. J. Hum. Genet. 43:73-74. Tut TG, Ghadessy FJ, Trifiro MA, Pinsky L and Yong EL. 1997. Long polyglutamine tracts in the androgen receptor are associated with reduced trans-activation, impaired sperm production, and male infertility. J. Clin. Endocrinol. Metab. 82:3777-3782.  156  van Meurs JB, Schuit SC, Weel AE, van der Klift M, Bergink AP, Arp PP, Cohn EM, Fang Y, Hofman A, van Duijn CM, et al. 2003. Association of 5' estrogen receptor alpha gene polymorphisms with bone mineral density, vertebral bone area and fracture risk. Hum. Mol. Genet. 12:1745-1754. Vendola K, Zhou J, Wang J, Famuyiwa OA, Bievre M and Bondy CA. 1999. Androgens promote oocyte insulin-like growth factor I expression and initiation of follicle development in the primate ovary. Biol. Reprod. 61:353-357. Vendola KA, Zhou J, Adesanya 00, Weil SJ and Bondy CA. 1998. Androgens stimulate early stages of follicular growth in the primate ovary. J. Clin. Invest. 101:2622-2629. Weel AE, Uitterlinden AG, Westendorp IC, Burger H, Schuit SC, Hofman A, Helmerhorst TJ, van Leeuwen JP and Pols HA. 1999. Estrogen receptor polymorphism predicts the onset of natural and surgical menopause. J. Clin. Endocrinol. Metab. 84:3146-3150. Weiderpass E, Persson I, Melhus H, Wedren S, Kindmark A and Baron JA. 2000. Estrogen receptor alpha gene polymorphisms and endometrial cancer risk. Carcinogenesis. 21:623-627. Westberg L, Baghaei F, Rosmond R, Hellstrand M, Landen M, Jansson M, Holm G, Bjorntorp P and Eriksson E. 2001. Polymorphisms of the androgen receptor gene and the estrogen receptor beta gene are associated with androgen levels in women. J. Clin. Endocrinol. Metab. 86:2562-2568. Westberg L, Ho HP, Baghaei F, Nilsson S, Melke J, Rosmond R, Holm G, Bjorntorp P and Eriksson E. 2004. Polymorphisms in oestrogen and progesterone receptor genes: possible influence on prolactin levels in women. Clin. Endocrinol. (Oxf). 61:216223 Westberg L, Melke J, Landen M, Nilsson S, Baghaei F, Rosmond R, Jansson M, Holm G, Bjorntorp P and Eriksson E. 2003. Association between a dinucleotide repeat polymorphism of the estrogen receptor alpha gene and personality traits in women. Mol. Psychiatry. 8:118-122. Wittke-Thompson JK, Pluzhnikov A and Cox NJ. 2005. Rational inferences about departures from Hardy-Weinberg equilibrium. Am. J. Hum. Genet. 76:967-986. Xita N, Tsatsoulis A, Chatzikyriakidou A and Georgiou I. 2003. Association of the (TAAAA)n repeat polymorphism in the sex hormone-binding globulin (SHBG) gene with polycystic ovary syndrome and relation to SHBG serum levels. J. Clin. Endocrinol. Metab. 88:5976-5980. Xita N, Tsatsoulis A, Stavrou I and Georgiou I. 2005. Association of SHBG gene polymorphism with menarche. Mol. Hum. Reprod. 11:459-462.  157  Yang CQ, Chan KY, Ngan HY, Khoo US, Chiu PM, Chan QK, Xue WC and Cheung AN. 2006. Single nucleotide polymorphisms of follicle-stimulating hormone receptor are associated with ovarian cancer susceptibility. Carcinogenesis. 27:15021506. Yaron M, Levy T, Chetrit A, Levavi H, Sabah G, Schneider D, Halperin R, Ben-Rafael Z and Friedman E. 2001. The polymorphic CAG repeat in the androgen receptor gene in Jewish Israeli women with endometrial carcinoma. Cancer. 92:1190-1194. Yu H, BharajB, Vassilikos EJ, Giai M and Diamandis EP. 2000. Shorter CAG repeat length in the androgen receptor gene is associated with more aggressive forms of breast cancer. Breast Cancer Res. Treat. 59:153-161.  158  Chapter 4: X chromosome inactivation skewing in primary and secondary POF  3  4.1 Introduction Diagnosis of premature ovarian failure (POF) is based on at least 6 months of amenorrhea and elevated FSH in the menopausal range (>40 IU/1) before the age of 40 (Conway et al. 2000). Most patients with POF present with secondary amenorrhea, in which menstruation spontaneously ceases after a period of normal menses, and many studies of POF in the literature confine the term to this group (Simpson and Rajkovic. 1999; Vegetti et al. 2000). Others, however, include patients with primary amenorrhea, who do not undergo menarche, within the POF diagnosis (Anasti. 1998; Goswami and Conway. 2005; Kalantaridou et al. 1998; Nelson et al. 2005). Known causes of POF include genetic causes, autoimmune disorders, metabolic dysfunction, infection, resistant ovary syndrome, and iatrogenic causes such as chemotherapy and radiation (Davis. 1996). However, for the large majority of women with POF, a cause is not identified (Conway. 2000). Genetic causes of POF include both chromosomal abnormalities and autosomal and X-chromosome gene mutations. Disruption of a number of regions on both Xp and Xq, as well as 45,X/46,XX mosaicism, can lead to primary or secondary amenorrhea (POF) through accelerated oocyte atresia (Simpson and Rajkovic. 1999). It is not yet clear whether these effects are due to disruption of specific X-linked genes, position effects such as the inactivation of autosomal genes translocated to the X chromosome, or  3  A version of this chapter has been published: Bretherick KL, Metzger DL, Chanoine JP, Panagiotopoulos C, Watson SK, Lam WL, Fluker MR, Brown CJ, Robinson WP. 2007. Skewed X-chromosome inactivation is associated with primary but not secondary ovarian failure. Am J Med Genet A 143(9):945-51.  159  disruption of pairing at meiosis (Toniolo. 2006). Even when cytogenetically 'normal' chromosomes are found, it has been suggested that the X-chromosome is probably involved in the majority of familial and sporadic POF cases (Davis et al. 2000; Davison et al. 1999). Therefore, small X chromosome deletions, not detectable by conventional cytogenetic methods, may also contribute to POF. X-chromosome inactivation (XCI) is the dosage compensation achieved by the random inactivation of one of the two X chromosomes in every female cell during development, which is stably transmitted in every subsequent cellular division (Lyon. 1961). Skewed XCI occurs when there is preferential inactivation of either the maternal or paternal X chromosome. Several possible causes of skewed XCI have been described, one of which is the preferential inactivation of an X chromosome bearing a gene mutation or structural abnormality (Brown and Robinson. 2000). If undetected X-chromosome aberrations account for a significant percentage of POF cases, it is possible that there is an increased incidence of skewed XCI in women experiencing POF (Sato et al. 2004). We therefore assessed XCI skewing in our population of POF patients presenting with primary or secondary amenorrhea. Furthermore, we analyzed skewed cases by high resolution DNA microarray to determine whether undetected X-chromosome abnormalities are responsible for the skewed XCI observed in POF patients.  4.2 Methods 4.2.1 Samples The collection of samples for this study was approved by the University of British Columbia Clinical Research Ethics Board (CREB), approval number CO1-0460 (Appendix 1). Young women with primary ovarian failure (N=4) were ascertained at the 160  Endocrinology and Diabetes unit of the Children's & Women's Health Centre of British Columbia. All four patients had primary amenorrhea with elevated FSH, and were karyotypically normal on routine diagnostic analysis. Clinical details on these patients are summarised in Table 4.1. Notably, all four primary amenorrhea patients had some pubertal development, indicating that estradiol secretion was present initially, but stopped early. Women experiencing idiopathic secondary amenorrhea (N=54) were ascertained from the POF clinic at the Women's Health Centre of British Columbia. POF diagnosis was made based on the absence of menses for at least 3 months and two serum FSH results of >40 IU/L obtained more than one month apart, prior to age 40. For all patients diagnosis of POF was made before the age of 40; however DNA was obtained many years after diagnosis in some cases. Exact age at the time of blood draw was known for 36 of the secondary amenorrhea patients and the average age among these was 35.4 yrs (range 21-50). Control women in this study (N=109) are the same as those described in previous chapters. The average age of controls at the time of blood draw was 34.9 yrs (range 17-45). 4.2.2 XCI skewing assay The degree of XCI skewing was determined by assaying methylation at a  polymorphic CAG repeat at the Androgen Receptor (AR) locus (Allen et al. 1992) in DNA extracted from peripheral blood, with the use of MIC2 amplification to ensure complete digestion as described in Chapter 2. For samples homozygous at the AR locus, the two X chromosomes can not be distinguished from one another, therefore the degree of X chromosome inactivation skewing can not be determined and the samples are considered uninformative for the assay. In such cases the degree of skewed XCI can be  161  assessed at a polymorphic CGG repeat at the FMRI locus (Beever et al. 2003b). However, in some cases XCI skewing results from AR and FMR1 are not well correlated (Beever et al. 2003a), and since the FMR1 gene specifically is associated with risk for POF (Chapter 2), the degree of skewed XCI assayed at the FMR1 loci in the POF population may reflect the role of FMR1 in POF disease pathogenesis rather than the degree of XCI skewing. As the AR gene has not been shown to be associated with POF, we limited the XCI skewing results in this study to those samples that were informative at the AR locus only. XCI skewing was, however, assessed at the FMR1 locus for one primary amenorrhea patient that was uninformative for skewing at AR, as there was such a small sample size in this group. 4.2.3 DNA microarray To determine whether skewed X inactivation in POF patients is due to cryptic Xchromosome deletions or insertions, high resolution DNA microarray was performed on selected samples with skewed XCI (two primary amenorrhea cases with skewed XCI of 100% (P0F54) and 90% (P0F58), and one secondary amenorrhea case with skewed XCI of 97% (P0F62)). As we lacked parental DNA, or cell lines to follow up the numerous small changes now well-established to be common in the human genome (Conrad et al. 2006; Eichler. 2006; McCarron et al. 2006), the purpose was simply to identify an obvious change that may have been missed in conventional karyotyping of these individuals. DNA microarray was performed as described by Ishkanian et al. (2004). Briefly, arrays consisted of 32,433 BAC tiling array elements spanning the entire genome at —1.5fold coverage. Clones were robotically spotted onto slides in triplicate, and equal  162  quantities of POF patient and male genomic reference DNA were labelled, denatured and hybridized for —36 hrs. Images were obtained with the ArrayWorx CCD scanner and SoftWorx (Applied Precision, Issaquah WA) image analysis program. Data was normalized for spatial bias using CGH norm (Khojasteh et al. 2005) and visualized with seeGH (Chi et al. 2004). Regions of alteration were detected with aCGH-Smooth (Jong et al. 2004) as well as by moving average and log2 ratio cut off of 0.2. Hybridization quality was ascertained with male-female X chromosome ratio (log2 average 0.24+/-0.12 relative to the autosome loge ratio of 0.003+/-0.07). Select deletions and duplications suggested by this analysis were followed up by conventional fluorescent in-situ hybridization (FISH). FISH was performed using a commercially available probe, AR for the region Xq 11.2 following the protocol described by the manufacturer (Vysis, Abbott Laboratories Inc., Mississauga, ON), and with BAC clones (clone name indicated in brackets) as probes for the following regions Xp11.21 (RP11-1150121), Xq13.2 (RP11-13M9), Xq21.31 (RP11-1150018), Xq27.3-28 (RP11963J21). BAC clones were labelled with Spectrum Red (Vysis,) using a nick-translation labelling kit (Vysis), precipitated by salt extraction, suspended in LSI/WCP buffer (Vysis), and hybridized overnight at 37°C to metaphase cells. Slides were washed, dried and counter stained with DAPI (Vysis). For each probe, an image of a metaphase spread was captured to confirm the approximate location and >300 interphase cells were scored. An example of FISH results obtained is shown in Figure 4.1  163  4.3 Results 4.3.1 XCI skewing in patients with secondary amenorrhea The percentage of individuals informative for skewing was similar in the patient and control groups; 50 out of 58 (86%) POF patients and 97 out of 109 (89%) control women were heterozygous at the AR polymorphism. There was no difference in XCI skewing between the patients with secondary amenorrhea and the control women. Four out of 47 (8.5%) women with secondary amenorrhea, and 8 out of 97 (8.2%) control women had skewed XCI >90% (p=n.s.) (Figure 4.2). The mean degree of skewing was also the same in the two groups (71.2% in the women with secondary amenorrhea and 70.8% in control women).  4.3.2 XCI skewing in patients with primary amenorrhea Skewed XCI was substantially increased in the women experiencing primary ovarian failure. All three individuals with primary amenorrhea who were informative for the AR polymorphism had skewed XCI >90% (p=0.001 vs. control women, one tailed Fisher's exact test). For the sample with primary amenorrhea that was uninformative for the AR polymorphism, XCI skewing was assessed at the FMR1 locus and found to be 54% (not skewed). Nonetheless, even if this non-skewed primary amenorrhea sample is included in the analysis, the primary amenorrhea group still has a significantly higher level of skewed XCI than the controls (p=0.004 vs. control women, one tailed Fisher's exact test). Specific skewing values for all four primary amenorrhea patients are presented with clinical details in table 4.1.  164  4.3.3 DNA microarray  Array analysis was performed on two primary amenorrhea cases with skewing of 90% and 100%, and in one secondary amenorrhea case with skewing of 97%. No regions of copy number alteration were detected on autosomal chromosomes. Hybridization results for the X chromosome for these three samples are shown in Figure 4.3. There are no large scale regions of copy number variation apparent in the microarray results for any of the POF samples analysed. Any detected regions of alteration within the POF samples have been shown to be normal copy number variation (Wong et al. 2007). Selected suspected regions of copy number alteration were followed up by FISH (Figure 4.1). For all three samples the majority (>90%) of interphase cells displayed two signals for each probe and had very few (<5%) cells with either one or three or more signals (Table 4.2). This is within the expected range for normal healthy controls (Devi et al. 1998) and therefore no duplications or deletions (either complete or mosaic) were confirmed. 4.4 Discussion  POF may occur with either primary or secondary amenorrhea; however, the presentation and management of these two conditions is quite different (Davis. 1996), therefore it is expected that they may have related but distinct etiologies. Skewed XCI was associated with POF in patients presenting with primary but not secondary amenorrhea, suggesting that if cryptic X-abnormalities do contribute to POF they may be limited to the early-onset group.  