Abstract
Five classical designations of sickle haplotypes are made on the basis of the presence or absence of restriction sites and are named after the ethno-linguistic groups or geographic regions from which the individuals with sickle cell anemia originated. Each haplotype is thought to represent an independent occurrence of the sickle mutation rs334 (c.20A>T [p.Glu7Val] in HBB). We investigated the origins of the sickle mutation by using whole-genome-sequence data. We identified 156 carriers from the 1000 Genomes Project, the African Genome Variation Project, and Qatar. We classified haplotypes by using 27 polymorphisms in linkage disequilibrium with rs334. Network analysis revealed a common haplotype that differed from the ancestral haplotype only by the derived sickle mutation at rs334 and correlated collectively with the Central African Republic (CAR), Cameroon, and Arabian/Indian haplotypes. Other haplotypes were derived from this haplotype and fell into two clusters, one composed of Senegal haplotypes and the other composed of Benin and Senegal haplotypes. The near-exclusive presence of the original sickle haplotype in the CAR, Kenya, Uganda, and South Africa is consistent with this haplotype predating the Bantu expansions. Modeling of balancing selection indicated that the heterozygote advantage was 15.2%, an equilibrium frequency of 12.0% was reached after 87 generations, and the selective environment predated the mutation. The posterior distribution of the ancestral recombination graph yielded a sickle mutation age of 259 generations, corresponding to 7,300 years ago during the Holocene Wet Phase. These results clarify the origin of the sickle allele and improve and simplify the classification of sickle haplotypes.