The disposable soma theory of the evolution of ageing proposed by Kirkwood (1977) and developed further here, in itself provides a strong argument in favour of the hypothesis that ageing is in large part due to the accumulation of defects in macromolecules. There is now compelling evidence that the accuracy of synthesis of these depends not only on the specificity of enzymes, but also on a series of additional mechanisms which remove, in one way or another, incorrect products. These mechanisms are themselves mediated by proteins, the structure of which are, of course, genetically determined. It is therefore impossible to dispute the conclusion that the overall accuracy of macromolecular synthesis is dependent on the genotype of the organism. This being so, it is clear that organisms will evolve to an optimum level of accuracy. If errors were too frequent, it would be impossible, a completely error-free organism, with no possibility of mutation, would no longer be capable of evolution. In the initial evolution of increasing accuracy the law of diminishing returns must apply and there will always be an optimum point where there is no further advantage in investing energy and cellular resources to reduce the error levels. Since organisms have very different environments and life styles we would certainly expect them to evolve to different optimum error levels, and our basic hypothesis is that these will be directly related to the time the species must survive in the natural environment in order to reproduce itself successfully. The theory described makes to important predictions, both of which can be tested experimentally. First, the authors would predict that germ-like cells synthesize macromolecules with greater accuracy than somatic cells. So far, no direct comparison of the accuracy of protein synthesis in these two types of cell have been made, but measurements of this type should soon be possible. The second basic prediction is that the accuracy of macromolecular synthesis in somatic cells is directly related to the lifespan of the species. This could also be tested at the level of protein, RNA or DNA synthesis, but so far no appropriate experiments have been done.