165  The present results confirm those of Bione et al. (2006), who found no significant increase in the degree and frequency of skewed XCI in a population of 151 Italian POF patients with secondary amenorrhea. In contrast Sato et al. (2004) reported a significant increase in skewed XCI in Japanese women experiencing POF after 'previous menstrual regularity'. Specifically, XCI skewing >90% was observed in 5/24 patients (21%) but in 0/29 controls (p=0.02). It should be noted that they found less skewing in controls than reported elsewhere for healthy women (Beever et al. 2003b; Uehara et al. 2001), which may be due to the small sample size in the control group. They also report a somewhat higher incidence of 45,X mosaicism and X-chromosome abnormalities in their patient group than is reported elsewhere for POF patients (Devi and Benn. 1999). Although they excluded karyotypically abnormal patients from their XCI skewing analysis, it is possible that their mode of ascertainment resulted in the acquirement of more severe POF cases. These patients may present with an earlier onset of symptoms and be more likely to have skewed XCI. There are several possible explanations for the increased incidence of skewed XCI observed in POF patients with primary amenorrhea. Skewed XCI itself may be responsible for the primary amenorrhea, as could be the case if there was chance preferential activation of an X chromosome with a mutation in a gene directly affecting ovarian development or function. As this would require 2 independent events, a more parsimonious explanation would be a common cause that led to both skewed XCI and primary amenorrhea, such as selection against cells activating an X chromosome carrying a deleterious mutation or deletion. However, all POF patients in our study were karyotypically normal and we were unable to detect obvious X-chromosome copy  166  number alterations by high resolution DNA microarray in any of the POF cases with skewed XCI. An alternative explanation for skewed XCI is a reduction or effective reduction in precursor cell number, due to a small number of cells present at the time of inactivation or a decrease in the number of cells contributing to the fetus (Robinson et al. 2001). One cause of a decrease in the number of precursor cells that may also result in primary amenorrhea is trisomy mosaicism confined predominantly to the placenta (Bretherick et al. 2005; Robinson et al. 2001). Skewed XCI is commonly observed in the diploid cell line arising from a trisomy rescue event (Lau et al. 1997; Penaherrera et al. 2000), and it has been shown that the oocytes may contain high levels of trisomic cells even when other fetal tissues are not affected by trisomy (Stavropoulos et al. 1998). These cells may be more likely to suffer early atresia, resulting in a reduced follicular pool size, and the occurrence of primary amenorrhea. Low birth weight, a frequent finding in trisomy mosaicism resulting from trisomic zygote rescue (Robinson et al. 1997), is reported for two of the primary amenorrhea patients in this study (Table 4.1), providing further support for this hypothesis. If trisomy mosaicism could also cause secondary amenorrhea in some persons, we might also expect to see an increased incidence of skewed XCI in secondary amenorrhea patients. Individuals with trisomy 21 have an earlier age at menopause (Seltzer et al. 2001) possibly due to cell atresia resulting from abnormal meiotic pairing (Cheng et al. 1998). However, trisomy mosaicism for other chromosomes that are less compatible with cell survival may be more likely to cause early follicular atresia, resulting in a more severe phenotype such as primary amenorrhea. Trisomy 21 makes up only —10% of  167  clinically recognized trisomic pregnancies (Hassold et al. 1996) and trisomy 21 mosaicism occurs in less than 0.01 % of live births (Devlin and Morrison. 2004), therefore it would not be expected to be a major contributor to trisomy mosaicism in the secondary amenorrhea patient population. A severe effect of mosaicism for chromosomes other than trisomy 21, and the relative prevalence of these cases may explain the lack of an increase in skewed XCI among secondary amenorrhea patients in this study. In conclusion, we report that skewed XCI is associated with primary but not secondary ovarian failure, and that the co-occurrence of skewed XCI and POF is not due to X-chromosome copy number alterations detectable by microarray. The primary amenorrhea population was quite limited in this report and these findings must be reproduced in larger populations to confirm this association. Nonetheless, these results suggest that primary and secondary amenorrhea may have distinct etiologies and highlight the fact that they should be examined separately in studies looking for POF disease gene associations. Furthermore, skewed XCI may help identify a subset of POF patients with differing clinical features.  168  Table 4.1 Clinical details for POF patients presenting with primary amenorrhea.  Age at work-up (yr)  POF18 16  POF19 15  POF54 17  POF58 15  Tanner Stage' FSH (U/L) LH (U/L) Estradiol (pmol/L) Birth weight3 Maternal age at birth Anti-ovarian antibodies Karyotype FMR1 repeat4 Other features  3 97.52 19.2 <35 10th Not available Negative 46,XX Normal Pre-pubertal uterus  3 92.22 23.62 73 <3rd  3 79.3 28.5 <40 <3rd  2 85.2 27.1 53 75th  28 Negative 46,XX Normal Scoliosis 28° max; Learning disability  25 Negative 46,XX Normal  22 Negative 46,XX Normal BMI >97th  XCI skewing at AR  U15  92%  100%  90%  'Tanner stage for breast development is reported An average of 2 values obtained >1 month apart is given. 3 A11 patients were born at or past term, and the birth weight percentile is given 4Normal indicates both FMR1 alleles were below 30 repeats 5 U1—uninformative for skewing at the AR locus; skewing assayed at the FMR1 locus was 54% 2  Table 4.2 FISH results for microarray follow-up Number of cells with: Sample and probe'  Cells counted (N)  one signal (%)  two signals (%)^three signals (%)  >three signals (%)  RP11-1150121  312  4 (1.3)  300 (96.2)  8 (2.6)  0 (0.0)  AR  500  6 (1.2)  491 (98.2)  3 (0.6)  0 (0.0)  RP11-13M9  304  4 (1.3)  298 (98.0)  2 (0.7)  0 (0.0)  RP11-963J21  366  4 (1.1)  360 (98.4)  2 (0.5)  0 (0.0)  AR  514  5 (1.0)  500 (97.3)  8 (1.6)  1 (0.2)  RP11-13M9  347  4 (1.2)  322 (92.8)  16 (4.6)  5 (1.4)  AR  360  7 (1.9)  342 (95.0)  9 (2.5)  2 (0.6)  RP11-13M9  357  5 (1.4)  336 (94.1)  15 (4.2)  1 (0.3)  RP11-1150018  302  5 (1.7)  288 (95.4)  7 (2.3)  2 (0.7)  RP11-963J21  339  7 (2.1)  322 (95.0)  9 (2.7)  1 (0.3)  POF54  POF58  POF62  'probes are listed in order from Xptel ---+ Xptel  Figure 4.1 Example of FISH results evaluated in microarray follow-up. Spectrum red labelled Androgen receptor (AR) FISH probe hybridized to metaphase cells from primary amenorrhea patient POF54 and counterstained with DAPI.  171  35% 30% 25% 20% 15% 10% 5% 0%  ■ controls N=97 O secondary amenorrhea patients N=47  Ii  1  50-59^60-69^70-79^80-89^90-100 Degree of skewed XCI (AR)  Figure 4.2 Degree of skewed XCI in secondary POF patients and controls.  172  POF 54  GT  -0.5^0.5  ^  POF 58  POF 62  -0.5^0.5  -0.5^0.5  ^  ^  X-2  -0.5^0.5  =231 4.2222  45231  42222  :  Xp22.13 V.12 422.11 Xp212 4212 Xp21.1  xpitt xpila xpl123 Xp11 22 Xq1121 X1)11.1 X,311.1 X911 2 Xq12 Xq13.1 Yn11,  Xq 133 Xq21.1  X4212 Xq21.31 Xq21.32 Xq21.33 Xq22.1 X0222 Xq223  xq 23 xq23 Xq25 Xq26.1 Xq262 Xq263 Yr17? I Xg2t 2 xq27  xqx  Figure 4.3 DNA microarray data for select POF patients with skewed XCI. Male reference DNA was used for comparison; therefore baseline values for normal copy number complement on the X chromosome are shifted to +0.25. Regions followed up by FISH are indicated with an asterisk. POF 54 and POF 58 are primary amenorrhea patients with skewed XCI of 100% and 90% respectively; POF 62 is secondary amenorrhea patient with 97% skewed XCI; X-2 is an individual with an X deletion and duplication (46,XX,inv dup(X)(pter—>q13.3::q21.3—>pter)) shown as a control to indicate the degree of shift expected in cases of copy number gain or loss.  4.5 References Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM and Belmont JW. 1992. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgenreceptor gene correlates with X chromosome inactivation. Am. J. Hum. Genet. 51:1229-1239. Anasti JN. 1998. Premature ovarian failure: an update. Fertil. Steril. 70:1-15. Beever C, Lai BP, Baldry SE, Penaherrera MS, Jiang R, Robinson WP and Brown CJ. 2003a. Methylation of ZNF261 as an assay for determining X chromosome inactivation patterns. Am. J. Med. Genet. A. 120:439-441. Beever CL, Stephenson MD, Penaherrera MS, Jiang RH, Kalousek DK, Hayden M, Field L, Brown CJ and Robinson WP. 2003b. Skewed X-chromosome inactivation is associated with trisomy in women ascertained on the basis of recurrent spontaneous abortion or chromosomally abnormal pregnancies. Am. J. Hum. Genet. 72:399-407. Bione S, Benedetti S, Goegan M, Menditto I, Marozzi A, Ferrari M and Toniolo D. 2006. Skewed X-chromosome inactivation is not associated with premature ovarian failure in a large cohort of Italian patients. Am. J. Med. Genet. A. 140:1349-1351. Bretherick K, Gair J and Robinson WP. 2005. The association of skewed X chromosome inactivation with aneuploidy in humans. Cytogenet. Genome Res. 111:260-265. Brown CJ and Robinson WP. 2000. The causes and consequences of random and nonrandom X chromosome inactivation in humans. Clin. Genet. 58:353-363. Cheng EY, Chen YJ, Bonnet G and Gartler SM. 1998. An analysis of meiotic pairing in trisomy 21 oocytes using fluorescent in situ hybridization. Cytogenet. Cell Genet. 80:48-53. Chi B, DeLeeuw RJ, Coe BP, MacAulay C and Lam WL. 2004. SeeGH--a software tool for visualization of whole genome array comparative genomic hybridization data. BMC Bioinformatics. 5:13. Conrad DF, Andrews TD, Carter NP, Hurles ME and Pritchard JK. 2006. A highresolution survey of deletion polymorphism in the human genome. Nat. Genet. 38:75-81. Conway GS. 2000. Premature ovarian failure. Br. Med. Bull. 56:643-649. Davis CJ, Davison RM, Payne NN, Rodeck CH and Conway GS. 2000. Female sex preponderance for idiopathic familial premature ovarian failure suggests an X chromosome defect: opinion. Hum. Reprod. 15:2418-2422. Davis SR. 1996. Premature ovarian failure. Maturitas. 23:1-8.  174  Davison RM, Davis CJ and Conway GS. 1999. The X chromosome and ovarian failure. Clin. Endocrinol. (Oxf). 51:673-679. Devi AS, Metzger DA, Luciano AA and Benn PA. 1998. 45,X/46,XX mosaicism in patients with idiopathic premature ovarian failure. Fertil. Steril. 70:89-93. Devi A and Benn PA. 1999. X-chromosome abnormalities in women with premature ovarian failure. J. Reprod. Med. 44:321-324. Devlin L and Morrison PJ. 2004. Mosaic Down's syndrome prevalence in a complete population study. Arch. Dis. Child. 89:1177-1178. Eichler EE. 2006. Widening the spectrum of human genetic variation. Nat. Genet. 38:911. Goswami D and Conway GS. 2005. Premature ovarian failure. Hum. Reprod. Update. 11:391-410. Hassold T, Abruzzo M, Adkins K, Griffin D, Merrill M, Millie E, Saker D, Shen J and Zaragoza M. 1996. Human aneuploidy: incidence, origin, and etiology. Environ. Mol. Mutagen. 28:167-175. Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, Coe BP, Snijders A, Albertson DG, Pinkel D, Marra MA, et al. 2004. A tiling resolution DNA microarray with complete coverage of the human genome. Nat. Genet. 36:299-303. Jong K, Marchiori E, Meijer G, Vaart AV and Ylstra B. 2004. Breakpoint identification and smoothing of array comparative genomic hybridization data. Bioinformatics. 20:3636-3637. Kalantaridou SN, Davis SR and Nelson LM. 1998. Premature ovarian failure. Endocrinol. Metab. Clin. North Am. 27:989-1006. Khojasteh M, Lam WL, Ward RK and MacAulay C. 2005. A stepwise framework for the normalization of array CGH data. BMC Bioinformatics. 6:274. Lau AW, Brown CJ, Penaherrera M, Langlois S, Kalousek DK and Robinson WP. 1997. Skewed X-chromosome inactivation is common in fetuses or newborns associated with confined placental mosaicism. Am. J. Hum. Genet. 61:1353-1361. Lyon MF. 1961. Gene action in the X chromosome of the mous (Mus musculus L.). Nature. 190:372-373. McCarron SA, Hadnott TN, Perry GH, Sabeti PC, Zody MC, Barrett JC, Dallaire S, Gabriel SB, Lee C, Daly MJ, et al. 2006. Common deletion polymorphisms in the human genome. Nat. Genet. 38:86-92. Nelson LM, Covington SN and Rebar RW. 2005. An update: spontaneous premature ovarian failure is not an early menopause. Fertil. Steril. 83:1327-1332. 175  Penaherrera MS, Barrett IJ, Brown CJ, Langlois S, Yong SL, Lewis S, Bruyere H, Howard-Peebles PN, Kalousek DK and Robinson WP. 2000. An association between skewed X-chromosome inactivation and abnormal outcome in mosaic trisomy 16 confined predominantly to the placenta. Clin. Genet. 58:436-446. Robinson WP, Barrett IJ, Bernard L, Telenius A, Bernasconi F, Wilson RD, Best RG, Howard-Peebles PN, Langlois S and Kalousek DK. 1997. Meiotic origin of trisomy in confined placental mosaicism is correlated with presence of fetal uniparental disomy, high levels of trisomy in trophoblast, and increased risk of fetal intrauterine growth restriction. Am. J. Hum. Genet. 60:917-927. Robinson WP, Beever C, Brown CJ and Stephenson MD. 2001. Skewed X inactivation and recurrent spontaneous abortion. Semin. Reprod. Med. 19:175-181. Sato K, Uehara S, Hashiyada M, Nabeshima H, Sugawara J, Terada Y, Yaegashi N and Okamura K. 2004. Genetic significance of skewed X-chromosome inactivation in premature ovarian failure. Am. J. Med. Genet. A. 130:240-244. Seltzer GB, Schupf N and Wu HS. 2001. A prospective study of menopause in women with Down's syndrome. J. Intellect. Disabil. Res. 45:1-7. Simpson JL and Rajkovic A. 1999. Ovarian differentiation and gonadal failure. Am. J. Med. Genet. 89:186-200. Stavropoulos DJ, Bick D and Kalousek DK. 1998. Molecular cytogenetic detection of confined gonadal mosaicism in a conceptus with trisomy 16 placental mosaicism. Am. J. Hum. Genet. 63:1912-1914. Toniolo D. 2006. X-linked premature ovarian failure: a complex disease. Curr. Opin. Genet. Dev. 16:293-300. Uehara S, Hashiyada M, Sato K, Sato Y, Fujimori K and Okamura K. 2001. Preferential X-chromosome inactivation in women with idiopathic recurrent pregnancy loss. Fertil. Steril. 76:908-914. Vegetti W, Marozzi A, Manfredini E, Testa G, Alagna F, Nicolosi A, Caliari I, Taborelli M, Tibiletti MG, Dalpra L, et al. 2000. Premature ovarian failure. Mol. Cell. Endocrinol. 161:53-57. Wong KK, deLeeuw RJ, Dosanjh NS, Kimm LR, Cheng Z, Horsman DE, MacAulay C, Ng RT, Brown CJ, Eichler EE, et al. 2007. A comprehensive analysis of common copy-number variations in the human genome. Am. J. Hum. Genet. 80:91-104.  176  Chapter 5: Age-related chromosome factors and POF 4 5.1 Introduction Both human epidemiological studies and animal models provide support for the suggestion that longevity is associated with an increase in reproductive lifespan. In humans, higher total fecundity, albeit an indirect measure of reproductive lifespan, is positively correlated with longevity (Manor et al. 2000; Muller et al. 2002). A woman's age at the birth of her last child is correlated with overall lifespan and increased postreproductive survival (Doblhammer. 2000; Helle et al. 2005; Muller et al. 2002; Perls et al. 1997; Smith et al. 2002). Late menopause, a more direct measure of reproductive senescence, is associated with decreased mortality and increased post-reproductive lifespan (Cooper and Sandler. 