ANIMAL MODELS are widely used in medical research, sometimes in testing new drugs for safety before human trials, other times as model systems for human diseases. Like all mammals, humans and mice share most of their genes, and maintain high sequence similarity. These factors suggest that many of these genes should share the same role. A new study in Proceedings of that National Academy of Sciences examines this hypothesis.
Many human diseases can be linked to a specific gene, and most of these genes have a mouse ortholog. Some of these mouse genes have been experimentally knocked out, and the null mutants studied for any effects of the absence of that gene. Liao and Zhang examined a set of 120 genes determined to be essential in humans—when the gene is rendered useless due to a null mutation fitness is zeroed out due to death before sexual maturity or sterility. The authors examined the records for mouse knock-outs for these genes, this time looking for genes that are not essential in mice in spite of their orthologs being necessary in humans. They found that 27 of these genes are not essential in mice. In these cases knock-out mice survived to maturity and were able to reproduce normally, at least until the age of 6 months. Additionally, in a few cases the knock-out mice showed no detrimental effects of the null mutation at all.
An examination of the human ortholog of these genes shows that many (44.4%) of the human essential/mouse nonessential genes localize to the cell vacuole. This organelle is involved in waste processing. Mice have a very rapid metabolic rate compared to humans, meaning that we might expect their vacuoles to have more stringent efficiency requirements. On the other hand, humans reproduce at a much later age than mice, meaning that the vacuole has to process more wastes before successful reproduction.
The authors additionally examined the nonsynonymous distance and the synonymous distance between the orthologs. This is a measurement of the frequency of nonsynonymous (affecting protein product sequence) or synonymous (redundant mutations not affecting protein product sequence) mutations. We would expect essential genes to have a low nonsynonymous distance because of purifying selection restricting divergence. Nonessential orthologs should be more free to mutate without causing detrimental effects. Indeed, the nonsynonymous distance between orthologs essential in both species (HeMe) was low. The nonsynonymous distance between genes essential in humans and their nonessential mouse orthologs (HeMn) was higher. Indeed, this was even higher than the average nonsynonymous distance between any human gene and its nonessential ortholog (HaMn).
The authors propose two factors affecting gene sequence divergence. First, the greater nonsynonymous distance between human essential genes and their nonessential mouse orthologs could be due to a slackening of purifying selection in the mouse orthologs. Alternatively, it could be due to positive selection for mutations in the genes essential in humans. The comparison of the HeMn set to HaMn reveals that this is the case. The HaMn set is composed of HeMn genes and HnMn genes. The genes nonessential in both humans and mice should be relatively free of purifying selection, so if the mere absence of purifying selection is the explanation for the nonsynonymous distance, the nonsynonymous distance for HeMn should be about equal to the nonsynonymous distance for HaMn. Since its nonsynonymous distance is actually greater, that must mean that natural selection was favoring certain nonsynonymous substitutions in human orthologs during evolution.
Indeed, a study of branch-specific nonsynonymous distance/synonymous distance ratios shows accelerated nonsynonymous substitution in primate evolution. For genes involved in vacuole function, this evolutionary route combined a general increase in body size with later sexual maturity. This would change functional restraints upon the vacuole, and favor certain beneficial mutations. Other genes essential in humans and nonessential in mice do not have as obvious a functional divergence, but probably experienced similar changes in function.
The gene orthologs that are essential in humans and not essential in mice are easily visible to us because of the dire diseases that result from mutation of these genes in humans. We can then explore the effects of knocking out these genes in mice. Obviously we cannot do the opposite. Currently we do not have the technology to selectively knock out specific genes in humans, and even if we could, knocking out genes in human embryos and then following the individuals to adulthood to determine their longevity and reproductive success would be unconscionable. Though we cannot study these mouse essential/human nonessential genes, they no doubt exist. Evolution is not targeted towards production of the human species, and other species have their own evolutionary specializations.
Gene orthologs carry out similar functions in most cases, and often the gene product from one species can substitute directly for its ortholog in a different species (as I mentioned in my discussion of genes in volvocine algae and their unicellular relatives). However, developmental and functional constraints will differ in different organisms, and a gene variant that is lethal in one species may be viable in another. Liao and Zhang point out that while knowledge of orthologs in other species can be illuminating, it should serve as a first approximation of gene function. Studies in mice will continue to provide us with important information with less expense and fewer ethical worries than other animal studies. However, in some cases we may need to ultimately move to primate disease models. This is probably most important for disorders of vacuole function, which appears to differ significantly between mice and humans.
Liao, B., Zhang, J. (2008). Null mutations in human and mouse orthologs frequently result in different phenotypes. Proceedings of the National Academy of Sciences, 105(19), 6987-6992. DOI: 10.1073/pnas.0800387105