Using Copy Number Variation to Identify Severe Obesity Genes in Utah Pedigrees

University Of Utah, Salt Lake City UT

Investigators

Linked publications & trials

Abstract

Severe obesity (BMI&#8805;35 kg/m2) often has early onset, is accompanied by strong risk factors for disease and early mortality, is difficult to treat short of surgery, and long-term maintenance of weight loss is usually unsuccessful. While there is clearly a significant genetic component to severe obesity, linkage studies, candidate gene studies, and recent genome-wide association studies using single nucleotide polymorphisms (SNPs) have detected genes that each explain less than 1% of the variance in obesity. Obviously, there must be significant additional genetic variation that can be uncovered by other methodologies. This application proposes to detect regions along each chromosome that have either been deleted or which have additional copies of the region inserted that could explain significant portions of the unexplained genetic variance in severe obesity. These variable chromosomal regions are called copy number variants (CNVs). This study proposes to detect CNVs by comparative genome hybridization of 60 basepair oligonucleotides on NimbleGen 720K tiling arrays specifically designed for CNV detection. The advantages of CNV methodology are that SNP arrays often fail to capture CNV regions, as there is often evidence of lack of Hardy-Weinberg equilibrium in these regions (e.g. deletions). These conditions usually cause SNPs to be discarded. The remaining nearby SNPs often have low linkage disequilibrium with the CNV regions. Once association is found with a CNV, it is likely to be of greater functional significance than a SNP and has better defined boundaries within which to look for specific gene copy numbers or mutations. The samples to be used for the genome-wide CNV association study include 107 three-generation Utah pedigrees previously ascertained for high familial aggregation of severe obesity (five or more relatives with BMI&#8805;35 kg/m2). Pedigrees have the advantage of reducing the nondisease-related CNV noise in the genome. They also limit the false positive findings in such a large scan by requiring that severe obesity and the CNV variant cosegregate with each other in most affected pedigree members and seldom in the unaffected members. We will type 1500 members of these pedigrees who are either severely obese or normal weight (BMI<25 kg/m2). The significant regions detected will be validated using denser custom arrays on a casecontrol series of 2000 Caucasian unrelated subjects who are either severely obese or normal weight and 1000 White and 1000 African-American subjects with the full range of BMI values. In addition, three European population studies, two of them prospective and one having available CT scans for visceral fat, will be used as an additional replication sample (8582 subjects) for a total of 14,082 subjects in this study. This highly informative set of pedigrees will be used to further unravel the complexity of the genetic underpinnings of severe obesity and may contribute to the understanding of less severe obesity.