1998; Jacobsen et al. 1999; Snowdon et al. 1989). In both mice (Nagai et al. 1995) and flies (Hutchinson and Rose. 1991), selection for females with the capacity for later reproduction results in an overall increase in years of life, supporting a role for genetic factors in this process. In contrast, the 'disposable soma theory' of aging predicts that there is a trade-off between fertility and longevity such that female lifespan is negatively correlated with reproductive capacity (Kirkwood and Rose. 1991). This is supported by studies in both C. elegans (Hsin and Kenyon. 1999) and human populations (Westendorp and Kirkwood. 1998). These findings do not necessarily contradict the hypothesis that reproductive lifespan is associated with longevity, rather they highlight the necessity for a genetic component in the model. If the  Data in this chapter is included in a manuscript that is being prepared for publication: Hanna CW, Bretherick KL, Gair JL, Fluker MR, Stephenson M, Lansdorp P, Robinson WP. 2008. Telomere length and age-related infertility in women (in preparation). This chapter also contains data on XCI inactivation skewing, AR methylation, and APOE genotype that will not be included in this publication. 4  177  same genes contribute to both extreme lifespan and reproductive longevity, they would allow the trade off between somatic maintenance and reproduction to occur later in life, predicting that the exceptionally aged would, in fact, have delayed reproductive senescence (Penis and Fretts. 2001). There are two general explanations for the correlation between lifespan and reproductive longevity. Firstly, hormones produced by a functional ovary may be important in prolonging lifespan. This is supported by the finding of increased lifespan in mice undergoing transplantation with ovaries obtained from younger mice (Cargill et al. 2003). Alternatively, the rate of somatic cell aging and ovarian aging may be correlated due to common underlying factors; extreme longevity may not have been directly selected for, but may rather be a consequence of selection for genes that maximize reproductive years (Perls et al. 2002; Perls and Fretts. 2001). As individuals age, a number of chromosomal changes take place which reflect the aging process. In human blood, cellular markers associated with aging include: decreased telomere length, increased X-inactivation skewing in females, and altered DNA methylation. Telomere length decreases with age, but is highly variable within a group of individuals of the same age. This variability is a product of differences in the rate of cell turnover, original telomere length at conception, and telomerase activity during early development. Telomere length is negatively associated with smoking (Valdes et al. 2005), poor social status (Cherkas et al. 2006) and both oxidative and life stress (Epel et al. 2004; Kotrschal et al. 2007). In addition, there is a genetic influence on telomere length, with twin studies estimating the heritability to be 36-84% (Andrew et al. 2006; Jeanclos et al. 2000; Slagboom et al. 1994). However, these estimates may be inflated  178  due to an inability to extricate the shared pre- and post-natal environmental effects (Gilley et al. 2008), which may be the primary factors responsible for correlation in telomere length in twin pairs (Huda et al. 2007). Telomere length is also associated with paternal age (Unryn et al. 2005), presumably by influencing initial telomere length at conception, which could also partially explain the increased correlation in twins. Regardless of additional influences, in all individuals telomere length decreases with age due to mitotic cell division. The telomere theory of senescence proposes that when telomeric DNA falls below a critical length, chromosome integrity is compromised, resulting in cell death (Harley et al. 1992). This limitation on replicative capacity could limit the number of cell divisions possible for a specific cell type, such as primordial germ cells, and could therefore determine reproductive lifespan (Aydos et al. 2005). Skewed X chromosome inactivation (XCI) is a detectable bias in the proportion of cells with either the paternal or maternal X chromosome inactivated (see Chapters 1 and 4). The frequency of skewed X chromosome inactivation, particularly in blood, increases with age (Busque et al. 1996; El Kassar et al. 1998; Gale et al. 1997; Hatakeyama et al. 2004; Sharp et al. 2000; Tonon et al. 1998). Explanations for the increase in skewed XCI with age include selective differences over time and exhaustion of the hematopoietic stem cell pool (Brown and Robinson. 2000). Both of these phenomena can be viewed as indirect indicators of the rate of cell division, since selective differences would become more pronounced with increased cell division and stem cell pool would exhaust after a finite number of mitotic events. DNA methylation of CpG residues occurs throughout the genome, both in nongenic regions where it is associated with a heterochromatic chromatin state, and at CpG  179  islands in gene promoters where it is responsible for gene silencing. As individuals age there is global demethylation of the genome in a number of species including humans (Golbus et al. 1990; Richardson. 2003) although it is unclear whether these methylation changes occur in gene promoters or non-genic regions (Richardson. 2003). Age-related changes in methylation of CpG islands in gene promoters appear to be gene specific. Some gene promoters (including tumor suppressor genes) become hypermethylated (Issa et al. 1994), while others, such as those associated with retrotransposons and endogenous retroviruses, become demethylated (Barbot et al. 2002). DNA methylation changes with age may be the result of reduced expression of DNA methyltransferases (Xiao et al. 2007; Zhang et al. 2002), or environmental or dietary influences (Cooney. 2001). Oocyte maturation and fertilization involves a number of chromatin reprogramming events; if the cellular processes involved in DNA methylation are affected by age, these changes might also affect reproductive senescence. Furthermore, the increase in CpG island methylation in certain tissues (including endometrium) has been proposed as a somatic cell "molecular clock" because it is correlated with age and expected rate of cell division (Chu et al. 2007; Kim et al. 2005). If there is a relationship between somatic aging and reproductive aging, genes that influence lifespan may be associated with reproductive longevity. The single gene most commonly associated with aging is the apolipoprotein E gene, APOE. Apolipoprotein E is an important factor in lipid metabolism and the APOE gene has three common isoforms e2, e3, and e4. The E4 allele is associated with a moderate increase in yearly mortality due to increased risk of Alzheimer's disease, dyslipidemia, and coronary heart disease, while the APOE2 allele has a protective effect (Christensen et al. 2006). The 64  180  allele has also been associated with increased risk for trisomic pregnancy (Avramopoulos et al. 1996; Nagy et al. 2000) a reproductive phenomenon that increases in prevalence with female age (Hassold et al. 1996). APOE s4 may influence risk of trisomy by increasing the rate of cell death in the ovary. Cell death could result from oxygen deficit at the ovary as a result of atherosclerosis (Avramopoulos et al. 1996), negative effects of oxidative stress influenced by APOE2 genotype, or meiotic spindle disruption by specific APOE isoforms (Nagy et al. 2000). An increased rate of cell death within the ovary could result in a reduction in follicular pool size and lead to premature ovarian senescence. Women with premature ovarian failure (POF) may be considered "prematurely aged" (Pal and Santoro. 2002). Telomere length, degree of skewed XCI, and DNA methylation are indicators of cell division and may therefore be correlated with one another. If onset of reproductive senescence is related to rate of cell division, these factors would also be associated with the presence of POF. In order to determine if there is a link between markers of biological aging and reproductive senescence, we assayed telomere length, X chromosome inactivation skewing, and DNA methylation at the Androgen Receptor (AR) gene CpG island in POF patients and control women. In addition, APOE was genotyped in POF patients and controls to assess whether this agerelated gene also plays a role in ovarian senescence.  5.2 Methods 5.2.1 Samples Collection of samples for this study was approved by the University of British Columbia clinical ethics review board, approval number H01-70460 (Appendix 2). 181  Women with idiopathic POF (N=56) were recruited from the POF clinic at BC Women's Hospital. POF diagnosis was made based on criteria outlined in previous chapters. Exact age at the time of blood draw was known for 37 of the POF patients and the average age among these was 36.2±6.5years (range 21-57). Two control groups were used in this study, as described in previous chapters. Control group 1 (N=109) consisted of healthy women from the general population for whom reproductive history is unknown. The average age of this group at the time of blood draw was 36.3 years (range 17-55). Additional controls (N=29) meeting the same criteria were added to this group to assess a broader range of ages for some age-related factors. These consisted of 13 subjects between ages 2 and 19, and 16 subjects between ages 57 and 81. Control Group 2 (n = 47) consisted of women who had had a healthy pregnancy after the age of 37 years and had not experienced any pregnancy loss. Average age at time of blood draw for women in Control Group 2 was 41.9 years (range 37-58).  5.2.2 Telomere length analysis Relative telomere length was determined using the method described by (Cawthon. 2002). Telomere repeat copy number was determined by quantitative real time PCR (qPCR) amplification of the telomeric repeat region and expressed relative to a single copy gene (36B4, encoding acidic ribosomal phosphoprotein PO, located on chromosome 12) copy number determined by the same means in a parallel reaction (Figure 5.1). This telomere to single copy (T/S) ratio should be proportional to the average telomere length of the sample.  182  Telomeric repeats and single copy genes were amplified using a home-made PCR master mix due to the specific reaction conditions required for amplification of this highly repetitive region. PCR reactions were performed on 96 well plates in 20uL reactions. Each well contained 5ng genomic DNA (0.5uL of a dilution of 1 Ong/uL in TE-4 ), lx AmpliTaq GOLD Buffer (Applied Biosystems), 2mM MgC1 2 (Applied Biosystems), 0.2mM each dNTP (Invitrogen), 5mM Dithiothreitol (Fisher), 0.5uM ROX Reference Dye (Invitrogen), 0.2uL 20X SYBR Green (diluted in DMSO from 10 000x SYBR Green I nucleic acid gel stain, Invitrogen), 0.65 units Amplitaq GOLD (Applied Biosystems), and either 13.5nM tel 1 (5'-GGT TTT TGA GGG TGA GGG TGA GGG TGA GGG TGA GGG T-3') and 45nM tel 2 (5'-TCC CGA CTA TCC CTA TCC CTA TCC CTA TCC CTA TCC CTA-3') for telomere amplification, or 15nM 36B4u (5'CAG CAA GTG GGA AGG TGT AAT CC-3') and 25nM 36B4d (5'-CCC ATT CTA TCA TCA ACG GGT ACA A-3') for single gene amplification. Amplification conditions were 50°C for 2 minutes, 95°C for 10 minutes, 30 cycles of (95°C for 15 seconds and 54°C for 2 minutes) for telomere repeats and 50°C for 2 minutes, 95°C for 10 minutes, 40 cycles of (95°C for 15 seconds and 58°C for 1 minute) for single copy genes. Each amplification was followed by a dissociation gradient from 95°C to 60°C to verify amplification of a single DNA species (Figure 5.2a). Relative telomere and single gene quantity was determined based on a standard curve composed of a single sample freshly diluted in TE -4 to lOng, 5ng, 2.5ng, 1.25ng, and 0.625ng. Short and long controls were run with each plate, to ensure consistency between runs. Amplifications for each sample, standard, and control were performed in triplicate. Average quantity of the triplicates for telomere amplification was divided by average quantity of the triplicates for single gene  183  amplification to determine relative telomere repeat length. All samples were assayed at least twice, on separate plates, with separate DNA dilutions. The assay reproducibility is shown in Figure 5.2b (R 2 =0.19, p<0.0001, one-tailed test for significance of the correlation based on the t-distribution). In cases where the standard deviation of the two replicates was greater than 0.25, samples were assayed a third and occasionally fourth time and an average of all results was used in further analyses. To assess the validity of this assay, the relative telomere repeat length values were compared to telomere restriction fragment lengths obtained by Southern blot. Southern blot has historically been the gold standard for telomere length analysis, however it has several limitations and is rapidly being replaced by more sophisticated, high-throughput techniques. A subset of samples from the POF patient group and control group 2 (N=25) for which there was sufficient DNA, were assayed by both quantitative PCR and Southern blot. Southern blot for telomere restriction fragment length was performed as described in the Te1oTAGGG telomere length assay kit (Roche), blots were imaged on a Lumi-Imager, and averaged fragment length was determined with Quantity One software (BioRad). Relative telomere repeat length and telomere restriction fragment length were weakly, but significantly, correlated (Figure 5.2c, R 2 =0.4, p=0.0003, one-tailed test for significance of the correlation based on the t-distribution). The lack of a strong correlation between these two assays may be due in part to inter-individual variability in subtelomeric regions that are assayed as part of the telomere restriction fragment as well the intrinsic technical variability in each of the assays.  184  5.2.3 X chromosome inactivation skewing The degree of XCI skewing was determined by assaying methylation at a polymorphic CAG repeat at the Androgen Receptor (AR) locus (Allen et al. 1992) as described in Chapter 2.  5.2.4 DNA methylation at Androgen Receptor To assess methylation, I designed and optimized a quantitative DNA methylation assay for the Androgen Receptor (AR) CpG island. Genomic DNA was digested with a methylation-sensitive restriction enzyme and amplified by quantitative PCR (qPCR) with primers that flank two restriction sites at AR. Quantity determined at AR was normalized to quantity at a region of the XIST gene that lacked restriction cut sites and was used to control for the amount of input DNA. The normalized quantity of digested DNA is expressed as a percentage of the normalized quantity of the undigested sample assayed in parallel, to determine the percent DNA that is methylated (and therefore uncut) at both AR cut sites (Figure 5.3). For each digest, 750 ng of genomic DNA was incubated over night at 37°C with lx NEB buffer #1 (New England Biolabs) 23 units of HpaII (New England Biolabs), and 10 units of RsaI (New England Biolabs), in a 50uL reaction. An undigested sample with an equal quantity of DNA was incubated with RsaI but not HpaII. MIC2 PCR amplification was performed as described in Chapter 2, to ensure complete enzyme digestion. Following digestion, samples were purified with the Wizard DNA clean-up system (Promega) according to manufacturer's instructions, and eluted in 30uL of 75°C H20. Cleaned digestion products were re-quantified by spectrophotometer to ensure they fell within the standard curve used in qPCR analysis. 