Meneely: Advanced Genetic Analysis 
Topical updates
Figures from the bookMeneely: Advanced Genetic Analysis
Do essential genes in humans have essential functions in mice?
Chapter 4, December 2008
The mouse is the best and most widely-used model organism for understanding human genetics. Nearly all genes in one organism have an orthologous gene in the other, and the function of a gene in humans can often be understood by examining the phenotype of a mutant in the orthologous mouse gene. On the other hand, many gene families have more members in one species than in the other so that one-to-one correlations in function are not always possible (or, indeed, informative). A systematic comparison of mouse mutant phenotypes with human genetic diseases has not yet been done.
As a first step towards this goal, a study by Liao and Zhang (2008) asked if genes that have essential function in humans are also essential in mice. They began with 1450 genes that cause genetic diseases in humans and that have an easily identified mouse ortholog. Of these genes, 756 have been deleted in the mouse, so information on the null mutant phenotype is available. This number is further reduced by eliminating genes that cause mild diseases in humans or that may not be null mutations in humans (612 genes) and by excluding genes involved in immune function, which could be very different between a mouse living in an animal room and a human living in a pathogen-rich environment (24 genes). With these adjustments, the comparisons were based on mutations in 120 human genes that cause death before puberty or infertility.
Of this group, 93 genes are also essential in the mouse, suggesting that these genes are likely to have similar functions in both species. However, 27 genes are essential in humans but not in mice as evidenced by a deletion having no apparent mutant phenotype. Thus, the human disease is not being effectively modeled by the mouse mutant.
The authors tested and ruled out a number of possible explanations for the phenotypic difference between the two species. For example, the genes that are essential in humans but not in mice do not have duplicate mouse paralogs, so it is unlikely that the functions are being taken over by another gene. Likewise, the genes as a whole do not have notably different expression patterns in the two species, so changes in regulation do not appear to explain the difference in phenotype.
One difference that was noted is that the genes that are essential only in humans are slightly more diverged in inferred amino acid sequence from their mouse orthologs than those genes that are essential in both species. The greater divergence may not be too surprising since the non-essential mouse gene is predicted to have less selective pressure to maintain its sequence and function. However, the authors favor another interpretation for this greater sequence divergence, one in which there has been a selection for essentiality in humans rather than a loss of essentiality in mice. They make this inference by comparing the rate of amino acid replacement (the non-synonymous substitution rate) for these genes to the rate for the average human-mouse gene. The rate of substitution is greater for the essential human genes than for the average gene, suggesting that there has been selection for the sequence (and function) in the human.
They further support this inference by examining the orthologous genes in some other mammalian species; although the effect of a mutation is not known for most of these other mammals, the rate of substitution is more similar to the mouse than to the human. Thus, it appears that there has been selection for an essential function in humans for these genes.
Do these genes have any functions in common that could explain the selection pattern? The authors used Gene Ontology (discussed on pages 156 and 338 in the book) to ask if any particular cellular functions are over-represented among the genes that are essential in humans but not in mice. Of the 27 such genes, 12 of them have products associated with vacuole function. The vacuole is involved in waste and toxin accumulation and disposal. The authors speculate that these genes may have an essential function in humans because we have a much longer life span than a mouse so waste accumulation could have a more deleterious effect. This does not explain the other 15 genes (or 12.5% of the starting number) that are essential in humans but not in mice, so these may be genes to study individually.
In summary, most of the genes that are essential in humans are also essential in mice, re-enforcing the value of the mouse model for human genetic diseases. However, a surprisingly high number of genes that are essential in humans are not essential in mice. While some of this difference can be explained by differences in life span, other genes are not simply explained. This suggests that it may not always be easy to draw a correspondence between a human disease mutation and a mouse mutant phenotype.
Reference:
Liao, B-Y, and J. Zhang, 2008. Proc. Natl Acad. Sci. ( USA) 105: 6987-6992
View print version