So each olfactory neuron (ON) expresses only one of its 2000 olfactory receptor (OR) genes on its surface and hence can only react (bind to) a limited range of odorant molecules. The combination of ONs that can bind to an odor molecule is a combinatorial expression code for the brain: the combination of signals is remembered when it first occurs and recognized when sometime later the same smell is encountered.
But how does this 'map' from nose to the brain?
In ways not yet understood, the receptor molecule on the surface of each ON helps guides the axon from the nose through the base of the skull into the brain. There it meets a structure called the olfactory bulb. Neurons expressing the same OR may be located in different parts of the nerve, but their axons recognize each other, and bundle together, as their axons migrate to one or two specific locations in the olfactory bulb. These locations are called glomeruli. A glomerulus is a kind of neural knot of axons from all the ONs that express the same OR and hence that respond to the same odor molecule.
From there, neurons travel to various parts of the brain. Their routes are not precise and there is no longer a simple correspondence between the neural endings in a given part of the brain and the odor molecules they respond to. The brain can remember which ONs fire for a given odorant, but there's no spatial map that corresponds to the spatial pattern of the ONs in the nose. Instead, the brain simply remembers where 'lemon' is, and we each have this little factoid in different parts of our brains.
Now it was thought that evolution had programmed a fixed location for each glomerulus, so that in some sense there would be at least a map of where in the olfactory bulb axons from cells using the same OR would travel to. There is logic in terms of the development of brain terminals for sound and sight information, since for example, light travels from a tree, lion, potential mate in an orderly way. But there's no such natural order to odors, and glomeruli-like units are not found in other sensory systems. So why would such a system be needed and how did it evolve?
Recent work reviewed by Zou et al, the paper we referred to in our previous post, shows that it may be true that axons expressing the same OR molecule do recognize each other and bundle together. They end up in a glomerulus in the olfactory bulb. But it's not the same place in different individuals, even different genetically identical mice from the same inbred strain. The idea of forming a glomerulus, as a kind of developmental ordering that clusters similar neurons together, and may increase the signal strength as a result, seems to be programmed. But a particular location is not.
By requiring less specific order, and one forced on a system of information (odors) that didn't have any natural order, the olfactory 'wiring' system may thus have been easier to evolve. But it means that odor responses would be somewhat less stereotypical, less precisely evolved, and that we respond to odors more as a result of experience than hard-wiring.
In fact, Zou et al. report that some studies have shown that the wiring that's observed depends in important ways on olfactory experience--the usage of the neurons--during early life. Since each animal's experience is different, it's no surprise that the results also differ.
This picture does not make complex traits simple, but it helps show how complexity can be made simply. Tractable, reasonable processes can end up generating complex structures by assembling them bit by bit in stages during development. Olfactory organization is an example.
Further, this story shows that environments and their associated variation and stochasticity affect the traits we all bear--even when they are programmed genetically to develop by orderly processes. In that way, even inbred animals--like human identical twins--can be different from each other.