185  Quantity of the digested and undigested samples remaining intact at the Androgen Receptor (AR) locus was determined by qPCR amplification on an ABI 7000 using TaqMan primers and probes. Amplifications were performed on 96 well plates. Each 20uL reaction contained luL purified digested DNA sample, lx TaqMan Universal PCR Master Mix (Applied Biosystems), and either 1.8uM ARF (5'-TGC GCG AAG TGA TCC AGA A-3') , 1.8uM AR(s)R (5'-CTG CAG CAG CAG CAA ACT G-3') and 0.1uM AR probe (5'-6FAM-AGG CAC CCA GAG GC-MGB-NFQ-3'), for AR amplification, or 0.9uM XRTF (5'TCT CAA GGC TTG AGT TAG AAG TCT TAA G3'), 0.9uM XRTR (5'-CGT TGG CCT TGT GTC ACA AGT-3') and 0.1uM XIST probe (5'-VIC-CTG GGA CAG GAC ACA TG-MGB-NFQ-3') for XIST amplification. AR and XIST amplifications were performed simultaneously on the same plate with the following cycling conditions: 50°C for 2 minutes, 95°C for 10 minutes, and 40 cycles of (95°C for 15 seconds, 60°C for 1 minute). Each plate was run with a standard curve consisting of undigested genomic DNA in concentrations of 100, 50, 25, 5, and lng/uL in duplicate. Average quantity of DNA uncut at AR for each set of triplicates was determined based on the standard curve and normalized by average quantity at XIST to adjust for quantity of input DNA. Percent methylation was calculated as normalized AR quantity of digested / normalized AR quantity of undigested for each sample. To assess the validity of this assay the linearity, reproducibility, and percent methylation in controls was determined. The linearity of this assay was assessed by determining the methylation status of mixes of digested and undigested male DNA. In a male, the single X chromosome is active and unmethylated; therefore nearly all X chromosome DNA should be cut by methylation sensitive restriction digestion. Mixes of  186  0%, 10%, 25%, 50%, 75%, 90%, 100% ratios of digested and undigested male DNA were made and the percent methylation assayed by this quantitative method was determined in each mix. The assay was linear across the range of methylation from 10100% as shown in Figure 5.4a (R 2 =0.99, p<0.0001, one-tailed test for significance of the correlation based on the t-distribution). The slope of 0.90 suggests that this assay underestimates methylation, and all methylation results were therefore corrected by a factor of 0.90. A number of samples were assayed twice and the assay was poorly but significantly reproducible (R 2 =0.20, p=0.002, one-tailed test for significance of the correlation based on the t-distribution, Figure 5.4b). Finally, control DNA from mousehuman hybrid cell lines (known to carry either a single active or inactive human X chromosome) and human placental tissue (known to be hypomethylated) was assessed for methylation at AR using this assay and the expected level of methylation was seen (Figure 5.4c).  5.2.5 Genotyping FMR1 and ESR1 allele sizes were determined by fluorescent PCR amplification and fragment analysis on an ABI 310 genetic analyzer as described in Chapters 2 and 3. FMR1 biallelic mean and genotype repeat size were calculated as described in Chapter 2. APOE genotype was determined in patients and controls as described by (Wenham et al. 1991) and shown in Figure 5.5. Approximately 25 ng of genomic DNA was amplified in a 15piL PCR reaction containing: lx PCR buffer with MgC12 (Rose Scientific), 0.2Mm each dNTP, 1.5uL DMSO, 1.0 uM each APOE F (5'-TCC AAG GAG CTG CAG GCG GCG CA-3') and APOE R (5'-ACA GAA TTC GCC CCG GCC TGG TAC ACT GCC A-3') primers, and 0.25U Taq DNA polymerase (Rose Scientific).  187  Cycling conditions were 94°C for 5 minutes followed by 40 cycles of 94°C for 30 seconds 65°C for 30 seconds and 70°C for 90 seconds, and finished with an 10 minute extension at 70°C. 12uL of each PCR product was then digested by overnight incubation at 37°C with lx NEB Buffer #4 (New England Biolabs), 0.lug Bovine Serum Albumin (New England Biolabs), and 0.15 U HhaI (New England Biolabs) in a 15uL reaction. Approximately 8uL of digest was mixed with an equivalent amount of Urea loading buffer, loaded on an 8% polyacrylamide sequencing gel and run at 45W for 30 minutes. Bands were visualized by silver-staining: gels were fixed for 15 minutes in a 10% acetic, 10% methanol solution; washed for 10 minutes in 10% ethanol; cleaned with a 30 second rinse in 0.5% nitric acid; stained by shaking for 25 minutes in 0.012M silver-nitrate; developed by washing in 0.28M sodium bicarbonate with 0.05% formaldehyde until bands appeared; and fixed in 10% acetic acid. APOE genotypes were assigned based on banding patterns as diagrammed in Figure 5.5b.  5.3 Results 5.3.1 Telomere length There was significant variability in average relative telomere length between individuals and an overall decrease in relative telomere length with age in individuals aged 17-59 years (Figure 5.6). This decrease was significant for both Control group 1 (R2 =0.07, p=0.002) and POF patients (R 2 =0.17, p=0.007), but not for Control group 2 (R2=0.002, p=0.39) (one-tailed tests for significance of the correlation based on the tdistribution). Surprisingly, relative telomere length was longer in the POF patient group than in controls. In data not adjusted for age, and only including individuals between the ages of 188  17 and 55, average relative telomere length of the POF patient group was 0.95±0.18 as compared to 0.90±0.16 in Control group 1 (p=0.05, two-tailed t-test) and 0.89±0.14 in Control group 2 (p=0.08, two-tailed t-test; p=0.03 vs. combined controls, two-tailed ttest; Figure 5.7). Furthermore, age-adjusted mean telomere length was also significantly longer for POF patients than controls (p=0.04, ANCOVA, Table 5.1). Estrogen exposure may positively affect telomere length (Aviv et al. 2005; Lee et al. 2005; Sato et al. 2004b). Since genetic variants may influence cellular responses to estrogen, and ESRI genotype in particular has been associated with POF (Chapter 3; Bretherick et al. 2008) correlation between telomere length and ESRI genotype was examined. There was no association between telomere length and ESRI repeat genotype in POF patients, Control group 1, Control group 2, or all patient and control groups combined (Table 5.2). A recent report has suggested that male FMRI premutation carriers with and without Fragile X tremor ataxia syndrome (FXTAS) have shorter telomeres than agematched controls (Jenkins et al. 2007). I therefore evaluated whether telomere length was correlated with FMRI repeat length. There was no correlation between telomere length and FMRI allele size in either POF patients, Control group 1, Control group 2, or in all samples combined, when considering length of the long FMRI allele (Figure 5.8), FMRI biallelic mean, or FMRI genotype repeat size calculated at the AR or FMRI loci (Table 5.3). Four premutation (>50 CGG repeats) carriers were present in the POF patient and control groups, and the average telomere length of these four samples was no different from those with FMRI repeats within the normal range (0.95±0.16 vs. 0.91±0.17 respectively, p=0.6, two tailed t-test). Furthermore, there was no difference in telomere  189  length between individuals with an FMR1 allele of >35 repeats and those without (0.93±0.16 vs. 0.91±0.17 respectively, p=0.5, two tailed t-test). Thus, there is no evidence that telomere length is in anyway associated with FMRI repeat length in these samples.  5.3.2 X chromosome inactivation skewing Predictably, there was significant variability in degree of X chromosome inactivation (XCI) skewing between individuals and an overall increase in skewing with age in individuals aged 2-81 years (Figure 5.9). This correlation was significant for POF patients (R 2 =0.13, p=0.02), but not for Control group 1 (R 2 =0.1, p=0.09) 5 or Control group 2 (R2 =0.006, p=0.3) (one-tailed tests for significance of the correlation based on the t-distribution). Notably, the rate of increase in skewed XCI appears to be more rapid in POF patients than controls (Figure 5.9), although this difference is not statistically significant (p=0.17, ANCOVA test for homogeneity of regressions). There was no difference in the frequency of skewed XCI between POF patients and controls. 6 In the POF patient group 3 of 46 patients (6.1%) had skewed XCI >90%, whereas in Control groups 1 and 2, 9 of 95 (9.4%) and 3 of 38 (7.9%) women had XCI skewing above this cut off.  5.3.3 AR DNA methylation The Androgen Receptor gene, AR, is found on the X chromosome and due to X chromosome inactivation the expected level of DNA methylation at the AR promoter is 50%. This quantitative methylation assay suggests that deviation from the expected 50% 5 Control group 1 is comprised of a subset of samples for which the relationship between skewed XCI and age has been previously described (Hatakeyama et al. 2004). 6 This data is described in Chapter 4 and has been previously published (Bretherick et al. 2007).  190  is common at the AR locus, with methylation ranging from 15% to 77% in normal women from Control group 1 (Figure 5.10). Degree of methylation at the Androgen receptor promoter increases with age in individuals aged 2-81 years (Figure 5.10). This increase was significant for Control group 1 (R 2 =0.09, p=0.04) but not for POF patients (R2 =0.008, p=0.7) (two-tailed tests for significance of the correlation based on the tdistribution). Control group 2 had a non-significant trend towards decreased AR methylation with age (R 2 =0.001, p=0.9, two-tailed test for significance of the correlation based on the t-distribution). There was no difference in AR methylation between the POF patients and controls. In data not adjusted for age, and only including individuals between the ages of 17 and 55, average AR methylation of the POF patient group was 46.3±10.2% as compared to 42.1±11.6% in Control group 1 (p=0.3, two-tailed t-test) and 40.6±12.0% in Control group 2 (p=0.09, two-tailed t-test; p=0.09 vs. combined controls, two-tailed ttest; Figure 5.11). There was a trend toward higher AR methylation in the POF patients than in controls, however this difference remained non-significant when data was adjusted for age (p=0.23, ANCOVA, Table 5.4). 5.3.4 Correlation between age-related chromosome factors If differences in telomere length, XCI skewing, and AR methylation reflect differences in the number of cell divisions, these factors should be correlated within an individual. Telomere length was positively correlated with degree of XCI skewing in POF patients (R 2 = 0.06, p=0.05, one-tailed test for significance of the correlation based on the t-distribution, Figure 5.12, Table 5.5). Although both these variables change with age, and when age is controlled for the correlation is not significant (p=0.09, multiple  191  regression ANOVA). There was no correlation between telomere length and AR methylation in either POF patients, Control group 1, Control group 2, or all samples combined (Figure 5.13, Table 5.5). XCI skewing was assessed indirectly in this study by assaying methylation at the AR locus on the X chromosome. If the deviation from the expected 50% AR promoter methylation observed in this study occurs in an allele specific manner, alterations in AR methylation could influence XCI skewing assayed at the AR locus. There was no correlation between XCI skewing and AR methylation in any group or all samples combined (Figure 5.14, Table 5.5), suggesting that in individuals with variations from the expected 50% methylation at AR, the loss or gain of methylation does not occur in an allele specific manner. 5.3.5 APOE genotype APOE allele and genotype frequencies in POF patients, Control group 1 and Control group 2 are shown in Figure 15.5. There was no difference in overall allele distribution between POF patients and either control group, or both control groups combined (p=0.5, p=0.4 for comparisons to Control groups 1 and 2, respectively, p=0.9 for comparison to combined controls; x2 contingency test). Specifically, there was no increase in frequency of the age-associated APOE e4 allele in the POF patient group (p=0.1, 0.2 for comparisons to Control groups 1 and 2, respectively, p=0.4 for comparison to combined controls; one-tailed x2 test). Likewise, there was no difference in frequency of the APOE E4/ c4 genotype in the POF patient group (p=0.6, p=0.6 for comparisons to Control groups 1 and 2, respectively, p=0.7 for comparison to combined controls; one-tailed Fisher's exact tests). APOE genotype does not appear to be associated with risk for POF in this sample.  192  5.4 Discussion In order to identify if there is a link between markers of biological aging and reproductive senescence, age-related chromosome factors including telomere length, X chromosome inactivation, and gene-specific methylation were assessed in women experiencing POF and controls. Average relative telomere length decreased with age, as has been well established previously with other methods (Benetos et al. 2001; Slagboom et al. 1994). Telomere shortening with age is a result of the cumulative effects of mitotic cell divisions in which DNA polymerase fails to replicate to the end of each chromosome and a portion of the telomeric repeat sequence is lost. The degree of X chromosome inactivation skewing was found to increase with age, also confirming several previous cross-sectional and longitudinal studies (Busque et al. 1996; El Kassar et al. 1998; Gale et al. 1997; Sandovici et al. 2004). Explanations for the increase in skewed XCI with age include accumulation of selective differences over time and exhaustion of the hematopoietic stem cell pool (Christensen et al. 2000; Hatakeyama et al. 2004; Kristiansen et al. 2005; Sandovici et al. 2004). An alternative explanation, in which an age-related allele-specific loss of methylation at the AR locus causes an apparent increase in XCI skewing, has also been proposed (Hatakeyama et al. 2004). However, the AR promoter CpG sites commonly assayed to infer XCI skewing are the same sites examined in the quantitative methylation assay described in this study (Figure 5.3a). Although the results reported here suggest that significant deviation from the expected 50% AR methylation is common, they indicate that there is an increase, rather than decrease in AR methylation with age. Furthermore, the degree of X  193  chromosome inactivation skewing was not correlated with methylation at AR, suggesting that this mechanism is unlikely to play a role in the observed increase in XCI skewing with age. There is a trend toward an accelerated rate of XCI skewing in the POF patient population as compared to controls, although these results do not reach statistical significance. This trend could be an artifact of the age of the sample groups or a result of an increased rate of selection against cells activating an abnormal X chromosome in the POF patient group (Sato et al. 2004a), however results previously described in this thesis refute the latter hypothesis (Chapter 4; Bretherick et al. 2007). Alternatively, several groups have reported that an accelerated rate of XCI increase occurs around middle age in normal women (Hatakeyama et al. 2004; Kristiansen et al. 2005; Sandovici et al. 2004) leading to the suggestion that hormonal changes following menopause may contribute to this phenomena (Kristiansen et al. 2005). This would provide an explanation for an accelerated rate of XCI increase in POF patients who have already experienced menopause. DNA methylation at the AR promoter varied between 15 and 77% in normal control women. The variation may be due to the inherent variability in the assay, as is apparent with the large error bars (Figure 5.4b). Variation from the expected 50% methylation may also suggest that in some women (with methylation below 50%), there is a loss of methylation of the inactive X chromosome; while in others (with methylation above 50%), there is methylation of the active X chromosome. Although methylation at AR has been correlated with X chromosome inactivation for the purpose of determining  194  skewing patterns (Allen et al. 1992), this does not preclude these losses or gains in methylation at these sites, provided that they do not occur in an allele-specific manner. DNA methylation at AR increased with age in controls aged 2 to 81. An increase in AR promoter methylation may reflect changes in AR expression with age. DNA methylation of the AR promoter CpG island is correlated with loss of AR expression (Jarrard et al. 1998), likely via a mechanism involving blocking of transcription factor binding or changing chromatin structure to inhibit transcription factor accessibility (Kumar and Thakur. 