Palaeontologists, the people with hammers who look for fossils, may be the latest group to feel the hot breath of the molecular biologists on their necks. Within a few years of the discovery of the structure of DNA in 1953, it was realised that the differing structure of proteins could be used to learn something about the relationship between organisms back into the mists of time. Forty years on, it seems the ambition is being realised.
Scientific journals are slowly filling up with little family trees relating different species with each other simply on the basis of the differences between proteins performing the same function in all of them.
Conceptually, it is quite simple. Protein molecules are all constructed from simple chemicals called amino acids strung together. Only 20 different units turn up in natural proteins, and the order in which they occur is genetically determined. So it is simply necessary to count the number of places along the length of a protein molecule at which the amino acids differ, and that will be a measure of their taxonomic distance from each other. The virtual identity of proteins with similar functions in human beings and the great apes is a constant reminder of our common origin.
Constructing family trees showing the divergence between species is the standard use of this technique, but there are difficulties. One protein (insulin, say) may give one tree, another (haemoglobin, say) another. That does not invalidate the technique, but makes people wary of relying on a single result.
The variations of protein structure arise from changes in the structure of the corresponding DNA, which happen at random. These are the "inheritable small variations" at the root of Darwinism. The molecular taxonomists hope that they can avoid discordant family trees by assembling enough data about different proteins to iron out the randomness of the variations.
Russell Doolittle, one of the pioneers of molecular taxonomy, has now gone a big step further by using the techniques to chart the course of the evolution of living things back for at least 2 billion years. Earlier this year, Doolittle and colleagues at the University of California at San Diego published their results. Perhaps inevitably, they have now run into trouble.
First, for stripling molecular taxonomy, the good news: it seems to work if there is enough data to play with. Doolittle chose no fewer than 57 proteins with distinctive functions from extant organisms spanning the range of biological complexity-from mammals and trees through fungi and protozoa to the two kinds of bacteria that still inhabit the earth.
One finding is that the bacteria which now give people disease (or which live in the soil) separated from all other creatures roughly 2 billion years ago. The next big step in the evolution of life was when the "others" split into what are now called archaebacteria (which live in oil wells or on the sea bed) about 200m years later and then the protozoa split off from the mainstream about 600m years after that. Doolittle has plants establishing their separate existence marginally earlier than the fungi-about 1 billion years ago.
The flak is not easily distinguishable from professional jealousy. In the current issue of Current Biology, Brian Golding from McMaster University in Ontario takes Doolittle to task for his choice of proteins. If, Golding says, Doolittle had chosen different proteins, he would have come to different conclusions; his estimates of age could well have been twice as great.
So who is right? Both. Doolittle explained in his original account that his 57 varieties were chosen so that he could be sure of them. Golding's were not excluded deliberately, but because their nomenclature is suspect. But the protagonists have demonstrated that molecular taxonomy has come of age.
It is a sign of near maturity in a new branch of science that people should be disagreeing about a factor of two in an estimated age; it is not so long ago that cosmologists estimated the age of the universe as 2 billion years-the same as Doolittle's age for the separation of the two classes of bacteria (and at least 5 times too small). But it is also of some importance that a person (Doolittle) and his helpers can sit down at a computer and calculate the time when the two main classes of bacteria separated from each other. Much will flow from this. The same technique, for example, has pinpointed the divergence of humans from great apes at 5-6m years ago.
u u u
the weeping over the failure of Ariane 5 and the loss of its four "Cluster" satellites prompts a question: why did Arianespace, the builder and launcher of the rocket, load the first flight of a new rocket with such a precious cargo?
Graduate students throughout Europe are alarmed that there will now be nothing of substance for them to put in their doctoral theses. It will be little comfort for them to know that the rocket will probably fly well at the next attempt and go on to be a big success at launching communications satellites cheaply. They should ask the bosses at Arianespace why they did not think of loading the first rocket with ballast. Or are scientific instruments considered just that?
