Evidence Of Recent Natural Selection On The Southeast Asian Deletion(--^SEA) Causing α-thalassemia In Southern China

Background and Objectivea-thalassemia is one of the most common hemoglobinopathies and is widely distributed at high frequencies throughout all tropical and subtropical regions worldwide. It results from reduced production of a-globin chain,eading toimbalance in theproportions of α-and β-chains in hemoglobin. Most documented cases ofa-thalassemia are caused by deletions of a-globin.They can be divided into α0-thalassemia (--) and a+-thalassemia(-a), based on the number of remaining functional a-globin genes. The more functional genes are lost, the more severethe phenotype caused. Homozygotes for α0-thalassemia are lethal and stillborn. The Southeast Asian (--SEA) deletion is the most common type of α0-thalassemia in the populations of Southeast Asia, including southern China, Epidemiological surveys in Guangdong and Guangxi provinces suggest that the-SEA variant is found at a frequency of around4-8%in southern ChinaIts local presence and still relatively high frequency, considering theseverity of its phenotype, indicate that this allele may havebeeninfiuencedbyselectivepressure.exerted as a result of malaria. Many other variants of hemoglobinopathies, such as HbS, HbE, HbC and βCD4I/42, have been demonstrated to contribute to malarial resistance by bot h epidemiologicalconfirmation and/or novel haplotype-based molecular evidence. Howe ver, no credible evidence of association between malaria and-SEA or other variants of a0-thalassemia has been reported to date. Fortunately, withtheadventofdensemapsofhum angeneticvariation such as HapMap project and SGVP project, and the development of various kinds of rapid and low-cost genotype techniques, itisnowconvenienttodetectwh ether positivenaturalselection operates on variants of interest.In this study, we report the results of our molecular study, based on a population from southern China. We conducted analyses, based on haplotype, to examine whether the--SEA alleles has been the target of recent natural selection.Materials and MethodsWe typed214individuals from southern China. They consisted of76-SEA/aa carriers and138wild-type individuals. The structure of the sample was non-random, so a constructed random sample (CRS) was built on the basis of the population frequency of the-SEA deletion. Specifically, we randomly excluded53-SEA/aa carriers for a total frequency of7%. In further analyses, we carried out some tests with both data sets. We scanned a fragment from the beginning of chromosome16to470kb, about300kb behind the downstream deletion breakpoint, to screen appropriate SNPs based on the data of CHB population in HapMap. Additionally, three extra SNPs in the proximal region upstream of the deletion breakpoint, from an earlier study in our laboratory, were also included. As a result, in total,28SNPs encompassing the-SEA deletion and covering a～410kb region were determined. For genotyping these SNPs, PCR was performed and amplicons were analyzed usinghigh-resolution melting (HRM) analysis after sequencing about10samples for each SNP.Haplotypes were inferred by a Bayesian statistical method with the PHASE2.1software after Hardy-Weinberg Equilibrium Test (HWE). Reconstructed haplotypes wereinserted into Haploview v.4.2to estimate LD statistics andD’. The minimum number of recombination events RM between haplotypes with the--SEA deletion and wild-type were compared using R language. Relative extended haplotype homozygosity (REHH) values were calculated using the Sweep1.1software. The inferred haplotype network was constructed using NETWORK4.6.1.0by a median-joining method.Results and DiscussionNone of the markers, including the--SEA deletion, deviated from HWE at a significance level of0.05except rs216091(χ2=4.66, P=0.031) in CRS. A double check.sequencing22individuals for rs216091, ruled out the possibility of genotyping errors. Additionally, the genotype frequency of rs216091was similar to that of the SGVPCHS population.Result of LD analysis suggests that the--deletion still showeda high degree of|D’|with other SNPs, although a recombination hotspot breaks down the LD downstream of the SEA deletion. Strong LD is hardly observed for SNPs more distant than～80kb in Asians, much smaller than the observed LD from the--SEA to SNPs outside the deletion locus (～200kb). Two SNPs, rs77308790and rs3760053, that are rare variants (<0.1) in the normal population but were predominant in carriers, showeda pattern similar to--SEA, indicating ahitchhiking effect due to linkage with the--SEA deletion.The increased LD might be due to a lack of recombination in the deletion region on chromo somes carrying the causal allele rather than natural selection. To assess this, we scanned the genetic maps of CHS and CHB. No hotspot was found in the deletion region. Furtherevaluation of the two regions upstream and downstream of the deletion can be seen that recombination events seem to occur more actively on normal chromosomes than--SEA chromosomes in both regions. These two observations indicated LD of--SEA chromosomes was not likely the consequenceof a lack of recombination in the deletion region on chromosomes. LRH test showed a significantly high REHH value (P=0.004for total samples and P=0.03for CRS). indicating selection favoring the-SEA deletion.This signature probably results from the malarial environment through estimating the age of the--SEA allele, which is consistent with records of malaria epidemiology.Phylogenetic network that--SEA appears to come from an Asian-specific haplotypic background, is consistent with the localized distribution of this mutation, indicating a recent origin of the deletion.Thus, to sum up the points which we have just indicated, our data support the hypothesis that the-SEA allele has been subjected to recent balance selection, caused by malaria.