2004). Regulation of AR expression is complex and varies by sex, tissue type, and developmental stage. In some cases it appears to be regulated by circulating sex steroid level (Kumar and Thakur. 2004; McAbee and Doncarlos. 1999) although in other tissues and developmental stages it is not (Vottero et al. 2006). The age-related increase in AR methylation may be reflecting age-related changes in gene expression that may or may not be regulated by changes in steroid levels. A detailed study examining AR methylation in a larger sample of women across the lifespan is needed to model this change and to postulate further regarding the mechanism. There was no association with degree of XCI skewing, total AR methylation, or APOE genotype and risk for POF in this study. There was however, a significant increase in telomere length in the POF patient population. This result was surprising and suggests that women with POF do not experience premature cellular aging, at least not as substantiated by a reduction in telomere length. Similarly, women with few oocytes retrieved after ovarian stimulation, who likely have "aged ovaries", also have longer telomeres than expected for their age (Dorland et al. 1998a) and long telomeres have been associated with low fertility and fecundity in Drosophila (Walter et al. 2007). In  195  contrast, mothers of children with trisomy 21, have normal telomere length (Dorland et al. 1998b) and women with increased reproductive lifespan have longer telomeres than expected for their age (Aydos et al. 2005). This suggests that the finding of increased telomere length in women with POF is not a common feature of conditions associated with reproductive aging, but rather may be specific to POF. Explanations for the increase in telomere length observed in the POF patient population include a slower rate of cell division and positive influences of estrogen on telomere length. A slower rate of cell division could lead to the establishment of a reduced follicular pool during early development. Fewer cell divisions, perhaps as a result of a prolonged cell cycle, would result in longer telomeres. If this reflects fewer mitotic divisions of the initial germ cells, it could suggest a smaller follicular pool which would result in early menopause (Dorland et al. 1998a). Alternatively, long telomeres in women with POF may be the result of the positive influence of estrogen on telomere length. It is well established that women have longer telomere lengths than age-matched men (Benetos et al. 2001; Jeanclos et al. 2000; Slagboom et al. 1994), a phenomenon that has been attributed to higher estrogen levels in women (Aviv et al. 2005). In addition, telomere lengths are longer in postmenopausal women with a history of long-term hormone replacement therapy, than in women not on hormone replacement (Lee et al. 2005). Two mechanisms by which estrogen may positively regulate telomere length have been proposed. Firstly, estrogen ameliorates the negative effects of reactive oxygen species (Aviv et al. 2005) which reduce telomere length by inducing single strand breaks (von Zglinicki. 2000). Secondly, estrogen may influence telomere length by stimulating telomerase activity (Aviv et al. 2005). Ovarian  196  telomerase activity is reportedly low in POF patients with follicular depletion but high in POF patients with ovarian dysfunction (Kinugawa et al. 2000). Although POF patients undergo early menopause and are therefore in a low-estrogen state earlier in life, prior to onset of ovarian failure women with POF may recruit abnormally large cohorts of follicles with each cycle. This could lead to both higher circulating estrogen levels with a positive influence on telomere length, in addition to a premature reduction in follicular pool. In conclusion, these results provide no evidence to support a link between markers of biological aging and reproductive senescence. They suggest that regulation of age-related chromosome factors including telomere length, X chromosome inactivation, and AR DNA methylation is likely much more complex than rate of cell division. The unexpected finding of elongated telomeres in the POF patient population may reflect an altered rate of cell division or influences of abnormal hormone environment. Further studies are necessary to confirm these findings and elucidate the mechanism behind them.  197  Table 5.1 Age-adjusted mean telomere length in POF patients and controls  N  Mean age  Observed mean telomere length  Adjusted mean telomere length'  POF patients  34  35.2  0.981  0.970  Control group 1  106  36.2  0.898  0.891  Control group 2  46  41.5  0.894  0.917  p=0.04, ANCOVA for comparison of age-adjusted mean telomere length  198  Table 5.2 Average telomere length and ESR1 genotype.  ESR1 genotype  POF patients  Average relative telomere length (N samples) Control group 2 combined groups Control group 1  SS  0.95±0.15 (3)  0.91±0.18 (30)  0.94±0.07 (13)  0.92±0.15 (46)  SL  0.97±0.17 (39)  0.90±0.14 (46)  0.90±0.18 (20)  0.93±0.16 (105)  LL  0.89±0.20 (11)  0.88±0.17 (14)  0.86±0.12 (7)  0.88±0.17 (32)  p value'  0.42  0.80  0.67  0.33  'ANCOVA for comparison of age adjusted mean telomere length -  199  Table 5.3 Correlation between relative telomere length and FMR1 repeat length.  POF patients  Correlation with relative telomere length Control group 1^Control group 2 combined groups  Long repeat N  53  83  47  183  R2  0.0004  0.006  0.03  0.007  0.9  0.5  0.3  0.3  N  53  83  47  183  R2  0.0007  0.003  0.04  0.008  0.9  0.6  0.2  0.2  N  47  72  nd2  119  R2  0.009  0.004  0.0001  0.5  0.6  0.9  N  53  78  R2  0.02  0.004  0  p value'  0.3  0.6  0.9  p value' Biallelic mean  p value' GRS AR  p value' GRS FMR1  nd2  131  'two-tailed test for significance of the correlation based on the t-distribution nd=not done  2  200  Table 5.4 Age-adjusted AR methylation in POF patients and controls Observed mean  Adjusted mean  N  Mean age  AR methylation %  AR methylation % I  POF patients  26  35.8  46.1  46.3  Control group 1  19  35.8  42.1  42.3  Control group 2  24  41.3  40.6  40.2  1 p=0.23,  ANCOVA for comparison of age adjusted mean telomere length -  201  Table 5.5 Correlations between age-related chromosome factors. POF patients  Control group 1  Control group 2  combined groups  N  46  98  38  182  R2  0.05  0.01  0.5  0.002  p value r  0.05  0.1  0.08  0.6  N  25  22  25  72  R2  0.05  0.009  0.07  0  p value 2  0.3  0.7  0.2  0.9  N  23  44  23  90  R2  0.03  0.001  0.08  0.004  p value2  0.5  0.8  0.2  0.6  Telomere length and XCI skewing  Telomere length and AR methylation  XCI skewing and AR methylation  i 2  one-tailed test for significance of the correlation based on the t-distribution two-tailed test for significance of the correlation based on the t-distribution  Telomere associate repeats  3-20kb of TTAGGG repeats Denaturation  5'1  111 111111 1111 I H4__4911 1  11111  ^  1111111 1 11111111  III 4-  4---^4-  Primers 11111111111^ H 11111^1111  .-;11^  11111^11111111111  I  PCP, amplification  1111111^  11I1111IIIII 4-  1  1111111111 1 111 11111111 111111111111111  Quantification NimailiN  =7m■NNMED  c=sommiri  Figure by Courtney Hanna, used with permission  Figure 5.1 Telomere measurement by quantitative PCR (Cawthon 2000). Telomere  repeat sequence is amplified by quantitative PCR using primers that are specific for the TTAGGG repeat. The shortest product will be preferentially amplified and therefore most products will be the length of both primers (76bp). The amount of product is compared to the quantity of product from a single gene amplification and is therefore representative of the relative quantity of telomere repeat sequence present in the original template  203  •  0. 10  A. 0. 0  0. 0e  0.02  -0.02  05  Temperature (C)  B.  •  1.6 -  •• • • • •  zs 1.4 -  • • • t" i •  :.* •* e•• ••-• •i..•00,••••  •rip 1.2'17) 1.0 7, 0.8 cD g 0.6  •  •  .4..• • • •••••^ 4^• • •••• • 4. •••• • •• • • •4s • • .^•  •  0.4 0.4 0.6 0.8 1.0^1.2^1.4^1.6  telomere length first run C.  ,rz?  1.6 -  L ) 1.4 P-1^• 1.2  -  g) 1.0 a)0.8  g  -  0.6 0.4 ^ 5000  7000  9000  11000^13000  Telomere restriction fragment length (Southern blot)  Figure 5.2 Validation of telomere length assay. A) Dissociation analysis from the quantitative PCR reaction shows a single peak. B) Reproducibility of the quantitative PCR assay (R 2 =0.2, p<0.0001). Note that this shows results of the first and second runs, in cases where there was significant discrepancies between these, samples were repeat a third time. C) Comparison of quantification by qPCR and Southern blot (R 2 =0.4, p=0.0003). (one-tailed tests for significance of the correlation based on the t-distribution)  204  ^ A. attcagccaa gctcaaggat ggaagt•cag ttagggctgg gaagggtcta ccctCggcCg A AKF cCgtccaaga cctacCgagg agctttccag aatctgttcc agagCgt otCg Cqaaqtqatc  caqaaccCgg t•ct•ct  gCgagCgcag cacctccCgg Cgccaqtttg agcagca gcagcagcag cagcagcagc agcagcagca gcaagagact  agccccaggc agcagcagca gcagcagggt gaggatggtt ctccccaagc ccatCgt A a4aB ggccccacag gctacctggt cctggatgag gaacagcaac cttcacagcc gcagtcggcc  B.  ••9 9I  '  99 9•, ^• ' • •^ ,  ^Methylation sensitive^Unmethylated CpG ^restriction digest^sites will be cut t=2c=====]  ^qPCR amplification^Uncut DNA will ^ ^of locus of interest be amplified  1-  •• ^• •^ •• '  ,  '  methylation = Quantity of the digested sample (D) Quantity of the undigested sample (U) Figure 5.3 Quantitative DNA methylation assay for AR. A) Region of the AR exon 1 at which methylation is assessed. PCR primers (AR F, AR R) and probe used for quantitative DNA methylation assay are shown in blue, in relation to primers (AR A, AR B) shown in red which encompass the polymorphic CAG repeat (in red) and are used for X chromosome inactivation skewing analysis. CpG sites are capitalized and in bold, and the HpaII cut sites are highlighted in yellow. B) Solid circles represent HpaII sites with methylated CpGs; open circles, unmethylated HpaII sites. Genomic DNA is digested with methylation sensitive enzymes and amplified by quantitative PCR with primers that flank two restriction sites at the locus of interest. The quantity of digested DNA is expressed as a ratio of the quantity of the undigested control, to determine the percent DNA that is methylated at both cut sites.  205  A 100%  0  -  80% -  .21 60% CL)  40% 20% 0% 0%  20%^40%^60%^80% % undigested male DNA in nix  ^  100%  B 100% -to  80% -  a.i  60% -  O  40% 20% 0% 0%  20%^40%^60% AR methylation first run  80%^100%  C  Expected  Observed  methylation  methylation  DNA sample  Description  Xi t48-lalDaz4a  Mouse-human hybrid cell line, human X,  100  100.0±14.4  Xi t11-4Aaz5  Mouse-human hybrid cell line, human X,  100  96.0±18.1  Xa AHA llaBl  Mouse-human hybrid cell line, human X a  0  0.0±0.0  PM MC 1 str3  Human placental stroma  unknown  12.3±1.7  5.4 Validation of AR methylation assay. A) Methylation in mixes of undigested and digested male DNA(R 2 =0.99, p=0.0001). B) Reproducibility of quantitative PCR assay (R2 =0.2, p=0.002). C) Expected and observed methylation in mouse-human hybrid cell lines with a single human X chromosome and placental DNA. (one-tailed tests for significance of the correlation based on the t-distribution).  206  A 21 _1 161^91^18 I^81  82  ('^  21^161^91^1181^48^33  83^1 21 161 19i^72^ 1181^48^33  84 B  6262 6263 6383 63 84 6484 6264 undigested  Figure 5.5 APOE genotyping assay (Wenham et al.1991). A) three APOE alleles are shown with HhaI cut sites indicated by arrows, and expected fragment sizes (in base pairs) indicated between arrows B) APOE genotypes and expected bands on gel electrophoresis following digestion.  207  ▪  1.6 1.4  o^  • al  0  1.2-  ••••••  •  •  LL)  0.8 .  ▪•  •$  1.0 O  •  -  •  •  0.6  •  •  0  •  o ••  ••  o • • 0^ 0_ • • 0• • ♦  ••  •  •  o  •  $ 0  0•  •  0  0 0  0^• •  0  0  Z  0  0  •o  •  •  •  ••  • 0 • •♦ 0 8 6  •  •  •$ •  •  •  • o  ••  • •  • •  •  0  0.4 15^20^25^30^35^40  ^  45^50^55^60  Age at blood draw (years) Control Group 1 N=112 Control group 2 N=47^0 POF patients N=35 Linear (Control Group 1 N=112) - - - Linear (Control group 2 N=47)^— — Linear (POF patients N=35) 5.6 Average relative telomere length by age. Correlations between telomere length and age are R 2 =0.17, p=0.007 for POF patients, R2 =0.07, p=0.002 for Control group 1, and R2 = 0.002, p=0.39 for control group 2 (one-tailed tests for significance of the correlation •  based on the t-distribution).  1.6 -  0  •  1.4 -  0 0  0  • • •  .0 i 1 8  o  0.6 -  0  0 0  0 8  0  I I • •  0  0.4 0  POF patients N=53 • Control group 1 N=106 0 Control group 2 N=46  Figure 5.7 Telomere length in POF patients, Control group 1 and Control group 2. Raw data, unadjusted for age for samples between the ages of 17 and 55 only. Average telomere lengths for POF patients, Control group 1 and Control group 2 are 0.95±0.18, 0.90±0.16, and 0.89±0.14 respectively (p=0.08 vs. Control group 1; p=0.03 vs. combined controls, two-tailed t-tests)  209  1.6 -  o•  1.4 -  o60 • •^o • •  02 0•  •  0  0 0^•  8 c. & • —6— —  • 0.6 -  0  a  •  0  • o  • •^08 • ••0 •  •• •  0  •  0  • 0  0.4 15^25^35^45^55^65^75 Repeat length of longer FMR1 allele •  Control Group 1 N=83 Control group 2 N=47^0 POF patients N=53 Linear (Control Group 1 N=83)^- - - • Linear (Control group 2 N=47)^— — Linear (POF patients N=53)  1..)^Figure 5.8 Telomere length and CGG repeat length of the longer FMR1 allele.  bl) 90% -  ^  80%  •  •  •  •  •  0  •  ct 70%  •  •  0  •  0  • •  •  50% 0^1 0  •  •  •  8/  •  ♦^•^•  • •  0  • •  •  •  •  •  • •  •  • •  • •  •  • • •  0 o ♦:•  0  ••  •  O 9.04 -  • .0/fr^* •  •  •  *0.eV•  •^ •0. OgO`' 0 •  20  •  •  0 • /  •••  0^4 • $ •.*•  60% -  00  • 0 ••  •$•  -  • o^  • •  •  •  0  • •• •  •  0  •  •  100% -  30^40^50  •  • •  60^70^80  Age at blood draw (years) •  Control Group 1 N=124^Control group 2 N=45 Linear (Control Group 1 N=124) ^- - - Linear (Control group 2 N=45)  0 POF patients N-33 — — Linear (POF patients N=33)  Figure 5.9 X chromosome inactivation skewing and age. Correlations between XCI skewing and age are R2 =0.13, p=0.02 for POF patients, R2 =0.1, p=0.09 for Control group 1, and R 2 =0.006, p=0.3 for control group 2 (one-tailed tests for significance of the  correlation based on the t-distribution).  