Molecular approach to common causes of female infertility.

Pivotal genetic information has been derived for a host of rare genetic disorders, but progress has been much slower in relation to the common causes of female infertility. In this chapter, we shall illustrate the approaches being applied in elucidating conditions causing infertility that are inherited in a polygenic/multifactorial fashion. The task is to determine the number of genes responsible and their chromosomal location(s). The first approach is to use genome-wide quantitative linkage analysis, searching throughout the genome with no prior expectation that a given gene or chromosomal region is casually involved. A second approach is to search across the genome for altered gene expression, for example comparing endometriosis and normal (non-endometriosis)cells. The third approach is less indiscriminate and more focused, depending upon identifying specific candidate genes. Aromatase, calhedrin, oestrogen receptor, galactose-1-phosphate uridyl transferase (GALT) and tumour suppressor genes such as p53 are attractive candidate genes for endometriosis. Endometriosis, which has long been suspected to possess a familial tendency, has been subjected to genome-wide linkage analysis in Oxford, UK, where sib-pair analysis uses polymorphic DNA markers and fluorescence-based automated analysis. Several regions of exclusion have been found, but no linkages have so far been reported. A candidate gene approach focuses on the presence of chromosomal aberrations, the assumption being that endometriosis parallels neoplasia. At Baylor College of Medicine, we thus began by showing chromosome alterations involving trisomy 11, monosomy 16 and monosomy 17 in late-stage endometriosis. A loss of only the p53 tumour suppressor gene, rather than a loss (monosomy) of chromosome 17 per se, however, seems to be the pivotal event. A second representative polygenic/multifactorial disorder causing female infertility is polycystic ovarian syndrome. Both quantitative linkage analysis and candidate gene approaches are being pursued. In the far more commonly observed 'idiopathic' variety (non-adrenal polycystic ovarian syndrome and hirsutism), consensus has long existed that one or more dominant genes causes the condition. Although the mode of inheritance in 'essential' polycystic ovarian syndrome remains uncertain, dominant tendencies are clearly more pertinent than recessive ones. Genes for adrenal biosynthetic enzymes, insulin receptors, leptin and leptin receptors, follistatin, activin and inhibins are attractive candidates for polycystic ovarian disease. A linkage to 37 candidate genes was sought using affected sib-pair analysis and transmission/disequilibrium methods.