January 16, 2005 – In a paper published in the online edition of Nature Genetics, deCODE scientists describe the discovery of a large and ancient stretch of inverted sequence on chromosome 17. The characteristics and geographic prevalence of this inversion, analyzed in the context of deCODE’s population genetic and genealogical data, has provided a novel opportunity to directly observe natural selection at work in a human population. The paper, entitled “A common inversion under selection in Europeans,” can be accessed on the web at www.nature.com/ng, and will appear in the February print edition of Nature Genetics.

The 900-kilobase (900,000 nucleotides) inversion on the long arm of chromosome 17 is the second largest known common inversion in the human genome. In an inversion, a segment of DNA is flipped around on a chromosome, creating a new version running in the opposite orientation to the original. The authors discovered vast differences between the sequences of the original and inverted orientations, which are referred to as H1 and H2, respectively. The divergence between H1 and H2 was so great that, assuming standard rates of mutation, the inversion was estimated to have occurred at least 3 million years ago. This long predates the emergence of our species, Homo sapiens, about 150,000 years ago, and even the emergence of the first archaic humans belonging to the genus Homo. However, for such an ancient variant the inversion presents a number of striking characteristics, identified through analysis of H1 and H2 in various populations around the world. First, very little variation exists in H2, compared with the multitude of variations in H1. Second, H2 is common in Europe and the Middle East – about one-third of individuals from these regions appear to have one or two copies – but it is found in only about 10% of Africans and is virtually non-existent in Asia.

The homogeneity of H2 and its high prevalence in just one part of the world are suggestive not of an old genetic variant, which over the course of millions of years would likely have generated a large number of mutations and either become predominant or disappeared. Rather it is more consistent with H2 being a new variant that has recently expanded in frequency due to positive selection, enabling it to become quite common in this one part of the world. To directly examine whether H2 is presently undergoing positive selection in a European population, the authors analysed genotypic and genealogical data from some 30,000 Icelanders participating in the company’s genetics research and asked the fundamental question at work in positive selection: Do people who carry H2 tend to have more children than do those who do not? Indeed, individuals carrying one copy of H2 have on average approximately 3.5% more children than do those with two copies of H1. This may seem a modest increase, but would have a substantial impact over time.

What accounts for this impact? One possible factor is recombination, the shuffling of chromosomal material that takes place in the formation of eggs and sperm. In a paper published in the November 2004 edition of Nature Genetics, deCODE scientists demonstrated that higher recombination rate was positively linked to reproductive success in women. To test whether recombination rate might contribute to the observed selective advantage conferred by H2, the authors correlated H2 status and recombination rate through the analysis of genotypic information on some 21,000 Icelanders. The results demonstrate that those with the inversion do tend to have higher recombination rates than do those without it, although the increase in recombination likely accounts for only a fraction of the impact that H2 has on increased number of children.

The ability to directly observe that H2 is linked with an increased number of children and with higher recombination rate is a milestone in human genetics. Never before has a specific genetic variant been linked to number of children and recombination rate; nor has it been possible to observe the effect of such a variant on reproductive success in real time – to observe natural selection as it takes place in a human population.

Yet the findings in the paper also raise important further questions. Perhaps the most tantalizing is how a genetic variant that appeared so long ago could have such limited variability and yet be so common, in only one branch of the human family. The authors offer two possible explanations for this. The first is that H1 and H2 were both maintained in the gene pool in a state of so-called “balancing selection,” in which individuals with one copy of both the H1 and H2 orientations enjoyed some minor selective advantage, preventing either from being lost as would ordinarily happen by sheer chance over such a long period of time. H2 could then have expanded more recently in Europe through some episode of positive selection.

The second explanation is that H2 may have emerged in the ancestral gene pool of an archaic human species such as Homo heidelbergensis or Homo erectus. These species coexisted for a time with early Homo sapiens, and H2 could have been introduced into the modern human genome through interbreeding between these archaic humans and our direct forbears, before the migration of anatomically modern humans out of Africa some 60,000 years ago.

Whether maintained through balancing selection or introduced through admixture with archaic humans, the relatively high prevalence and extreme homogeneity of H2 in the populations of Europe and the Middle East is most consistent with carriers of H2 in this region having experienced some form of positive selective pressure after their ancestors migrated out of Africa.