0  80% -  •  •  70% -  •  60% -  •  •  30% 20% -  •  •  10% 0  10  20^30^40^50  60^70^80  Age at blood draw (years) •  Control Group 1 N=45^Control group 2 N=25 Linear (Control Group 1 N=45)^- - - . Linear (Control group 2 N=25)  POF patients N=26 — — Linear (POF patients N=26) Figure 5.10 AR methylation and age. Correlations between AR methylation and age are R2 =0.008, p=0.7 for POF patients, R2 =0.09, p=0.04 for Control group 1, and R 2 =0.001, p=0.9 for control group 2 (two-tailed tests for significance of the correlation based on the tdistribution). 0  80% -  0  70% -  0  60% -  8 0  8 80 80 80 0 0  0 0 8 0 0  30% -  S  • •• • •• $ • • • • • • •  20% -  O 0 a O 0 0 0 0 e El B 0 0 0  B 0 0 a  10% 0  POF patients N=27 • Control group 1 N=19  ° Control group 2 N=24  Figure 5.11 AR methylation in POF patients and control women. Raw data,  unadjusted for age for samples between the ages of 17 and 55 only. Average AR methylation for POF patients, Control group 1 and Control group 2 are 46.3±10.2%, 42.1±11.6%, and 40.6±12.0% respectively (p=0.3 vs. Control group 1; p=0.09 vs. combined controls, two-tailed t-test)  213  1.4  •  1.3 -0  1.2 -  •  o^  •  • o^o^ 0  •^ • ^o•^ •^ 0 •^• • • •,^♦^•^o^ • • •^ • •^ • o o •^ o^ -•• • •o 0 • • • 0 o  o^  ^o  o  •  •^ mmmmmm  .....^  0 -"L  71') .  c=4 0.7 0.6 -  •  0  0  ••^• •  • • ^•  •  •  4  • •  •0  •  ^Ir  -  0.) .t 0.8  •  ^  ♦ o^  -  •  •••••• ....111"  *0  • 0  o•  +1 ■... .111•••■ .wm  ••• • • • co • • •  0 $  O  •  •  •  0  0.5 -  1 ^1 0.4 ^ ^ 100% 60%^70%^80%^90% 50% XCI skewing •  0 POF patients N=46 Control group 2 N=38 Control Group 1 N=98 Linear (Control Group 1 N=98)^- - - Linear (Control group 2 N=38)^— — Linear (POF patients N=46)  N^Figure 5.12 Correlation between telomere length and XCI skewing.  •  1.4 1.3 0  1.2 -  0  •  • •  0  ^0  • - -^o ♦ • 0  • 0  0.9  •  .....  0.8 -  •  TL)'  • 0.7 -  -s'  •■^.  0 0  ..... .....^ ""'"•,„. 41.  0^•  0  •  •  ■•••••■• ••••••Il  •  ♦  •  •  ••••••■  0  0.6 0.5 -  0  0.4 10%  20%  30%  40%^50%  60%  70%^80%  AR methylation  •  Control Group 1 N=22^Control group 2 N=25 0 POF patients N=25 Linear (Control Group 1 N=22) - - - Linear (Control group 2 N=25)^— — Linear (POF patients N=25)  Figure 5.13 Correlation between telomere length and AR methylation.  1 00% 0 0  • 90% -  •  • •  to^80% 5  •  ••  •  • •  0  •  •  0  •  •  • .`•""  ,^•^  ■■••••  o  (  50%  •  •-  • o  Z3 ><^70%-  • •  • 0  •  0  • •  0^ 0^0 0  •  •  0  • •  0  0 0  60% -  •  0  •  •  o  •  10%^20%^30%^40%^50%^60%^70%^ AR methylation •  Control Group 1 N=44^Control group 2 N=23^0^POF patients N=23 Linear (Control Group 1 N=44) ^- - - Linear (Control group 2 N=23)^— — Linear (POF patients N=23)  N Figure 5.14 Correlation between XCI skewing and AR methylation.  80%  A 100% 90% 80% 70% 60% 50% ,.6 40% 30% 20% 10% 0%  POF patients N alleles=110 ^ ■ Control group 1 N alleles=194 Control group 2 N alleles=86  -  )  E2  ^  E3^E4  APOE allele  B 90% 80% 70% 60% 6 50% 40% fz1 30% 20% 10% 0%  O POF patients N=55 ■ Control group 1 N=97 El Control group 2 N=43  -  -  ,  ,  E2E2^E2E3^E3E3^E3E4  E4E4^E2E4  APOE genotype  Figure 5.15 APOE A) allele and B) genotype frequencies in POF patients and controls.  217  5.5 References Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM and Belmont JW. 1992. Methylation of HpaII and HhaI sites near the polymorphic CAG repeat in the human androgenreceptor gene correlates with X chromosome inactivation. Am. J. Hum. Genet. 51:1229-1239. Andrew T, Aviv A, Falchi M, Surdulescu GL, Gardner JP, Lu X, Kimura M, Kato BS, Valdes AM and Spector TD. 2006. Mapping genetic loci that determine leukocyte telomere length in a large sample of unselected female sibling pairs. Am. J. Hum. Genet. 78:480-486. Aviv A, Shay J, Christensen K and Wright W. 2005. The longevity gender gap: are telomeres the explanation? Sci. Aging Knowledge Environ. 2005:pe16. Avramopoulos D, Mikkelsen M, Vassilopoulos D, Grigoriadou M and Petersen MB. 1996. Apolipoprotein E allele distribution in parents of Down's syndrome children. Lancet. 347:862-865. Aydos SE, Elhan AH and Tukun A. 2005. Is telomere length one of the determinants of reproductive life span? Arch. Gynecol. Obstet. 272:113-116. Barbot W, Dupressoir A, Lazar V and Heidmann T. 2002. Epigenetic regulation of an IAP retrotransposon in the aging mouse: progressive demethylation and de-silencing of the element by its repetitive induction. Nucleic Acids Res. 30:2365-2373. Benetos A, Okuda K, Lajemi M, Kimura M, Thomas F, Skurnick J, Labat C, Bean K and Aviv A. 2001. Telomere length as an indicator of biological aging: The gender effect and relation with pulse pressure and pulse wave velocity. Hypertension. 37:381-385. Bretherick KL, Hanna CW, Currie LM, Fluker MR, Hammond GL and Robinson WP. 2008. Estrogen receptor a gene polymorphisms are associated with idiopathic premature ovarian failure. Fertil. Steril. 89(2):318-324. Bretherick KL, Metzger DL, Chanoine JP, Panagiotopoulos C, Watson SK, Lam WL, Fluker MR, Brown CJ and Robinson WP. 2007. Skewed X-chromosome inactivation is associated with primary but not secondary ovarian failure. Am. J. Med. Genet. A. 143:945-951. Brown CJ and Robinson WP. 2000. The causes and consequences of random and nonrandom X chromosome inactivation in humans. Clin. Genet. 58:353-363. Busque L, Mio R, Mattioli J, Brais E, Blais N, Lalonde Y, Maragh M and Gilliland DG. 1996. Nonrandom X-inactivation patterns in normal females: lyonization ratios vary with age. Blood. 88:59-65. Cargill SL, Carey JR, Muller HG and Anderson G. 2003. Age of ovary determines remaining life expectancy in old ovariectomized mice. Aging Cell. 2:185-190.  218  Cawthon RM. 2002. Telomere measurement by quantitative PCR. Nucleic Acids Res. 30:e47. Cherkas LF, Aviv A, Valdes AM, Hunkin JL, Gardner JP, Surdulescu GL, Kimura M and Spector TD. 2006. The effects of social status on biological aging as measured by white-blood-cell telomere length. Aging Cell. 5:361-365. Christensen K, Johnson TE and Vaupel JW. 2006. The quest for genetic determinants of human longevity: challenges and insights. Nat. Rev. Genet. 7:436-448. Christensen K, Kristiansen M, Hagen-Larsen H, Skytthe A, Bathum L, Jeune B, Andersen-Ranberg K, Vaupel JW and Orstavik KH. 2000. X-linked genetic factors regulate hematopoietic stem-cell kinetics in females. Blood. 95:2449-2451. Chu MW, Siegmund KD, Eckstam CL, Kim JY, Yang AS, Kanel GC, Tavare S and Shibata D. 2007. Lack of increases in methylation at three CpG-rich genomic loci in non-mitotic adult tissues during aging. BMC Med. Genet. 8:50. Cooney CA. 2001. Dietary selenium and arsenic affect DNA methylation. J. Nutr. 131:1871-1872. Cooper GS and Sandler DP. 1998. Age at natural menopause and mortality. Ann. Epidemiol. 8:229-235. Doblhammer G. 2000. Reproductive history and mortality later in life: a comparative study of England and Wales and Austria. Popul. Stud. (Camb). 54:169-176. Dorland M, van Kooij RJ and te Velde ER. 1998a. General ageing and ovarian ageing. Maturitas. 30:113-118. Dorland M, van Montfrans JM, van Kooij RJ, Lambalk CB and te Velde ER. 1998b. Normal telomere lengths in young mothers of children with Down's syndrome. Lancet. 352:961-962. El Kassar N, Hetet G, Briere J and Grandchamp B. 1998. X-chromosome inactivation in healthy females: incidence of excessive lyonization with age and comparison of assays involving DNA methylation and transcript polymorphisms. Clin. Chem. 44:61-67. Epel ES, Blackburn EH, Lin J, Dhabhar FS, Adler NE, Morrow JD and Cawthon RM. 2004. Accelerated telomere shortening in response to life stress. Proc. Natl. Acad. Sci. U. S. A. 101:17312-17315. Gale RE, Fielding AK, Harrison CN and Linch DC. 1997. Acquired skewing of Xchromosome inactivation patterns in myeloid cells of the elderly suggests stochastic clonal loss with age. Br. J. Haematol. 98:512-519.  219  Gilley D, Herbert BS, Huda N, Tanaka H and Reed T. 2008. Factors impacting human telomere homeostasis and age-related disease. Mech. Ageing Dev. 129:27-34. Golbus J, Palella TD and Richardson BC. 1990. Quantitative changes in T cell DNA methylation occur during differentiation and ageing. Eur. J. Immunol. 20:1869-1872. Harley CB, Vaziri H, Counter CM, Allsopp RC. 1992. The telomere hypothesis of cellular aging. Exp Geront. 27:375-382. Hassold T, Abruzzo M, Adkins K, Griffin D, Merrill M, Millie E, Saker D, Shen J and Zaragoza M. 1996. Human aneuploidy: incidence, origin, and etiology. Environ. Mol. Mutagen. 28:167-175. Hatakeyama C, Anderson CL, Beever CL, Penaherrera MS, Brown CJ and Robinson WP. 2004. The dynamics of X-inactivation skewing as women age. Clin. Genet. 66:327332. Helle S, Lummaa V and Jokela J. 2005. Are reproductive and somatic senescence coupled in humans? Late, but not early, reproduction correlated with longevity in historical Sami women. Proc. Biol. Sci. 272:29-37. Hsin H and Kenyon C. 1999. Signals from the reproductive system regulate the lifespan of C. elegans. Nature. 399:362-366. Huda N, Tanaka H, Herbert BS, Reed T and Gilley D. 2007. Shared environmental factors associated with telomere length maintenance in elderly male twins. Aging Cell. 6:709-713. Hutchinson EW and Rose MR. 1991. Quantitative genetics of postponed aging in Drosophila melanogaster. I. Analysis of outbred populations. Genetics. 127:719-727. Issa JP, Ottaviano YL, Celano P, Hamilton SR, Davidson NE and Baylin SB. 1994. Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat. Genet. 7:536-540. Jacobsen BK, Knutsen SF and Fraser GE. 1999. Age at natural menopause and total mortality and mortality from ischemic heart disease: the Adventist Health Study. J. Clin. Epidemiol. 52:303-307. Jarrard DF, Kinoshita H, Shi Y, Sandefur C, Hoff D, Meisner LF, Chang C, Herman JG, Isaacs WB and Nassif N. 1998. Methylation of the androgen receptor promoter CpG island is associated with loss of androgen receptor expression in prostate cancer cells. Cancer Res. 58:5310-5314. Jeanclos E, Schork NJ, Kyvik KO, Kimura M, Skurnick JH and Aviv A. 2000. Telomere length inversely correlates with pulse pressure and is highly familial. Hypertension. 36:195-200.  220  Jenkins EC, Tassone F, Ye L, Gu H, Brown WT, Hagerman RJ and Hagerman PJ. 2007. Shorter Telomeres in Older Male Individuals with the Fragile X Premutation, FXTAS, and FXTAS with Dementia. Presented at the annual meeting of The American Society of Human Genetics, October 26, 2007, San Diego, California. Available from http://www.ashg.org/genetics/ashg07s/index.shtml Kim JY, Tavare S and Shibata D. 2005. Counting human somatic cell replications: methylation mirrors endometrial stem cell divisions. Proc. Natl. Acad. Sci. U. S. A. 102:17739-17744. Kinugawa C, Murakami T, Okamura K and Yajima A. 2000. Telomerase activity in normal ovaries and premature ovarian failure. Tohoku J. Exp. Med. 190:231-238. Kirkwood TB and Rose MR. 1991. Evolution of senescence: late survival sacrificed for reproduction. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 332:15-24. Kotrschal A, Ilmonen P and Penn DJ. 2007. Stress impacts telomere dynamics. Biol. Lett. 3:128-130. Kristiansen M, Knudsen GP, Bathum L, Naumova AK, Sorensen TI, Brix TH, Svendsen AJ, Christensen K, Kyvik KO and Orstavik KH. 2005. Twin study of genetic and aging effects on X chromosome inactivation. Eur. J. Hum. Genet. 13:599-606. Kumar RC and Thakur MK. 2004. Androgen receptor mRNA is inversely regulated by testosterone and estradiol in adult mouse brain. Neurobiol. Aging. 25:925-933. Lee DC, Im JA, Kim JH, Lee HR and Shim JY. 2005. Effect of long-term hormone therapy on telomere length in postmenopausal women. Yonsei Med. J. 46:471-479. Manor 0, Eisenbach Z, Israeli A and Friedlander Y. 2000. Mortality differentials among women: the Israel Longitudinal Mortality Study. Soc. Sci. Med. 51:1175-1188. McAbee MD and Doncarlos LL. 1999. Estrogen, but not androgens, regulates androgen receptor messenger ribonucleic acid expression in the developing male rat forebrain. Endocrinology. 140:3674-3681. Muller HG, Chiou JM, Carey JR and Wang JL. 2002. Fertility and life span: late children enhance female longevity. J. Gerontol. A Biol. Sci. Med. Sci. 57:B202-6. Nagai J, Lin CY and Sabour MP. 1995. Lines of mice selected for reproductive longevity. Growth Dev. Aging. 59:79-91. Nagy B, Ban Z, Toth-Pal E, Papp C, Fintor L and Papp Z. 2000. Apolipoprotein E allele distribution in trisomy 13, 18, and 21 conceptuses in a Hungarian population. Am. J. Clin. Pathol. 113:535-538. Pal L and Santoro N. 2002. Premature ovarian failure (POF): discordance between somatic and reproductive aging. Ageing Res. Rev. 1:413-423.  221  Perls TT, Alpert L and Fretts RC. 1997. Middle-aged mothers live longer. Nature 389:133. Perls T, Levenson R, Regan M and Puca A. 2002. What does it take to live to 100? Mech. Ageing Dev. 123:231-242. Perls TT and Fretts RC. 2001. The evolution of menopause and human life span. Ann. Hum. Biol. 28:237-245. Richardson B. 2003. Impact of aging on DNA methylation. Ageing Res. Rev. 2:245-261. Sandovici I, Naumova AK, Leppert M, Linares Y and Sapienza C. 2004. A longitudinal study of X-inactivation ratio in human females. Hum. Genet. 115:387-392. Sato K, Uehara S, Hashiyada M, Nabeshima H, Sugawara J, Terada Y, Yaegashi N and Okamura K. 2004a. Genetic significance of skewed X-chromosome inactivation in premature ovarian failure. Am. J. Med. Genet. A. 130:240-244. Sato R, Maesawa C, Fujisawa K, Wada K, Oikawa K, Takikawa Y, Suzuki K, Oikawa H, Ishikawa K and Masuda T. 2004b. Prevention of critical telomere shortening by oestradiol in human normal hepatic cultured cells and carbon tetrachloride induced rat liver fibrosis. Gut. 53:1001-1009. Sharp A, Robinson D and Jacobs P. 2000. Age- and tissue-specific variation of X chromosome inactivation ratios in normal women. Hum. Genet. 107:343-349. Slagboom PE, Droog S and Boomsma DI. 1994. Genetic determination of telomere size in humans: a twin study of three age groups. Am. J. Hum. Genet. 55:876-882. Smith KR, Mineau GP and Bean LL. 2002. Fertility and post-reproductive longevity. Soc. Biol. 49:185-205. Snowdon DA, Kane RL, Beeson WL, Burke GL, Sprafka JM, Potter J, Iso H, Jacobs DR,Jr and Phillips RL. 1989. Is early natural menopause a biologic marker of health and aging? Am. J. Public Health. 79:709-714. Tonon L, Bergamaschi G, Dellavecchia C, Rosti V, Lucotti C, Malabarba L, Novella A, Vercesi E, Frassoni F and Cazzola M. 1998. Unbalanced X-chromosome inactivation in haemopoietic cells from normal women. Br. J. Haematol. 102:996-1003. Unryn BM, Cook LS and Riabowol KT. 2005. Paternal age is positively linked to telomere length of children. Aging Cell. 4:97-101. Valdes AM, Andrew T, Gardner JP, Kimura M, Oelsner E, Cherkas LF, Aviv A and Spector TD. 2005. Obesity, cigarette smoking, and telomere length in women. Lancet. 366:662-664.  222  von Zglinicki T. 2000. Role of oxidative stress in telomere length regulation and replicative senescence. Ann. N. Y. Acad. Sci. 908:99-110. Vottero A, Capelletti M, Giuliodori S, Viani I, Ziveri M, Neri TM, Bernasconi S and Ghizzoni L. 2006. Decreased androgen receptor gene methylation in premature pubarche: a novel pathogenetic mechanism? J. Clin. Endocrinol. Metab. 