Many genetic studies have assumed that modern human populations were panmictic and genetically unstructured for much of their prehistory, especially in Africa. However, some recent evidence has suggested that prehistoric populations within continental regions were more subdivided with distinct genetic lineages and patterns of linkage disequilibrium (LD). Thus, human demographic history may arguably have been more complex than previously depicted by more simple models. However, more research is necessary in order to characterize the extent of this structure across a broader range of geographically and ethnically diverse populations. The demographic history of human populations also has implications for the evolution of genetic variation underlying complex diseases. Particularly, variants for late-onset disease (not under strong selection) are subject to the same demographic forces as neutral variation. Thus, under a model of population subdivision, alleles which predispose individuals to disease-related traits are predicted to occur at different frequencies in diverse populations, affecting the risk for acquiring disease. This project will examine SNP haplotype variation in the SORT1 and SORL1 genes, two newly identified risk factors for late-onset Alzheimer's Disease (LOAD): (a) to characterize the genetic structure and patterns of LD within major continental regions (~1500 individuals from 56 populations) (b) to infer demographic models that explain observed patterns of diversity, and (c) to determine if genetic structure contributes to differential risk for LOAD in African-Americans, Caribbean-Hispanics and European-Americans (~900 cases and ethnically-matched controls). In particular, this study will test for geographic variation in haplotype frequencies among diverse populations within continents, as well as correlations between population subdivision and variance in LD. This project will also test whether or not genetic risk for LOAD is different in populations with distinct population histories. The researchers in this project will select SNPs located in coding and non-coding regions of SORT1 and SORL1 from the International HapMap Project for High Throughput Genotyping. Haplotypes will be reconstructed using two algorithms with different underlying assumptions and other analytical methods, such as Wright's Fst and Lewontin's D' parameter, will be used to calculate population divergence and variance in LD. Disease-marker association analyses between LOAD and the above genes will also be performed using cases and ethnically-matched controls. Intellectual Merit: This project represents the first study of SNP variation in both genes in a large sample of healthy individuals from diverse human groups, including 10 distinct African populations which are mostly underrepresented in genetic studies. By conducting this research, it will lead to more detailed knowledge of the demographic history of human populations, and ultimately modern human origins. Broader Impacts: On a broader scale, the effective design of therapeutic strategies for LOAD will be enhanced by more detailed information on genetic factors involved in disease susceptibility in different human groups. In the end, treatments can be developed that will effectively decrease the incidence of LOAD across the human species. This project will also directly support the training of a minority graduate student, from an underrepresented group, in molecular biology and population genetics. In addition, the data collected from this research, including SNPs, will be made available publicly to online databases.
