Investigating southern African genetic diversity and its role in TB susceptibility

[摘要] ENGLISH ABSTRACT: Recent genetic studies have established that the KhoeSan populations of southern Africa arethe earliest known indigenous inhabitants of the region and distinct from all other Africanpopulations. Owing to the region's unique history, population structure in southern Africareflects both the underlying KhoeSan genetic diversity as well as differential recentadmixture. This population structure has a wide range of biomedical and socioculturalimplications such as changes in disease risk profiles: there is a known correlation betweenancestry and tuberculosis (TB) susceptibility.Research presented in this thesis consolidates information from various population geneticstudies that characterized admixture patterns in southern Africa with the aim to improve theunderstanding of differences in adverse disease phenotypes observed among populations.Further to previous studies, genome-wide polymorphism data from more than 20 southernAfrican populations were analysed to investigate the fine-scale population structure in thearea. The analyses revealed fine-scale population structure in and around the KalahariDesert, which does not always correspond to linguistic or subsistence categories, but ratherreflects the role of ecogeographic boundaries. In addition, we showed that the Khoe adoptedtheir pastoralism through a process of largely cultural diffusion rather than demic diffusion as previously thought. The proportion and origin of KhoeSan genetic ancestry in southernAfrican populations is of particular relevance to disease, because the KhoeSan exhibitgreater variation in genetic diversity than other African populations, including unusualvariation in genes with demonstrable immune function.Utilizing data from several TB genome-wide association studies (GWAS), a bioinformaticspipeline was employed to detect regulatory polymorphisms in linkage disequilibrium withvariants previously implicated in TB susceptibility. A total of 133 predicted regulatory variantswere found. Association analyses were performed in TB cases and healthy controls andyielded six intronic functionally relevant variants. The post-GWAS approach, which included ancestry as a confounder, demonstrated the feasibility of combining multiple TB GWAS datasets with linkage information to identify regulatory variants associated with TB susceptibility.In addition to classical association studies, selection scans have the ability to identifygenomic regions associated with a phenotype. Signals of natural selection in southernAfrican populations was studied using high-coverage exome sequence data. Selectionsignals were identified in genes associated with immune response to foreign pathogensintroduced from the 12th century onwards. In addition, signals of selection were identified in pathways associated with focal adhesion and ECM receptor interaction. It is clear that thereare distinct immune-related signals of positive selection present in southern Africanpopulations.This research not only provided insight into the genetic basis and biology of human TBsusceptibility, but also harnessed the unique ancestry present in southern Africanpopulations. The addition of population genetics information was shown to greatly shape andimprove our investigations of TB susceptibility and may also apply to other phenotypesunique to southern Africa. Since the era of personalised medicine is imminent, moreinvestigations of understudied southern African populations most severely affected by TB arerequired.

[发布机构] Stellenbosch University