Tuesday, January 25, 2011

The genetics of olfaction -- is a rose a rose by any other nose?

Laura Spinney writes about the genetics of olfaction in The Independent this week, and about population variation in what we can smell.  The genetics of olfaction have been pretty well worked out; in fact, the 2004 Nobel Prize in Physiology or Medicine went to Linda Buck and Richard Axel for this work.  But there is much still left to understand. 

Genes for olfactory receptors (ORs) comprise the largest gene family in the mammalian genome -- at about 900 genes and pseudogenes in humans, they make up approximately 3% of our total genome. On average, we've each got about 400 working olfactory receptor genes, and the rest have accumulated enough mutations to no longer be functional, and are now pseudogenes.  This is not quite accurate, because we are diploid so that we actually have about 800 OR genes, 2 copies of the 400.  Of the hundreds of these genes, many are polymorphically pseudogenes, meaning their sequence works in some copies in the population, but has been mutated out of function in others.

In addition, the copies we each have of each of our functional OR are themselves highly variable in their odorant binding pockets (the part of the receptor that binds to odorant molecules that float by in the nose).
"When I give talks, I always say that everybody in this room smells the world with a different set of receptors, and therefore it smells different to everybody," says Andreas Keller, a geneticist working at the Rockefeller University in New York City. He also suspects that every individual has at least one odorant he or she cannot detect at all – one specific anosmia, or olfactory "blind spot", which is inherited along with his or her olfactory apparatus.
Each olfactory receptor responds to several odorants.  But with so many genes, how on earth could we actually distinguish what we're smelling?  The answer is remarkable and one of the most interesting unsolved problems in genetics.  Our hundreds of OR genes are located in many different clusters of a few to hundreds of adjacent duplicate OR genes.  And we have two copies of each.  But in each individual olfactory neuron, only one of these hundreds of genes is expressed!  The other 799 genes, all over the genome, are inactive.

From Genetics and the
Logic of Evolution
Weiss and Buchanan, 2004
Further, neurons expressing a given OR are in some way guided by that choice to collect in OR-specific locations in the olfactory bulb of the brain.  That's what makes it possible for the brain to tally what specific receptors have been triggered.  That is, the binding of an odorant to its receptor triggers its perception in the brain.  The wiring pattern is to some extent, at least, conserved among individuals, so in that sense we may share not only the ability to identify, say, 'lemon', but to experience it in somewhat similar ways.

Most odor perception involves a combination of signals from a number of different receptors.  And, some of those receptors may be non-functioning pseudogenes in some people, and if the subset of pseudogenes differs, the odor perceived will be different.
That genetic variability is reflected in behavioural variability, as Keller, with colleague Leslie Vosshall and others, recently demonstrated when they asked 500 people to rate 66 odours for intensity and pleasantness. The responses covered the full range from intense to weak, and from pleasant to unpleasant, with most falling in the moderate range – a classic bell curve in each case.
There is also variation in how intensely people can smell.  Some people are born with no ability to smell, while others are acutely aware of odors.  This may have nothing to do with odorant receptors, but instead how efficiently the odorant signal is transmitted to the brain.

One of us (Ken) spent a sabbatical in England working with Manolis Dermitzakis at the Sanger genome institute outside of Cambridge, trying to find genomic DNA sequence 'signals' that might help explain the extensive monoallelic and monogenic expression of only one OR per olfactory neuron.  Unfortunately, the search was frustratingly unfruitful. In fact, earlier we had done a study in the Pennsylvania Amish, collaborating with one of the experts cited in the Spinney story to see if, in this classic founder-effect population, we could work out the genetics of the ability to identify specific odorants, for which the ability to smell seemed, in other studies, to be polymorphic.  We couldn't find any pattern, because essentially everyone could smell all the test odorants.

The unusual expression pattern of this huge gene family, and the bookkeeping by which its detection of complex odorants allows us to identify our surroundings with a great deal of reliability (and, much more so in other species like dogs, who use the same system) is one of the remarkable, and incompletely understood, facts of genetic life.


Anonymous said...

Francesc here, for a bit of Catalan toilet humor and another bit of self-promotion...

The most famous example of OR pseudogenization may be that of the receptor of metanethiol, aka what-happens-when-you-go-to-the-toilet-after-eating-asparagus. Funny thing, the discoverers of such receptors were... 23andme, who just had to ask their customers.

Intriguingly, Elena Bosch's found that, at one OR, a presumably functional variant had spread in a mannner that was inconsistent with neutral evolution (Moreno-Estrada et al. 2008, PMID: 17981927), so what you can smell may help you have more babies.

Ken Weiss said...

Interesting! I didn't know these things, but what you say suggests that the pseudogenization is polymorphic, since some of us can detect the tell-tale asparagus aroma. Alternatively, might some people with the dead asparagus gene have combinations of variation in the other OR genes that enable detection?

I didn't know of the paper you cite but will take a look. Inferring selection is not easy, as a rule, but there is a long history of thought about the importance of olfaction in human reproductive success--especially to the extent that we don't make classical reproductive pheromones or have a functional vomeronasal pheromone receptor genes the way other mammals do.