For much of nature, 'survival of the fittest', the mantra of evolution, holds sway. A few thousand, even a few hundred years ago, the same was true for humans: only the 'fittest' – the best adapted – survived to reproduce. But today, in many parts of the world, improved hygiene and healthcare, food production, and heating and cooling systems have minimised the dangers in human lives. So if selective pressures have been taken away, is evolution over for humans?
It appears not: natural selection is known to be at work on some genetic variants, involved in some of the most fundamental influences on life: food, reproduction and infection. However, these variants are probably just the tip of the iceberg: "We only know about them because they are extreme," points out Dr Gil McVean, a biological statistician at the University of Oxford.
To find other, more subtle variants, researchers are delving into the data produced by the International HapMap Project, which catalogued millions of gene variants in 270 Yoruba Nigerians, Utah Europeans, Han Chinese and Japanese. (See News: International consortium completes map of human genetic variation .) For example, in a scan of the data, Dr Jonathan Pritchard and colleagues at the University of Chicago identified 700 variants that showed evidence of recent natural positive selection (see box, below). The genes were most commonly involved in the sense of smell, in reproduction, in the metabolism of carbohydrates and foreign compounds, and in brain development. They also found five genes involved in skin pigmentation that showed evidence of positive selection in Europeans.
Working out the role of the variants, and why they have been selected, is still a difficult task. "The hypothesis of natural selection is fundamentally untestable," points out Dr McVean. "We can't go back in time and measure the fitness of individuals with certain variants. All we can do is say that we think selection has happened."
Food and families
One of the best examples of recent selection involves milk, which adults in most parts of the world cannot drink. After weaning, the lactase gene, which produces the enzyme that breaks down the lactose sugar in milk, is usually switched off. Yet more than 70 per cent of European adults are quite happy drinking milk, as they carry a variation in the gene's control region that allows lactase production to continue. The genetic change appears to have happened between 5000 and 10 000 years ago, around the time that dairy farming was developed.
"We can speculate that in western Europe the ability to survive and reproduce must have depended on the ability to use milk as adults," says Dr Chris Tyler-Smith at the Wellcome Trust Sanger Institute. "This applied strong selection pressure in favour of the mutant form of the gene. It has become so common that Western medicine now sees the inability to digest lactose by adults as an illness – lactose intolerance. Presumably a few thousand years from now, if selection pressure remains the same, everyone will have the selected mutation."
A different type of genetic variation that appears to be under positive selection – a piece of chromosome 17 that in 20 per cent of Europeans is inverted – seems to affect family size. In Iceland, females with one copy of the inverted form were found to have about 3.5 per cent more children than those without, suggesting that the inversion it is in some way enhancing fertility. This variant seems to be becoming more common across Europe. "As long as there's variation in the number of offspring people have, we won't all be making equal contributions to the next generation, so we will continue to evolve," says Dr Tyler-Smith.
'Scars on our genome'
The strongest evolutionary pressure of all comes from infectious disease. "Selection is happening today: millions of people are still dying from infectious diseases," says Professor Adrian Hill of the Wellcome Trust Centre for Human Genetics in Oxford. People whose genetic make-up allows them to survive are most likely to pass on their genes to their offspring, but the genetic variants that help them to do so – the "scars on our genome", as Professor Hill describes them – can come at a cost.
A single copy of the sickle-cell variant of the b-globin gene brings protection against the Plasmodium falciparum malaria parasite, but two copies lead to life-threatening anaemia. Similarly, in Melanesia, people with one copy of a mutation in the AE1 gene that causes ovalocytosis - where red blood cells are slightly oval or elliptical - have mild anaemia and are protected against falciparum malaria, while two copies of the mutation are lethal before birth. Population-wide, the benefits of carrying a single copy of such variants outweigh the costs of carrying two, but losing both copies of a gene is not always detrimental. People who make no Duffy antigen on their red blood cells are completely resistant to the Plasmodium vivax malaria parasite. And almost all humans have lost the caspase-12 gene, probably because this improves resistance to bacterial sepsis.
Feature: Tracing human evolution
Other human genes that influence infection include: the H protein, secreted in saliva, as regards Norovirus (Norwalk or 'winter vomiting virus'); a prion gene variant found in 40 per cent of the UK population that, as Professor Hill describes, "gives people complete protection against infection with new variant CJD"; and the PTPN22 gene, which, intriguingly, seems to be linked both to risk of bacterial infection and to autoimmune disease.
Such genetic variants have presumably risen in frequency through hundreds or thousands of years of natural selection. But there is a modern-day force that may be driving human evolution at an unprecedented rate: HIV. In certain parts of South Africa, for example, nearly half of women are infected with the virus. In a study in Durban, Dr Philip Goulder (University of Oxford) and colleagues found that women with a certain combination of human leukocyte antigen (HLA) variants – HLA-B27 – were better at clearing HIV infection than those with HLA-A or HLA-C genetic subtypes. (HLAs are produced by the major histocompatibility complex, by far the most variable region of the human genome.) Infected mothers with protected HLA-B genes are more likely to survive and pass on those genes to their children.
It may be possible that the relatively low prevalence of HIV in western Europe is aided by a common variation in a co-receptor for the HIV virus particle (CCR5). The variant, which protects people almost completely against HIV, is found in 13 per cent of Europeans but is extremely rare in other populations, including Africans. As the origins of the variant in humans date to thousands of years ago, and the AIDS epidemic dates only from the late 1970s at the earliest, it is likely that the mutation may have been selected because it conferred resistance to other infections, such as bubonic plague or smallpox – or it may even be entirely neutral.
Where do we go from here? Some fear our genetic stock is weakening, as natural selection is no longer winnowing out the sickly and their detrimental genes (a view that has uncomfortable eugenic undertones to it). Another view is that the next stage in human evolution will not be wholly biological – that we will embrace new technologies to create enhanced forms of human being.
But a more pessimistic view is that nature has a way of challenging human hubris. Cataclysmic change to the planet, perhaps triggered by runaway global warming, could radically change human destiny. Natural selection could return with a vengeance, and who knows what features would be selected for in this human future.
How are positively selected genetic variants identified? Not easily, but they can be found by comparing the distribution of genetic variants in the genomes of different humans.
When variants are under positive selection and increase in frequency in a population, they leave patterns - or 'signatures' - in DNA sequences.
For example, as a variant starts to spread because of positive selection, it carries its surrounding regions with it, so haplotypes that are both long and common are one of the first signs that selection is acting. If selection occurs in one population but not in another, frequency differences between populations provide another early sign. As the selected allele starts to dominate, it can influence allele frequencies and even depress the overall level of variation in its vicinity. And if there are repeated alterations to the amino acid sequence, these stand out.
Unfortunately, even in the absence of selection, any of these patterns can turn up by chance, especially when the whole genome is examined. Even worse, events such as population expansions can mimic some of the same effects. There is no perfect way to recognise the signature of selection, but we sometimes get a very strong hint.