Physiological correlates of an enzyme polymorphism

ATTEMPTS to demonstrate the action of natural selection at enzyme loci have relied mainly on in vitro biochemical techniques to detect differences between allelic forms of the enzymes1–3. It is usually assumed that differences detected in this way will be relevant in vivo. The data on human inborn errors of metabolism do not support this view, however; in the majority of cases these are recessive. Heterozygotes, which usually have about 50% normal enzyme activity, rarely show clinical symptoms4. As any difference in activity between allozymes is likely to be much smaller, it cannot be assumed that such differences will necessarily be physiologically important. The properties of an enzyme molecule can only be biologically significant to the extent to which they affect the concentrations and flow of intermediates in its metabolic pathways. It was recently shown that changes in the concentration of most enzymes in a pathway at steady state will have little effect on flux5,6. Therefore, to demonstrate that differences in fitness between strains possessing different genotypes at an enzyme locus are caused by that gene substitution, it is necessary not only to show that differences in kinetic parameters can be detected in vitro, but also that changes in the activity of the enzyme will affect flux through that pathway. I previously described7 significant changes in genotype frequency at a phosphoglucomutase locus in a wild population of field mice (Apodemus sylvaticus) which occurred at a time of food shortage and declining population numbers. I report here the results of some in vivo experiments on glycogen metabolism in animals of different genotypes at this locus, inducing glycogenolysis by fasting. The results parallel survival data from the wild population and suggest that the genotype of an individual at this locus is of physiological significance and may therefore be subject to natural selection.