91:968-972. Walter MF, Biessmann MR, Benitez C, Torok T, Mason JM and Biessmann H. 2007. Effects of telomere length in Drosophila melanogaster on life span, fecundity, and fertility. Chromosoma. 116:41-51. Wenham PR, Price WH and Blandell G. 1991. Apolipoprotein E genotyping by one-stage PCR. Lancet. 337:1158-1159. Westendorp RG and Kirkwood TB. 1998. Human longevity at the cost of reproductive success. Nature. 396:743-746. Xiao Y, Word B, Starlard-Davenport A, Haefele A, Lyn-Cook BD and Hammons G. 2007. Age and gender affect DNMT3a and DNMT3b expression in human liver. Cell Biol. Toxicol. Zhang Z, Deng C, Lu Q and Richardson B. 2002. Age-dependent DNA methylation changes in the ITGAL (CD1 la) promoter. Mech. Ageing Dev. 123:1257-1268.  223  Chapter 6: Discussion This thesis has examined several candidate genetic factors for association with increased risk of premature ovarian failure (POF). This included examination of the role of the FMR1 gene, analysis of a number of hormone related gene variants, evaluation of X chromosome inactivation skewing and assessment of a number of age-related chromosomal factors. Here I will briefly review the findings of this research, discuss the strengths and limitations of this study, summarize current knowledge and future investigations required in this field, and highlight the significance of this research.  6.1 Genetic factors assessed in this thesis The results in this thesis confirm the association of POF with FMRI premutation alleles. Moreover, a higher prevalence of alleles at the high end of the normal range was found in POF patients suggesting that risk for POF may extend outside the classic premutation range. These results have subsequently been supported by at least one additional study (Bodega et al. 2006). We found no evidence of altered FMR1 methylation or gene expression among the POF patients, suggesting that the FMR1 gene is not involved in POF pathogenesis in women without FMR1 repeat expansions. Analysis of common polymorphisms at several hormone related genes revealed a novel association between the estrogen receptor a gene, ESRJ, and POF. There was no evidence to support an association with the androgen receptor, estrogen receptor (3, FSH receptor, or sex hormone binding globulin genes in this population. A specific ESRI haplotype was associated with POF in a simple dominant manner in which carriers have a nearly 10 times increased risk of developing ovarian failure. Although the functional  224  effect of this haplotype could not be conclusively determined, it may confer a more active promoter that could influence POF risk by increasing the rate of follicular atresia. Skewed X chromosome inactivation (XCI) assessment in POF patients highlighted the distinctly different etiologies between patients presenting with primary amenorrhea and those presenting with secondary amenorrhea. There was no increase in skewed XCI in POF patients presenting with secondary amenorrhea, however there was a highly significant increase in a small group of patients presenting with primary amenorrhea. Skewed XCI in these patients was not due to X chromosome copy number alterations detectable by microarray, and may be the result of a trisomic rescue event in early development. Age-related chromosome factors including skewed XCI, telomere length, and DNA methylation were assessed to determine if POF patients, who experience premature reproductive aging, also demonstrate an increased rate of cellular aging. XCI skewing and AR methylation were found to increase with age; however they were not increased in POF patients. Although telomeres decreased in length with age, surprisingly they were increased rather than decreased in the POF patient group. This finding may not be a cause of POF, but rather a co-occurrence or consequence, and may be explained by altered level of estrogen or response to estrogen in these patients. In addition, genotype at the APOE gene was not found to be associated with increased risk of POF. There was no detectable relationship between any of these age-related chromosome factors suggesting that their regulation is likely much more complex than merely rate of cell division.  225  6.2 Strengths and limitations of this study The type of patient and control sample set assessed in this thesis presents a number of advantages. The most significant single gene association with POF is with the FMRI gene, and until recently the bulk research on POF focussed on assessing the risk of POF in carriers of FMR1 repeat expansions. Thus, at its inception, the sample set examined in this study was one of a handful (Kenneson et al. 1997; Marozzi et al. 2000; Murray et al. 1998; Patsalis et al. 1999; Uzielli et al. 1999) capable of assessing the range of FMR1 alleles in POF patients or determining other genetic factors influencing POF risk. In addition, POF patients were compared to not only normal women from the general population, but also to women who had demonstrated ovarian function at a relatively late reproductive age by having healthy pregnancy after the age of 37 years and not experiencing any pregnancy loss. Analysis of women from the general population allows genetic factors in the POF patient group to be compared to those expected in normal individuals, and the relative risk for POF above the population risk can be determined. Comparison to women at the opposite end of the spectrum of reproductive capacity from the POF patients acts as a more stringent control group and enables a more sensitive measure of whether of not the genetic factors studied have any bearing on ovarian function. There are, however, several limitations to the samples examined in this study that are important to consider. Firstly, the number of POF patient and control samples available for study was limited. This restricts the power of this study to detect effects of genes with small influences on phenotype and hampers attempts at identifying interactions between these genes that may be critical in POF pathogenesis. It has been  226  proposed that POF and age at menopause are complex genetic traits to which a number of additive genetic and environmental factors contribute (Snieder et al. 1998; van Asselt et al. 2004). A sample set of the size used in this study may be insufficient to adequately assess interactions between these factors. Secondly, the clinical criteria used for POF diagnosis was less stringent than the conventional standard. POF diagnosis for samples in this thesis was made based on the absence of menses for at least 3 months and two serum FSH results of >40 mIU/mL obtained more than one month apart, prior to age 40. Traditional criteria for POF requires an absence for menses for 6 rather than 3 months (Vegetti et al. 2000), however, since it is critical not to waste time in the management of infertility in aging patients it has been argued that an earlier cut off be used for diagnosis to facilitate timely treatment (Nippita and Baber. 2007). Using a more stringent criterion of 6 months would likely only serve to amplify any association reported in this study. Thirdly, limited clinical data was available for many of the POF patients in this study. All POF patients included in this study were normal on routine diagnostic karyotype, and had no known environmental cause (radiation, chemotherapy) for ovarian failure. However, due to ascertainment methods, details on the fertility workup for these patients were not available. In particular, information regarding the presence or absence of autoimmunity in POF patients was not available. Although the mechanisms are not well understood, autoimmune causes may be responsible for a significant percentage of POF cases (Geva et al. 1997; Hoek et al. 1997), and their presence may at least partially explain a number of cases in this study. Fourthly, for obvious logistical reasons analyses in this study were done on DNA extracted from peripheral blood of patients after POF diagnosis. This is sufficient for assessment of genetic factors such as SNP or  227  microsatellite polymorphisms that are independent of tissue and developmental timing. However, other genetic factors such as telomere length, DNA methylation, and RNA quantity inevitably vary by tissue type, life stage, and possibly menstrual cycle time point. Although assessing these factors in blood may provide clues to their nature at the point where they are critical to POF, it would likely be more powerful to assess them at the tissue and time point they are most relevant. This research was performed as a genetic association study using a candidate gene approach, a study design that has both strengths and weaknesses. As an advantage, this method is more sensitive than family-based linkage studies to detect genes variants with small effects on risk (Daly and Day. 2001; Jones et al. 2005), and may therefore partially compensate for the small sample size in this study. However, a limitation of this approach is the possibility of spurious results due to population stratification or biases in patient and control recruitment. Although we attempted to control for this in the samples assessed in this thesis (Chapter 3 methods) it is nonetheless an ongoing concern regarding this data. A distinct advantage of the candidate gene approach in this study was the ability to focus analysis on genetic factors that are good candidates for involvement in POF pathogenesis based on their role and function. This is one of the few studies to report on common hormone related gene variants or non-genic factors such as telomere length, or DNA methylation in POF patients. A parallel limitation is the inability to assess all possible candidates and therefore attain an estimate of the percentage of POF cases in which a genetic cause can be ascribed.  228  6.3 Current knowledge and future research on genetics and POF In the last ten years there has been an explosion of reports suggesting novel genes that influence risk for POF (reviewed in Chapter 1). These have been established primarily by linkage studies in affected families or candidate gene screening in patient and control populations. Nonetheless, the only gene that is consistently and reproducibly shown to significantly influence risk is FMR1. This may be explained in part by the fact that mutations in other genes have been found to be rare and cause POF in only very few patients. For the majority of cases it is likely that a number of small effect genes contribute to disease development. Particularly for patients presenting with nonsyndromic secondary amenorrhea, POF may be part the extreme end of the normal distribution of age at menopause and may therefore be viewed as a complex trait to which multiple genes contribute. The small scale association studies performed thus far (including this one) may simply not have had the power to consistently detect these small effect genes. Recent abstracts assessing age at menopause in Fragile X syndrome families have suggested that the FMR1 gene operates on a background of increased genetic risk to cause POF in premutation carriers (Hunter et al. 2007; Pearson et al. 2007), providing a possible explanation for the incomplete penetrance of FMR1. These findings suggest that the heritability estimate for age at menopause in these families after correction for FMRI repeat length is >80% (Hunter et al. 2007), supporting the suggestion that many other genes contribute to POF in FMR1 premutation carriers. Future research on the role of FMR1 in POF is necessary to clarify the etiology of this association. Several explanations for the mechanism of this association have been put forward in chapter 2. Some of these will require the use of molecular methods and  229  cell systems to determine how they are influenced by repeat length and how they negatively influence ovarian function. In particular, clarifying the explanation for the reduced penetrance of FMR1 expansions may illuminate further the mechanism involved. This may be initiated by examining populations of premutation carriers with and without POF for features such as FMRI CGG repeat interruptions and expression of alternative transcripts influenced by CGG repeat length. Identification of the mechanistic pathway may lead to other candidate genes that similarly influence POF risk. Furthermore, if, as suggested by (Pearson et al. 2007), FMR1 substantially increases POF risk in carriers of multiple genes with more minor influences on risk, FMR1 premutation carriers with POF may represent a high risk group ideal for the identification of other risk genes that play a role in non-carriers. Large scale studies may be necessary for the identification of further POF genes. An array-based whole genome scan of single nucleotide polymorphisms in a large population of clearly defined POF patients and controls could be useful in detecting genes with minor influences on risk and determining how these interact with one another. Furthermore, defining the environmental factors that increase risk of POF and elucidating gene-environment interactions that influence ovarian function may be of broader clinical significance. Alternatively, considering POF as a quantitative trait and performing a thorough study on a population of women experiencing menopause at a variety of ages may more accurately represent its pathophysiology and be a more powerful tool to isolate genes influencing risk. An important consideration in future genetic studies of POF is the necessity to precisely define the endophenotype under examination. Within the common POF  230  diagnosis exists considerable phenotypic heterogeneity and failing to address this significantly dilutes the power of any association study. The XCI skewing results presented here highlight the distinct etiology between patients presenting with primary and secondary amenorrhea. In addition, the distinction between POF cases resulting from a premature reduction in follicular reserve and those ensuing from dysfunctional ovaries would no doubt clarify genetic studies. Determination of ovarian reserve is most accurately accomplished by ovarian biopsy, and less reliably assessed by ultrasound or serum hormone markers (Vital-Reyes et al. 2006). While the feasibility of obtaining these parameters on large numbers of patients would be limiting, the additional power obtained from separating these phenotypes would be beneficial.  