| || |
| Out of Africa? |
As with determining the relationships of the apes, the first DNA-based studies of the relationships of human populations also used mitochondrial DNA. In mammals, mitochondria have an interesting inheritance pattern: they are transmitted nearly exclusively along maternal lines. Although males have mitochondria, they do not transmit them to their offspring. Thus, all of your mitochondria came from your maternal grandmother and, by extension, your maternal-maternal great-grandmother. In 1987 Rebecca Cann, Mark Stoneking, and Alan Wilson (thenat University of California-Berkeley) published a controversial and provocative paper in Nature, stating that they had located the common ancestor of all mitochondrial variants - the so-called Mitochondrial Eve (Fig. 4). They placed her in Africa approximately 200,000 years ago; subsequent studies have found similar results. Fifteen years after that first paper, the results remain a source of interest and controversy.
Why are the Mitochondrial Eve studies a continual source of controversy within the human evolutionary genetics community? That there is a common ancestor of mitochondrial DNA sequences is not a surprise. In fact, it is a consequence of Mendelian genetics: genes taken from any sample within a population will share a common ancestor. Take pairs of gene copies in the same population. Some of them will share the same ancestor from one generation ago; they are from siblings of the same parents. Some pairs of gene copies will trace a common ancestor two generations back. Some pairs will share common ancestry even further back. However, eventually, all copies will share a common ancestor. What is of interest is how long it takes all gene copies to coalesce to that ancestor.
What the debate focuses on is the timing and the location of the Mitochondrial Eve. The initial studies showed that there was one clade consisting only of African individuals, and one with African and other individuals. Hence, we can infer that the common ancestor lived in Africa. Numerous researchers challenged the methodology of the original study. For instance, the original study used African-American individuals instead of individuals from Africa. Most of the subsequent studies, using more data - including data from individuals from several African tribes - and better methodology seem to confirm that Africa is the location of the common ancestor.
To determine the age of the Mitochondrial Eve, biologists need to first make assumptions about the way evolution proceeds. The usual assumption is that changes in the DNA occur roughly in a clock-like fashion - that there is a so-called molecular clock. The molecular clock assumes that groups separated by twenty nucleotide changes have common ancestors that are roughly twice as old those separated by ten nucleotide changes. No one believes DNA evolution proceeds in a perfect clock-like manner. What is debated is the extent to which the clock assumption can provide an estimate about divergence times. The usefulness and the accuracy of molecular clocks have been controversial ever since Zuckerkandel and Pauling proposed them in the 1960s. Yet, most evolutionary biologists agree that the molecular clock concept has at least some validity.
Inferring dates based on a molecular clock also requires that one calibrate the clock. How quickly do the changes occur in the lineage(s) of interest? Different molecular clocks based on different regions of the genome or different types of organisms don't all tick at the same rate. To calibrate a molecular clock, researchers usually use a lineage split for which they have at least some degree of confidence about when it occurred. They then divide the amount of genetic divergence by the time when the groups last shared a common ancestor. In the case of mtDNA, the calibration was set by the human-chimp split of six million years. Because chimps and humans differ by about twelve percent of nucleotides in mtDNA, the rate of change for the hominid lineage for mtDNA is about two percent per million years. The average total divergence from contemporary sequences to the inferred sequence of the mtEve is about 0.4% and, thus, the divergence time is about 200,000 years. The confidence limits, however, of this estimate are rather large. Christopher Wills once concluded that it is possible that the upper-end of mtEve's age may be as much as 800,000 years; new data places his latest estimate at 400,000 years. 3
Different genes will often have different evolutionary histories. One should not expect the male equivalent ("the Y chromosome Adam") to have lived at the same place and the same time as the Mitochondrial Eve. Owing in part to having a lower mutation rate, the human Y chromosome generally has less variation than the mitochondria, which makes analysis more difficult. Nonetheless, recent studies suggest that the last common ancestor of all existing human Y chromosomes also lived in Africa - but more recently than Mitochondrial Eve.
Largely from the Mitochondrial Eve studies, one model - the out of Africa hypothesis - gained favor among anthropologists and human evolutionary geneticists. This hypothesis, which is sometimes called the "replacement hypothesis," postulates that modern Homo sapiens spread out of Africa, into Europe and Asia, and replaced archaic Homo sapiens living in those regions (Fig. 5). In contrast, Milford Wolpoff and others have proposed the multiregional hypothesis. They argue that the archaic Homo sapiens populations in the different regions (Europe, Asia, and Africa) all evolved together into modern Homo sapiens. While genetic changes would first occur in one locality, gene flow would spread those changes into the other localities.
The out of Africa and multiregional hypotheses make several distinct predictions. One would predict that under the out of Africa hypothesis, Africa would be the origin of the common ancestor of variants for most of the independent data sets (different genes) tested. The multiregional hypothesis would predict a random pattern. Under the out of Africa model, the divergence time between the African and the non-African populations would have an upper-limit of about 200,000 years. In contrast, the multiregional hypothesis would predict a divergence time of approximately one million years. One caveat is that the apparent age of the divergence could be reduced by the gene flow among the populations. Another caveat is that selection can also alter the apparent divergence times. The out of Africa hypothesis also predicts that there will be more genetic diversity within the African population than within the other populations.
As of 2003 the evidence seems to favor the out of Africa model though some intermediate positions cannot be ruled out. In nearly all of the studies more genetic diversity is seen in the African populations than in others. In addition, the divergence times appear more consistent with the out of Africa than the multiregional hypothesis. As we obtain more and more sequences from different regions from the genome, this debate should become resolved.