Premature ovarian failure : from phenotype to genotype

Abstract

Postponement of childbearing has led to increased rates of age-related female subfertility. Age-related decreases in ovarian follicle numbers and decay in oocyte quality influence the natural loss of fecundity and ultimately the start of menopause. The rate of ovarian ageing is highly variable among women, so it is relevant to be able to identify women who have severely decreased ovarian reserve for their age. Spontaneous Premature Ovarian Failure (POF) is the most extreme phenotype of ovarian failure at a young age. POF is characterized by spontaneous secondary amenorrhea before the age of 40. POF not only truncates the fertile lifespan, but also has major implications for long-term health (i.e. osteoporosis, cardiovascular health and cognition). The aetiology of the great majority of spontaneous POF cases remains unknown. Since family history has been shown to be the best predictor for early menopause and strong associations have been disclosed between the menopausal ages of mothers and daughters, sisters and twin pairs, idiopathic POF is most likely due to genetic factors. The aim of this thesis was to further elucidate the phenotype of premature ovarian failure and to explore the genetic mechanisms underlying this phenotype. Classically, ovarian dysfunction is categorized on the basis of cycle history and follicle-stimulating hormone (FSH). Our study indicated that anti-Mllerian hormone, a novel ovarian marker, may provide a more accurate assessment of the follicle pool in young hypergonadotropic patients, especially in the clinically challenging subgroups of patients with elevated FSH and regular menses or with cycle disturbances but not fulfilling the POF diagnostic criteria. Early menopause is associated with a higher incidence of cardiovascular events later in life. Concurrent with the menopausal transition, there is a shift in lipid profile. POF provides a route to studying the effect of cessation of ovarian function on lipid profile independent of the effects of advanced chronological age. We found that loss of ovarian function at a very young age (POF) coincides with subtle changes in lipid profile (higher triglyceride levels, marginal lower HDL-cholesterol). Androgens are better markers for unfavourable lipid changes than oestrogen levels or duration of oestrogen deprivation. A genome wide association study was initiated to identify new genetic variants involved in POF. We did observe a possible association with a single nucleotide polymorphism in ADAMTS19, a biologically plausible candidate gene. This finding warrants a larger follow-up study to investigate its role in POF. Besides SNPs, the human genome contains multiple copy number variants (CNVs), small, sub-microscopic deletions and duplications. POF is associated with macroscopic deletions, in particular on the Xq arm. Using the intensities of high-density, genome-wide oligonucleotide arrays, it is now possible to obtain an increase in resolution of more than 100 times compared to conventional karyograms, and it is thus possible to detect CNVs. Sixty-three percent of the long arm CNVs (90% deletion) mapped to locus Xq21. Our findings are preliminary although the abundant overrepresentation of CNVs on Xq21 further substantiates the important role of this locus in the genetics of POF.