Niimura Y and Nei M. 2005. Comparative evolutionary analysis of olfactory receptor gene clusters between humans and mice. Gene 346: 13-21.

On March 3, 2009, we discussed multiple topics regarding the production, efficacy, and reception of chemical signals. Despite being a modality less pertinent to human life and civilization, olfaction seems to be a dominant form of communication in my study species (harvestman). I also find the general lack of dependence on sense of smell in humans curious in comparison to the importance it has in other systems. Thus I was intrigued when we were presented in class with the relative number of olfactory receptor genes found in mice as opposed to humans—almost three times as many. Why would humans lose their sense of smell over evolutionary time? Have we been completely excommunicated from this gene family, or is there evidence of the presence of lost olfactory genes in humans? And how blind are humans really to intraspecific chemical communication?

To begin to answer some of the former questions, I consulted the Niimura and Nei paper on comparative analysis of olfactory receptor (OR) gene clusters in humans and mice. This paper backed up their previous work from 2003 on numbers of OR genes and pseudogenes (defunct genes with no protein coding ability) found in each species. Mice were found to have approximately 1400 OR genes with a fraction of about 25% percent of those pseudogenes. Humans by contrast had only 802 OR genes, but a whopping 52% were pseudogenes. These OR genes are found on every chromosome in the human genome, but tend to form clusters. So, could it be that the reason humans have so few olfactory receptors compared to other mammals is due to relaxed selection on human ancestors to exhibit strong olfactory sensitivity? How similar are human genes and pseudogenes to those of mice?

Niimura and Nei explored these questions by conducting phylogenetic analysis using the neighbor joining method on known OR functional and pseudogenes in banked human and mouse DNA sequences. They defined an OR genomic cluster as any situation where neighboring OR functional or pseudogenes are separated by 500 kbps or less. This inclusion uncovered a mouse genomic region dubbed Mm2.2, an area that seems to parallel the gene-rich human major histocompatibility complex (MHC), although Mm2.2 is way larger. The authors also defined orthologous gene pairs (genes that are related due to common ancestry) between mice and humans as those that were resolved in the phylogeny with at least 80% bootstrap resampling support. These definitions allowed the authors to break up the resultant phylogeny into gene clades supported by 90% or better bootstrap values. Each of these clades was given a letter name based on size and location in the phylogeny (for example: for clades A, B, and C, AA would be the largest sub-clade found within A).

Every OR gene clade the authors found included mouse and human genes, suggesting most mouse OR genes have a human equivalent (whether functional or not). These clades included functional and nonfunctional OR genes and the genes tended to form tandem arrays, although related genes could still be found in different clusters. However the relative size of each clade was specific to the species—some clades were mouse-heavy, and one sub-clade did not have any genes orthologous to human OR genes. Also, phylogenetic trees containing many of the identified pseudogenes had less bootstrap support. The authors suggested this was due to the fact that they expected many of the OR genes to have lost functionality due to deletion events over evolutionary time. In humans, it appears some of the pseudogenes (denoted H*) actually started to form arrays AFTER they lost functionality. All in all though, no cases in the paper indicated entire genomic clusters that were duplicated in the mouse genome but were not found in the humans. Niimura and Nei concluded that mice have generated more OR genes than humans due to gene duplication, forming the tandem arrays.

Likely for space reasons, this paper focused more on the contributions of putatively functional OR genes, but I thought it would be really interesting to examine in more detail the human pseudogenes and their orthologous, functional mouse genes. Perhaps this sort of study could give insight into ancestral mammals and the sensory perceptions that are adaptive at one scale and comparatively useless or even somehow genetically harmful to keep around for primates.