Q J Med 2003; 96: 545-546
© 2003 Association of Physicians
Biologic |
The molecular clock
Sea squirts (Ciona sp.), flies, ascidians and human beings all have vitamin D receptor and retinoid X receptor genes. However, thyroid hormone receptor, retinoic acid receptor, and peroxisome proliferator-activated receptor genes are found in sea squirts and Man, but not in flies, implying that these genes are an invention of the chordates. Similarly, sea squirts do not have oestrogen receptor genes, genes for androgen, glucocorticoid or mineralcorticoid receptors: these appear to be associated with the appearance of the vertebrates. Such steps may be of evolutionary significance, and their identification is part of a redefinition of taxonomy, much as new disease classifications (in oncology, say) depend on new genetic information.
A measure of the genetic distance between species (now often known with great accuracy) can be converted to a time estimate for their separation from a common ancestor by using a calibration factor applied to the number of genetic changes expected in a given time. Though this methodology may seem cumbersome, and to be of only theoretical value, it is worth looking at more closely; it has helped to show that the last common ancestor of the main pandemic strain of the Human Immunodeficiency Virus was around in the 1930s, and thus could not have arisen from the use of primate cell cultures in polio vaccines, as has been suggested.
Rates of molecular evolution are much less variable than rates of change of body form and function. The study of amino acid sequences in a given protein in different species shows that around one amino acid substitution per 1x109 years has occurred in haemoglobins. Of course, this rate of change per amino acid codon does not have to be divided by three to get the rate of change per nucleotide (any one of the three nucleotides could change), since many nuclotide changes are synonymous. The rates of amino acid replacement in different proteins indicate that this rate is inversely related to the specificity of individual amino acids for protein function. Histones have very precise physico-chemical constraints in their binding to DNA; any amino acid substitution will affect function dramatically. In contrast, the fibrinopeptides can function effectively despite many amino acid changes. In general, proteins with the fewest constraints that are dependant on amino acid composition evolve the fastest, a consideration also applicable to parts of a molecule. Regions of genes that are not translated evolve faster than translated parts, non-functional genes evolve faster than functional ones, and synonymous nucleotide substitutions are much more rapid than those that code for a different amino acid. One conclusion that has been drawn is that the smaller the effect caused by a nucleotide change, the more rapid that change is in evolutionary time.1
Another way of looking at this is to say that the greater the proportion of neutral sites in a protein (sites where changing the amino acid has no effect), the faster the rate of molecular evolution. Advantageous mutations are rare, deleterious mutations are rapidly eliminated by selection, and most changes have no effect on protein function. Thus the rate of accumulation of neutral mutations is influenced only by the mutation rate, and remains relatively constant. There is an apparent paradox in the fact that the less strongly a nucleotide or amino acid is selected for, the more rapidly evolutionary substitutions occur. This is resolved by the consideration that the smaller the effect of a mutation, the more likely it is to be beneficial. Since mutations occur randomly with regard to the functioning of the organism, they are usually harmful; a small change might result in an improvement, where a large one is almost certain to make things worse. The areas of a molecule that are critical to its function generally cannot be changed without producing something less efficient – rates of change in the A and B parts of the pro-insulin molecule are around 0.4x10-9/amino acid/year, but reach 2.4x10-9 in the discarded C fragment. Examination of fossil leaves, including a magnolia leaf from over 20 Mya (million years ago) has shown that the photosynthesis related gene rbcL has only 17 of 820 bp different in its modern version.
But of course there may be different drivers of the process of change. Kimura’s data2 shows that the remarkably constant rate of change in 
-haemoglobin over evolutionary time is in no sense comparable to the rate of change of form of the animal in which it is found. The shark has changed very little during the last 350x106 years, while the mammals have changed from a fish-like ancestor to a group with widely diverse morphology. Crow illustrates this point further with an imaginative comparison: human haemoglobins and ß show similar substitution rates. This gene duplicated, forming the present two, about 500 Mya; this is about 100 m years after the separation of the ancestral forms of Carp and Man. The divergence between human and ß chains is the same as that between haemoglobin in the two species, despite the differences between water and air as environments and the much more rapid change in form in the human lineage. As Crow puts it, 'haemoglobin and morphological evolution are marching to different drummers'.
In a massive study, Kumar and Hedges3 have established a time scale for estimating rates of molecular change and comparing them with morphology, based on an analysis of 658 nuclear genes from 207 vertebrate species. Their findings endorse many analyses of timings made from the fossil record (for the appearance of the amphibians, the separation of the mammals and birds from a lizard-like ancestor, the marsupial/placental split and so on), but they found that some mammalian divergences occurred in the Cretaceous (100 Mya) before the Cretaceous/Tertiary extinction of the dinosaurs.
Identifying significant amino acid substitutions that result in adaptive changes in proteins has proved difficult, despite the development of increasingly sophisticated techniques of examining protein structure/function relationships. Many proteins are phylogenetically ancient, and it is difficult to distinguish adaptive from near-neutral substitutions. Perhaps the best evidence for an adaptive change is found when sequence convergence has been found in homologous proteins that have independently acquired the same new functions in two different organisms. Stewart et al.4 have shown that lysozyme isolated from the cow stomach and from the leaf-eating langur monkey Presbytis entullus shows sequence convergence, a finding of great significance, since these two herbivorous mammals have independently evolved foregut fermentation in complex stomachs (as opposed to the hindgut fermentation of rabbits and horses). The two forms of lysozyme in these animals share an ability to function at low pH and a resistance to proteolysis by pepsin. This has developed as a result of evolution of the monkey stomach enzyme away from that of other monkeys and towards the cow sequence at five specific amino acid sites.
If you only count the number of differences between the DNA sequence for a particular gene or genes you will miss repeated changes; it is usually the case that a model of sequence evolution is needed to identify the true number of substitutions that have occurred. There are problems with this, since substitution rates vary; in mice they are up to three times faster than those in primates. This is probably due to the combination of different DNA repair mechanisms, generation times, metabolic rate and the size of the population considered. It is easy to fix, rather than eliminate, a deleterious mutation by chance events in a small or isolated population—an important cause of potentially confusing results. Many major changes in the form of animals may be dependent on duplications (of the Hox genes for example) for which different modelling attempts must be made.
Why does any of this matter? Well, within the next two years we will have complete genomes for Man, the chimpanzee, the mouse, the rat, the chicken, the zebrafish, the puffer fish, the sea squirt, the mosquito, the fruit fly and a number of nematodes. Comparative studies of these genomes enables the identification of a role for genes with previously unknown functions, and an estimation of their significance from their conservation over a range of separation times from 5 m to 70 m years. So far, comparative methods seem to be the simplest for defining phenotype/genotype correlation in areas of uncertainty. To return to today, evolutionary trees have shown that the cases of West Nile virus encephalitis in New York and New England are the first in the Western Hemisphere, helping to identify birds as the vector. The long-term concerns of evolutionary biologists have a useful immediate relevance.
References
1. Crow JF. Basic concepts in population, quantitative and evolutionary genetics. New York, W.H. Freeman, 1986.
2. Kimura M. The neutral theory of molecular evolution. Cambridge, Cambridge University Press, 1983.
3. Kumar S and Hedges SB. A molecular timescale for vertebrate evolution. Nature 1988; 392:917–20.
4. Stewart CB, Schilling JW, et al. Adaptive evolution in the stomach lysozymes of foregut fermenters. Nature 1987; 330:401–4.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||