¿ìè¶ÌÊÓƵ

In the latest round of the Eve debate, two molecular biologists in the
US present further evidence for the theory that homo sapiens evolved in
Africa. A dramatic population expansion about 60 000 years ago followed
the migration of the first modern humans out of Africa, say the researchers.
This is one interpretation of their most recent analysis of the variability
of human DNA (Proceedings of the National Academy of Sciences, vol 88, p
1597).

Allan Wilson, the chief exponent of the use of DNA in anthropology,
and Anna Di Rienzo at the University of California, Berkeley, analysed mitochondrial
DNA from the blood or hair of 117 living people from Sardinia and the Middle
East. Their goal was not so much to trace modern humans back to Africa (past
papers from Wilson’s team deal with this) but to find out when the great
migration from Africa occurred.

To do so, the researchers sequenced a fast-evolving part of the mitochondrial
genome of the 177 Caucasians. This allowed them to piece together a ‘high-resolution’
picture of Caucasian history. The most striking finding was a burst of variation
in the DNA sequence around 60 000 years ago. Biologists believe that such
a rapid accumulation of mutations is very unlikely in a population that
is more or less stable in number. So the most probable explanation is that
the population grew rapidly, possibly following the occupation of new territory.

Despite these results, the molecular approach to anthropology is likely
to remain extremely controversial. In previous studies Wilson and his collaborators
constructed an evolutionary tree for modern humans by analysing mitochondrial
DNA from individuals belonging to various Old World populations. In doing
so, they assumed that DNA mutates at a constant rate, and so makes a good
evolutionary clock. This assumption has been one of the main bones of contention
between critics and supporters of Eve.

The first challenge was to the rate of mutation that Wilson and his
colleagues claimed for the clock: about 3 per cent per million years. Such
a rate would mean that in any stretch of 100 bases in the mitochondrial
genome, about three bases would mutate in any million year period.

Critics, including Milford Wolpoff and his colleagues, first argued
that a figure of 1.5 per cent was more likely. This slower rate pushes the
‘DNA date’ for Eve back to around 430 000 years ago. Wolpoff has since conceded,
however, that Wilson’s laboratory got the rate ‘precisely right’.

The Wilson camp now faces a different central criticism: which is that
the maternal inheritance of mitochondrial DNA makes it a potentially misleading
marker of population history. Because lineages may be lost randomly, critics
say that there is every likelihood that all the older lineages of modern
humans have disappeared during history. Eve’s lineage may be the one we
see in mitochondrial DNA but it need not be the oldest.

The Wilson camp refutes this, arguing that because lineage loss is a
chance event – like the loss of surnames in a population – younger lineages
are just as likely to be lost as are the older ones; so at least some ancient
lineages should have survived. To date, however, no such lineages have been
discovered – even though researchers have tested mito-chondrial DNA from
some 4000 Asians.

Ah, say the critics, what if lineage loss was not random but was driven
by natural selection? Mitochondria with just one type of genetic make-up
might then be driven through the population, replacing existing types. This
is possible in theory, acknowledges John Avise, an expert on mitochondrial
dynamics in animal populations at Georgia State University. But he knows
of no instance of preferential replacement as a result of natural selection.

Mitochondrial DNA is, of course, just one source of genetic information
about human history. What of the testimony of nuclear genes? Inferring population
histories from nuclear genes is difficult because maternal and paternal
genes mix. Nevertheless, the results that are emerging support the Eve hypothesis.
Masatoshi Nei and colleagues in Pennsylvania and Luigi Luca Cavalli-Sforza
at Stanford have independently analysed several hundred variants of nuclear
genes, getting data consistent with the theory.