6.4 Significance The results of this thesis offer a significant contribution to the existing body of literature on genetic factors implicated in POF. Firstly the novel finding that FMR1 alleles at the high end of the normal range occur at increased frequency in POF patients extends the risk range into a category that is more common in the population and suggests that FMRI may play a larger role in POF than previously thought. As additional studies further clarify the relationship between FMR1 repeat size and POF risk, this will become increasingly valuable as a tool for counselling patients with a family history of POF or Fragile X syndrome. In addition, this finding provides insight into the mechanism of the POF association with FMR1, by highlighting that POF risk is associated with not only the expansion-prone premutation size alleles. Secondly, the previously unreported association between POF and estrogen receptor a has revealed that a relatively common haplotype in the population has significant influence on risk for  231  POF. The high frequency and incomplete penetrance of this haplotype suggests that it must interact with other genetic or environmental factors to influence risk and the result presented here provide a framework for further research into these interactions. Thirdly, existing literature on the relationship between POF and skewed XCI had arrived at contradictory conclusions. The results in this thesis suggest that patient ascertainment and phenotypic presentation may explain discrepancies in previously reported studies. Lastly, the perplexing finding of increased telomere length in the POF patient population may support suggestions in the literature that telomere length is influenced by circulating estrogen level. Currently there is a paucity of studies on this relationship that may have implications for our understanding of both telomere length regulation and the effects of hormonal environment. Identifying genetic factors associated with increased risk of POF is important for both patient care and elucidation of disease mechanisms. Defining the genetic causes of POF will provide explanation for more individual patient cases and be useful for predictive testing for family members at risk and wanting to make informed reproductive choices. This is increasingly relevant as more and more women are delaying child bearing until later in their reproductive life. Determining the genetic factors involved POF is also critical in the development of therapies and treatment options for restoring fertility in affected individuals. Furthermore, discovering the genetic and environmental factors associated with POF provides a starting point to further understand the mechanisms responsible in for this complex and poorly understood disease.  232  6.5 Conclusion POF is a relatively common disorder that affects approximately 1% of the female population and is thought to have a significant genetic contribution. However, little is known about the genetic etiology in the majority of cases. In this thesis several candidate genetic factors were examined for a possible influence on risk of POF via a population case control study. The findings have clarified some aspects of this field and may ultimately have implications for patient counselling, however perhaps more significantly this research has illuminated new areas of research in this field. This provides background for future research into the pathogenesis of this poorly understood and increasingly concerning reproductive health issue.  233  6.6 References Bodega B, Bione S, Dalpra L, Toniolo D, Ornaghi F, Vegetti W, Ginelli E and Marozzi A. 2006. Influence of intermediate and uninterrupted FMR1 CGG expansions in premature ovarian failure manifestation. Hum. Reprod. 21:952-957. Daly AK and Day CP. 2001. Candidate gene case-control association studies: advantages and potential pitfalls. Br. J. Clin. Pharmacol. 52:489-499. Geva E, Amit A, Lerner-Geva L and Lessing JB. 1997. Autoimmunity and reproduction. Fertil. Steril. 67:599-611. Hoek A, Schoemaker J and Drexhage HA. 1997. Premature ovarian failure and ovarian autoimmunity. Endocr. Rev. 18:107-134. Hunter J, Tinker S, Epstein M and Sherman S 2007. Familial aggregation of premature ovarian failure related to the FMR1 repeat expansion. Presented at the 13th International workshop on Fragile X and X-linked Mental Retardation, October 3-6, 2007, Venice, Italy P34: Jones R, Pembrey M, Golding J and Herrick D. 2005. The search for genenotype/phenotype associations and the phenome scan. Paediatr. Perinat. Epidemiol. 19:264-275. Kenneson A, Cramer DW and Warren ST. 1997. Fragile X premutations are not a major cause of early menopause. Am. J. Hum. Genet. 61:1362-1369. Marozzi A, Vegetti W, Manfredini E, Tibiletti MG, Testa G, Crosignani PG, Ginelli E, Meneveri R and Dalpra L. 2000. Association between idiopathic premature ovarian failure and fragile X premutation. Hum. Reprod. 15:197-202. Murray A, Webb J, Grimley S, Conway G and Jacobs P. 1998. Studies of FRAXA and FRAXE in women with premature ovarian failure. J. Med. Genet. 35:637-640. Nippita TA and Baber RJ. 2007. Premature ovarian failure: a review. Climacteric. 10:1122. Patsalis PC, Hettinger JA and Sismani C. 1999. FMR1 repeat analysis in patients with ovarian dysfunction or failure. Am. J. Med. Genet. 83:329-330. Pearson PL, Costa S and Vianna-Morgante A 2007. The contribution of genetic background and FMR1 premutation to menopausal age: FMR1 is only half the story. Presented at the 13th International Workshop on Fragile X and X-linked Mental Retardation, October 3-6, 2007, Venice, Italy P57: Snieder H, MacGregor AJ and Spector TD. 1998. Genes control the cessation of a woman's reproductive life: a twin study of hysterectomy and age at menopause. J. Clin. Endocrinol. Metab. 83:1875-1880.  234  Uzielli ML, Guarducci S, Lapi E, Cecconi A, Ricci U, Ricotti G, Biondi C, Scarselli B, Vieri F, Scarnato P, et al. 1999. Premature ovarian failure (POF) and fragile X premutation females: from POF to to fragile X carrier identification, from fragile X carrier diagnosis to POF association data. Am. J. Med. Genet. 84:300-303. van Asselt KM, Kok HS, Putter H, Wijmenga C, Peeters PH, van der Schouw YT, Grobbee DE, to Velde ER, Mosselman S and Pearson PL. 2004. Linkage analysis of extremely discordant and concordant sibling pairs identifies quantitative trait loci influencing variation in human menopausal age. Am. J. Hum. Genet. 74:444-453. Vegetti W, Marozzi A, Manfredini E, Testa G, Alagna F, Nicolosi A, Caliari I, Taborelli M, Tibiletti MG, Dalpra L, et al. 2000. Premature ovarian failure. Mol. Cell. Endocrinol. 161:53-57. Vital-Reyes V, Chhieng D, Rodriguez-Burford C, Tellez-Velasco S, Grizzle W, Chavarria-Olarte ME and Reyes-Fuentes A. 2006. Ovarian biopsy in infertile patients with ovarian dysfunction. Int. J. Gynecol. Pathol. 25:90-94.  235  Appendix 1: UBC Research Ethics Board Certificates of Approval CO1-0460  The University of British Columbia and Administration ..,=" .^of^Services Office^Research %SS^."^  Clinical Research Ethics Board  Certificate of Approval T^T  Rca hinst n, 11".P  Medical Gene  ,  J^ICI^u'-,,  Lirel I S & '  Women's Health .  aRs Bruyerc--, tltlene, Medical Genetics; Glair, Jane, ^al Crenrti cs A ,I1'  i ^!Oil ,  li,t`C  ,^1 [fl I< esearch  TliLi 7^'  l,^.^  ,,  ,_.,,,  flCHCli'diC' ''' 0 l '  M AR 1 3 2002  etitroill^l'ii&Oiliy  I  ' - 1^bru L y 200_^); .^II horsed  The protocol and consent form for the above-named project have been reviewed by the Committee and the experimental procedures were found to be acceptable on ethical grounds for research involving human subjects  Approval of the (.., finical^ese rch Ethics Board by one Dr. P. I oewen, Chair Dr. A. Hannatn, Associate Chair Angus Liv ingstone, Director Pro Tem, Research Services  This Certificate of Approval is valid for the above term provided there is no change in the experimental procedures  236  The University of British Columbia Office^Research of Services and Administration -- , 7^Clinical Research Ethics Board  worm rem^  Certificate of Expedited Approval: Renewal PRINCIPAL INVESTIGATOR  DEPARTMENT  Robinson, W.P.  Medical Genetics  NUMBER  CO1-0460  INSTITUTIONS) WHERE RESEARCH WILL BE CARRIED OUT  Children's & Women's Health Ctr CO•INVESTIGATORS:  Brown, Carolyn, Medical Genetics; Gair, Jane, Medical Genetics SPONSORING AGENCIES  Canadian Institutes of Health Research TITLE :  Genes, Chromosomes and Human Reproduction APPROVAL RENEWAL DATE  TERM (YEARS)  JUL 1 7 2003  1  AMENDMENT:  Consent forms dd 1 July 2003  AMENDMENT APPROVED:  JUL 1 7 2003  CERTIFICATION:  In respect of clinical trials:  1.The membership of this Research Ethics Board complies with the membership requirements for Research Ethics Boards defined in Division 5 of the Food and Drug Regulations. 2. The Research Ethics Board carries out its functions in a manner consistent with Good Clinical Practices. 3. This Research Ethics Board has reviewed and approved the clinical trial protocol and informed consent form for the trial which is to be conducted by the qualified investigator named above at the specified clinical trial site. This approval and the views of the this Research Ethics Board have been documented in writing.  The Chair of the UBC Clinical Research Ethics Board has reviewed the documentation for the above named project. The research study, as presented In the documentation, was found to be acceptable on ethical grounds for research involving human subjects and was approved for renewal by the UBC Clinical Research Ethics Board. The CREB approval for renewal of this study expires one year from the date of renewal.  Approval of the Clin^I R earth Ethics Board by one of Dr.^. Loewen, Chair Dr. Alain Gagnon, Associate Chair  237  UK  The University of British Columbia fRReesseeaar rchEStheics rvice  j 1;  Clinical^Board rd — Room 210, 828 West 10 th Avenue, Vancouver, BC V5Z 1 L8  Certificate of Expedited Approval: Renewal Clinical Research Ethics Board Official Notification  PRINCIPAL INVESTIGATOR  DEPARTMENT  Robinson, W.P.  Medical Gendtics  NUMBER  CO1-0460  INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT  Children's & Women's Health Centre CO-INVESTIGATORS:  Brown, Carolyn, Medical Genetics; Gair, Jane, Medical Genetics; Yew, Sandie, SPONSORING AGENCIES  Canadian Institutes of Health Research TITLE :  Genes, Chromosomes and Human Reproduction APPROVAL RENEWAL DATE  30 July 2004  TERM (YEARS)  1  AMENDMENT:  AMENDMENT APPROVED:  CERTIFICATION:  In respect of clinical trials:  1. The membership of this Research Ethics Board complies with the membership requirements for Research Ethics Boards defined in Division 5 of the Food and Drug Regulations. 2. The Research Ethics Board carries out its functions in a manner consistent with Good Clinical Practices. 3. This Research Ethics Board has reviewed and approved the clinical trial protocol and informed consent form for the trial which is to be conducted by the qualified investigator named above at the specified clinical trial site. This approval and the views of the this Research Ethics Board have been documented in writing.  The Chair of the UBC Clinical Research Ethics Board has reviewed the documentation for the above named project. The research study, as presented in the documentation, was found to be acceptable on ethical grounds for research involving human subjects and was approved for renewal by the UBC Clinical Research Ethics Board. The CREB approval for renewal of this study expires one year from the date of renewal.  Approval of the Clinical Research Ethics Board by one of Dr. P. Loewen, Chair Dr. Alain Gagnon, Associate Chair Dr. James McCormack, Associate Chair  238  Appendix 2: UBC Research Ethics Board Certificates of Approval HO1-70460 The University of British Columbia Office of Research Services Clinical Research Ethics Board — Room 110, 828 West 10th Avenue, Vancouver, BC V5Z 11.8  ETHICS CERTIFICATE OF EXPEDITED APPROVAL: RENEWAL PRINCIPAL INVESTIGATOR:  DEPARTMENT:  Wendy P. Robinson  UBC CREB NUMBER: H01 70460 -  INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT: Institution^ 1^  Site  Children's and Women's Health Centre of BC (incl.^Children's and Women's Health Centre of BC (incl. Sunny Hill)^ Sunny Hill) Other locations where the research will be conducted: N/A  CO-INVESTIGATOR(S): Jane L. Gair Carolyn J. Brown Monica Pearson SPONSORING AGENCIES: Canadian Institutes of Health Research - "Genes, Chromosomes and Human Reproduction" Unfunded Research - "Genetic Studies of Recurrent Trisomy" PROJECT TITLE: Genes, Chromosomes and Human Reproduction EXPIRY DATE OF THIS APPROVAL: November 7, 2007 APPROVAL DATE: November 1, 2006  CERTIFICATION: In respect of clinical trials: 1. The membership of this Research Ethics Board complies with the membership requirements for Research Ethics Boards defined in Division 5 of the Food and Drug Regulations. 2. The Research Ethics Board carries out its functions in a manner consistent with Good Clinical Practices. 3. This Research Ethics Board has reviewed and approved the clinical trial protocol and informed consent form for the trial which is to be conducted by the qualified investigator named above at the specified clinical trial site. This approval and the views of this Research Ethics Board have been documented in writing. The Chair of the UBC Clinical Research Ethics Board has reviewed the documentation for the above named ^_ project. The research study, as presented in the documentation, was found to be acceptable on ethical grounds for research involving human subjects and was approved for renewal by the UBC Clinical Research Ethics Board.  Approval of the Clinical Research Ethics Board by one of  11111 Dr. Bonita Sawatzky, Associate Chair 239  UK  The University of British Columbia Office of Research Services Clinical Research Ethics Board - Room 210, 828 West 10th Avenue, Vancouve BC V52 1L8  \-.W  ETHICS CERTIFICATE OF EXPEDITED APPROVAL: RENEWAL WITH AMENDMENTS TO THE STUDY PRINCIPAL INVESTIGATOR: Wendy P. Robinson  DEPARTMENT:  UBC CREB NUMBER: H01 - 70460  INSTITUTION(S) WHERE RESEARCH WILL BE CARRIED OUT: Institution^  I^  site  C hildren's and Women's Health Centre of BC (incl. ^Children's and Women's Health Centre of BC (incl.  Sunny Hill)^  Sunny Hill)  Other locations where the research will be conducted!  N/A  CO INVESTIGATOR(S): -  Jane L. Gair arolyn J. Brown  SPONSORING AGENCIES:  Canadian Institutes of Health Research (CIHR) - "Genes, Chromosomes and Human Reproduction" Unfunded Research - "Genetic Studies of Recurrent Trisom "  PROJECT TITLE:  Genes, Chromosomes and Human Reproduction  The current UBC CREB approval for this study expires: Decembe r 17, 2008 AMENDMENT(S): Document Name^  1 Version I ^Data  C3t_nSitnti9rM5_; Genetic Studis of Recurrent Trisomy Control ^December 3, 3^ Consent^ 2007 Genetic Studies of Recurrent Trisomy Subject^December 3, 3^2007 Consent^ Advertisements; November 8, Recruitment Flyer_ Reproductive Aging^4^ 2007 November 8, Pamphlet_ Reproductive Aging^ 2^ 2007  AMENDMENT APPROVAL DATE: ,December 17, 2007  CERTIFICATION: In respect of clinical trials:  1. The membership of this Research Ethics Board complies with the membership requirements for Research Ethics Boards defined in Division 5 of the Food and Drug Regulations. 2. The Research Ethics Board carries out its functions in a manner consistent with Good Clinical Practices. 3. This Research Ethics Board has reviewed and approved the clinical trial protocol and informed consent form for the trial which is to be conducted by the qualified investigator named above at the specified clinical trial site. This approval and the views of this Research Ethics Board have been documented in writing.  The Chair of the UBC Clinical Research Ethics Board has reviewed the documentation for the above named project. The research study, as presented in the documentation, was found to be acceptable on ethical grounds for research involving human subjects and was approved for renewal by the UBC Clinical Research Ethics Board.  Approval of the Clinical Research Ethics Board by:  11111P11111.  -  Dr. James McCormack, Associate Chair  240  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0066666/manifest

Comment

Related Items