Nature The number of harmful mutations that arise in each generation has been measured, and it is surprisingly high. This supports one theory of why evolution favours reproduction, but the consequences for human health are unclear.
The rate at which deleterious mutations occur in a genome is clearly a quantity of interest -- not least when the genome is our own. It becomes particularly important if, as some have argued1, the rate in some species is high, at several new mutations each generation. Yet the deleterious mutation rate has been notoriously difficult to measure, and no convincing estimates exist for any vertebrate. Eyre-Walker and Keightley give the first such estimates for ourselves, chimpanzees and gorillas.
Getting an estimate of the total mutation rate is relatively simple. Neutral mutations -- those which neither enhance nor impair the organism carrying them -- accumulate through the generations at a rate equal to the mutation rate. The mutation rate can be determined by the rate of change of presumed neutral regions: areas of the genome, such as introns and pseudogenes, which are not translated into proteins. For mammals, these rates, extrapolated to the whole genome, lead to enormous numbers, on the order of 100 new mutations per individual1. Of course these can't all be deleterious, but no one knows what the proportion is. This proportion could be determined in principle by comparing the slower rate of evolutionary change of the genome as a whole with the faster rate expected under neutral assumptions; but this involves statistical uncertainties and extensive sequencing.
Eyre-Walker and Keightley have made the analysis feasible by concentrating on protein-coding regions. They measured the amino-acid changes in 46 proteins in the human ancestral line after its divergence from the chimpanzee. Among 41,471 nucleotides, they found 143 nonsynonymous substitutions -- mutations where swapping one DNA base for another changes an amino acid, and therefore the final protein made by that gene. If these had evolved at the neutral rate, 231 would be expected. The difference, 88 (38%), is an estimate of the number of deleterious mutations that have been eliminated by natural selection and have therefore made no contribution to contemporary populations.
Translating these numbers into mutation rates gave a total rate of 4.2 mutations per person per generation, and a deleterious rate of 1.6. The rates for chimpanzees and gorillas were very similar, the deleterious rates being 1.7 and 1.2, respectively. The authors took 60,000 as the gene number and 25 years as the generation length. The number 1.6 is probably an underestimate, for various reasons. For instance, mutations outside the coding region are not counted and some of these regions -- such as those controlling gene expression -- are expected to be subject to natural selection. The gene number may also be an underestimate. If there have been mutations that increase fitness, they would also cause the number of deleterious mutations to be underestimated. A less conservative, and probably more realistic, estimate doubles the value, giving 3 new deleterious mutations per person per generation.
What's the significance? Every deleterious mutation must eventually be eliminated from the population by premature death or reduced relative reproductive success, a 'genetic death'. That implies three genetic deaths per person! Why aren't we extinct? If harmful mutations were eliminated independently, as in an asexual species, it has been estimated that this would lower population fitness to a fraction e-3, or 5%, of the mutation-free value, leading to the inevitable extinction of species with limited reproductive capacity. A way out is for mutations to be eliminated in bunches. This happens if selection operates such that individuals with the most mutations are preferentially eliminated, for example if harmful mutations interact. But such a process can only work in sexual species, where mutations are shuffled each generation by genetic recombination. The existence of a high deleterious mutation rate strengthens the argument that a major advantage of sex is that it is an efficient way to eliminate harmful mutations. It also raises again the possibility of fitness decline or even extinction in rare species from too many harmful mutations.
Presumably, we humans have profited in the past by reproduction's ability to reduce the effect of a high mutation rate on fitness. In the recent past the intensity of natural selection has been greatly reduced, especially where a high standard of living means that most infants reach reproductive age. From this it would seem that natural selection will weed out mutations more slowly than they accumulate. This effect may be accentuated by trends for males to start or continue reproducing later in life, because the sperm of older men contains more base-substitution mutations. In a time of rapid environmental improvement, how this genetic decline will affect our health can only be guessed at.
Eyre-Walker and Keightley noticed that the proportion of harmful mutations in the 46 genes in their study is greater in humans than in the equivalent genes of rodents. Their preferred explanation is that slightly deleterious mutations have become fixed in the population, by a process known as random genetic drift, during periods of human history when the breeding population size was low -- especially during genetic 'bottlenecks'. This would increase whatever effect the accumulated mutations are having on current human welfare. Are some of our headaches, stomach upsets, weak eyesight and other ailments the result of mutation accumulation? Probably, but in our present state of knowledge, we can only speculate.
JAMES F. CROW, Nature, 1999.
Última Atualização: 23/Set/99
