Genome study points to evidence of natural selection

The scientists,led by Li Jin of the Chinese Academy of Sciences in Shanghai,report in the journal Genome Research that certain disease-causing variant genes became more common in African-Americans after their ancestors reached US shoresperhaps because they conferred greater,offsetting benefits. Other gene variants have become less common,the researchers say,like the gene for sickle cell haemoglobin,which in its more common single-dose form protects against malaria. The Shanghai team suggests the gene has become less common in African-Americans because malaria is much less of a threat.

The purpose of studying African-American genomes is largely medical. Most searches for variant genes that cause disease take place in people of European ancestry,and physicians want to make sure they have not missed variants that may be more common in African-Americans and helpful for developing treatments or diagnosis. Such searches often reveal events in a populations history by pinpointing genes that have changed under the pressure of natural selection.

The unusually common variants identified by the Shanghai team are associated with higher risk of hypertension,prostate cancer,sclerosis and bladder cancer. Most of the genes associated with African-American ethnic diseases, they write,may have played an important role in African-Americans adaptation to local environment. But the authors have not yet been able to identify the benefits they believe such genes conferred.

Mark D Shriver,a geneticist at Penn State,said it was plausible that some versions of a gene would become more common as African-Americans adjusted to a new environment. Its very valid to expect that there will be factors subject to genetic adaptation and that are now more prevalent in contemporary African-Americans than in the ancestral group, he said.

The Shanghai researchers used a method for studying admixture,a geneticists term for when two populations or races intermarry; China has several such populations,perhaps accounting for the teams interest. Using gene chips that analyse common variations in the human genome,researchers can deconstruct the chromosomes of an African-American,say,assigning each chunk of DNA to an African or European origin.

The scientists found that of the African-American genomes in their sample,22 per cent of the DNA came from Europeans,on average,and the rest from African ancestors,a figure in line with other estimates. They then looked for sites along the genome where either European or African ancestry was present at statistically significant levels above the average,finding four regions with very common European ancestry and two with very common African ancestry. Most of these sites harboured genes of unknown function,but one,of European origin,holds a gene that combats influenza,suggesting it has become more common in African-Americans by conferring protection from the disease.

In another approach,the Shanghai team focused on all the DNA segments of the African origin in the African-American genomes,discarding all the European DNA. They then compared the African component of African-American genomes with the DNA of the Yoruba of Nigeria,a well-studied population that happens to be genetically very close to the West African population from which many slaves were taken.

The Shanghai team then asked how the African genome had changed after Africans arrived in the United States. They found that versions of some genes had become more common and others less so. The less common genes included several known to be involved in protection against malaria.

Researchers can analyse the ancestry of admixed populations because of the way the hereditary material is shuffled between generations. People have a double set of chromosomes,of which one member of each pair comes from the mother and one from the father. When the egg or sperm is made,the maternal and paternal copies of a chromosome line up and swap large chunks of DNA. The swapped segments are so large that it takes many generations before they are whittled down to a length too small to be recognised. Meanwhile,the ancestry of each segment can be identified from its pattern of single-nucleotide polymorphisms,or SNPs,the sites on the human genome where there is commonly variation in the A,T,C and G units that make up DNA.

Geneticists can thus deconstruct the genomes of admixed populations into a mosaic in which each segment can be traced back to one or the other of the two parent populations. This is the basis of the Shanghai teams approach. But proving that natural selection has been at work in very recent timesin this case,the last 300 yearsis very difficult,because the traces of selection are still small. To be sure of detecting such weak selection signals,Jin and his colleagues conclude,researchers in the future should analyse many thousands of genomes.