Q J Med 2004; 97: 57-58
© Association of Physicians 2004; all rights reserved.
Biologic |
Why?
Until closely studied, the human Y chromosome had been identified as the home of relatively few genes: the IEAT (Inability to Express Affection on the Telephone), HUH (selective hearing loss), and Flip (rapid TV channel selection) genes were among the few examples. More detailed and serious examination has revealed a great deal more. Y is not essential, about half of it consists of tandemly repeated satellite DNA, and it carries few genes, most of which do not recombine. Nonetheless, it can tell us a lot about ourselves.
Most of the Y escapes meiotic recombination (two segments do recombine with the X, but they represent < 3 Mb of its 60 Mb length, and I shall ignore them here). This means that its haplotypes (the combinations of allelic states of markers along the chromosomes) can change only by mutation, and not by reshuffling. So single-nucleotide polymorphisms on the Y tell us much more about our history than do their autosomal equivalents.
In a typical mating pair, there will be four autosomes, three Xs and one Y, so the effective population size of the Y is a quarter of an autosome, or a third of an X, and roughly the same as the mitochondrial DNA (mtDNA). If the mutation rate is the same on all chromosomes (see reference 1), there should be less sequence diversity on the Y, and there is. There is also more genetic drift, because of changes in the frequency of haplotypes brought about by random sampling from one generation to the next, and this helps to define populations of Y. So does patrilocality (women tend to move to the area where their man was born). Where matrilocality is the rule (e.g. Thailand), you find enhanced mtDNA differentiation.
Jump a bit. In 1912, Weinberg noticed that children with dominant achondroplasia were usually among the last-born in a family. With remarkable insight, he suggested a mutation as the cause; it was Penrose who showed the effect was due to paternal age. This suggested a much higher mutation rate in males vs. females, a finding true in haemophilia and autosomal dominant disorders. More recently, it has been shown that because mutations can be identified unambiguously on the Y, it is possible to say that the low average mutation rate in Man (2 x 10-8 per base per generation) cannot be applied to the whole chromosome. There are clearly vulnerable areas such as the highly polymorphic and atypical AT-rich locus, where a high mutation rate is thought to be due to unequal sister chromatid exchange, and those areas where replication slippage is favoured by the secondary structure of repeats.
So there is some need for caution in using the Y to estimate the time to the most recent common ancestor (TMRCA)—the so-called coalescence time. In this type of study, the assumptions that are commonly made include a constant population size and random mating; an assumption about what constitutes a generation must also be made. It is usually set at 20 years, although it is clearly different for men and women: two studies suggest 35 years male vs. 29 years female in Quebec, or 31 years male vs. 28 years female in Iceland. With the effective population size of Y a quarter of an autosome, or a third of an X, the TMRCA should be less than a quarter or a third of these, as the coalescence time of a locus is proportional to its population size in an exponentially expanding population (the common assumption). If not, there must be a reason. Sometimes the relationship between coalescence time and population can be affected by selection, but heterozygosity is impossible at a haploid locus (and thus cannot be favoured), and frequency-dependent selection has not been found. There are clear difficulties with these assumptions, but in a cowardly way, I have avoided areas of discussion where they may make a difference.
They are not important in talking about Genghis Kahn. About 8% of the Y chromosomes sampled from a region of Central Asia belong to a closely related cluster of lineages in haplogroup C with a coalescence time of 1000 years (a haplogroup is a grouping of binary markers which is more stable but less specific than one defined by microsatellites). The cluster was found in 16 populations including the Han Chinese, the largest ethnic group in the world, and drift could not have accounted for this high frequency. It appears that the cluster originated in Mongolia and matches the former Mongol Empire in distribution. So the 20 000 descendants of the Kahn reported in 1260 have spread it widely.
But what about less extravagant genetic donors? No ancient lineages (older than 200 000 years) of the Y chromosome have been found anywhere. It all began in Africa with a divergence into two lineages around 1.8 million years ago, but the modern Y arose fairly recently (about 50 000 years ago), again in Africa, and ‘replaced’ those in Homo erectus and heidelbergensis. There was probably an early southern migration to southeast Asia and Australia, and another to Asia proper and Russia 7000–10 000 years later. Haplogroup data support this pattern of movement, and suggest four main groupings: in Africa; in south-eastern Asia and Australia; Central and Western Asia and Europe; and the last in the regions that were settled later—the far North, the Pacific and the Americas.
More recently, evidence has been found for recent spread of haplogroup E chromosomes by Bantu-speaking farmers from West Africa, starting around two and a half thousand years ago (almost all African Ys are now in this group). In pre-history, a group has migrated back into Africa from Europe, bringing haplogroup R back to the Cameroons. This group is presently found in 37% of the population of Western Europe.
This is an overview in every sense. Detailed analysis tells far more, for example, in areas where population densities are low, such as Siberia. But the reinforcement of palaeological and anthropological data has been valuable, and has shed light on the role of developing agriculture in the success of Man.
References
1. Berry C. The molecular clock. Q J Med 2003; 96:545.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||