Investigation of loci influencing osteoporosis using DNA pooling
A major predictor of osteoporotic fracture is bone mineral density (BMD), and BMD variance has been shown through twin and family studies to be determined genetically in some cases up to 85% (the majority of genes as yet undefined). Attempts to identify these genes has proved challenging, genome-wide linkage studies have located a large number of BMD influencing loci (quantitative trait loci (QTLs)) but these regions usually remain large 1030 cM, subsequently requiring fine mapping to find positional candidate genes. The need for large scale genotyping studies is necessary to find genes which exert modest effects like those thought to act upon BMD regulation, in this thesis I show how DNA pooling of cases (low BMD) and controls (high BMD) can be compared to look for significant differences in allelic distribution. As different loci have been linked with different skeletal sites, two DNA pool sets were constructed one for lumbar spine (LS) BMD and the other for femoral neck (FN) BMD. The DNA pools were utilised to assess if five single nucleotide polymorphisms (SNPs) in a monogenic bone disease gene (SOST) influence normal LS BMD regulation, in this study the SOST SNP’s were shown to have no influence on LS BMD regulation, however the DNA pools were shown to successfully predict allele frequencies of the individuals pooled using PyrosequencingTM but not using direct sequencing methods. A LS BMD QTL localised in a mouse and then the syntenic human region (Xp22) was located and fine mapped using DNA pools, this successfully showed that the human QTL influenced LS BMD and narrowed the potential candidate genes to just six. Thus validating the use of cross-species research in the elucidation of complex diseases using DNA pooling and linkage disequilibrium mapping.