¿ìè¶ÌÊÓƵ

If the evolution of modern humans has left its mark in mitochondrial
DNA, what about earlier events? Researchers in the US have examined mitochondrial
DNA sequences for evidence of another evolutionary milestone: our divergence
from apes six million years ago.

It is now nearly a quarter of a century since researchers first used
genetic evidence to study how and when humans and apes evolved from a common
ancestor. For years researchers believed that humans, chimpanzees, and gorillas
diverged from each other at the same time, in a three-way split. Although
this was much less likely than a two-way split, the early genetic evidence
could neither prove nor disprove the idea.

Then, about five years ago, researchers used a technique known as ‘DNA
hybridisation’, which compares the genetic complements of two species without
analysing the sequences of individual genes. They uncovered no evidence
for a three-way split; gorillas evolved first, leaving humans and chimpanzees
to share a common ancestor for a while, before diverging in a two-way split.

Many anthropologists were disturbed by the notion that humans and chimpanzees
are more closely related to each other than either species is to gorillas:
the anatomies of the creatures seemed to argue against it. Nevertheless,
more and more genetic data accumulated supporting the human/chimpanzee association.

Yet there was still a problem. Some results, such as those from DNA
hybridisation and sequencing of ribosomal DNA, indicated that gorillas split
from the basic stock a long time before humans and chimpanzees separated
from each other. Not so various nuclear and mitochondrial genes, whose sequences
suggested that chimpanzees and humans shared their ancestory for just a
short spell.

Why were the sequences telling a different story? Perhaps a three-way
split was correct after all, and the genetic data an illusion. Maryellen
Ruvolo of Harvard University and her colleagues from various other laboratories
have gone a long way to settling these uncertainties (Proceedings of the
National Academy of Science, vol 88, p 1570). The biologists sequenced a
gene coding for a protein known as cytochrome oxidase subunit 11 from several
primates: the siamang (a relative of the gibbon), the gorilla, a chimpanzee,
and two Old World monkeys.

Ruvolo and her colleagues compared these results with data from the
equivalent human gene. They found strong support for the human/chimpanzee
association. Also, the interval between the origin of humans and chimpanzees
was about as long as that produced by any other technique. Assuming that
humans and chimpanzees separated six million years ago, gorillas might have
split as much as four million years earlier.

Ruvolo and her colleagues say there can be little doubt that gorillas
did split off first, followed later by a human/chimpanzee split. Exactly
how long humans and chimpanzees shared a common ancestor, however, is difficult
to say. The researchers suggest that it all depends on the difference between
what molecular evolutionists refer to as ‘species trees’ and ‘gene trees’.

When a common ancestor diverges to produce two sister species, a simple
Y-shaped species tree is produced. Suppose a molecular evolutionist now
compares the equivalent gene in the two sister species, and measures the
amount of sequence difference between them. The shape of the gene tree should
still be the same as the species tree, a simple Y shape. And, if the gene
in the common ancestor had existed in one form only just prior to the split,
then the accumulated mutations in genes in the sister species will reflect
the same time scale as in the species tree. The two trees will be the same
shape and have branches of the same length.

However, genes often split in the absence of speciation: so different
forms of the gene – polymorphisms – exist among the populations of a single
species. If an evolutionary event occurs later, producing two sister species
as in the earlier example, then a different picture emerges. The species
tree and the gene tree will again have the same shape, but because the genes
diverged before the new species evolved, the gene tree will have deeper
roots, and its branches will be too long.

Ruvolo and her colleagues believe that the shape of the human evolutionary
tree has now been settled, but as yet they see no easy way of determining
the length of the branches.