The way that the sense of smell works is that olfactory sensory neurons (OSNs) are wired up to olfactory receptors (ORs) in the nasal epithelium, from which they send signals to the brain. Once arrived there, a hierarchy of processing results in us experiencing the sensation of ‘smelling’. Exactly how the olfactory receptor-to-brain mapping works during development, and whether its physical pattern in the nasal epithelium is replicated in the brain, remained major questions until now. In a study published in Cell by [David H. Brann] and others, many of these questions have now been answered, at least for mice.
As it turns out, the mapping between OSNs and ORs isn’t performed by a random selection process, but instead creates a receptor map that’s closely matched between the nasal epithelium and the brain. What has complicated answering this question up till now is that the nasal epithelium isn’t a flat surface, but a convoluted labyrinth that maximizes surface area to smell better.
The second issue was linking the physical location of OSNs and gene expression in the nasal epithelium. Using a new approach, the researchers showed an intricate patterning in this epithelium, with the basal stem cells from which it regenerates maintaining this patterning. This makes for a system very similar to, for example, the auditory system, where the detection of frequencies in the inner ear, as a linear system, is found to be replicated in the brain.
Although it does not provide us with all the answers yet about how this genetic patterning works, it offers a glimpse at a fascinating system that would seem to be used repeatedly across sensory systems. It may also provide potential treatments for medical conditions affecting the olfactory system, whereby the sense of smell is missing, reduced, or oddly miswired, for example, after a SARS-CoV-2 infection of the olfactory nerve that leads to symptoms such as a constant sensation of a burning smell.
You have to wonder if a better understanding of the nose will revive interest in digitally creating and sending smells?
Having read the paper abstract, here’s what I think is going on.
The first part of the visual cortex, V1 in the back of the brain, is mostly a 1-to-1 mapping with the retina of the eye. If you present a figure to the eye, such as a capital “A”, the activated neurons in the V1 will be in a similar pattern.
Generally speaking, visual perception relies a lot on the adjacency of neighboring pixels. V1 recognizes lines and corners, higher levels recognize configurations of lines and corners as forming simple shapes, and even higher levels recognize configurations of shapes as objects.
It’s not at all as simple as the previous paragraphs might suggest – the optic nerve passes through several processing nodes, the individual nerves can mix randomly while the nerve grows towards the cortex (during development), and some basic processing is done in the retina – but if you map the activated neurons in V1 you can reconstruct what the subject is looking at.
The auditory cortex A1 is also topologically mapped to the sense of hearing. Low frequency tones are mapped to one end, high frequencies map to the other end.
It was an open question of whether the sense of smell is topologically mapped to the brain, or whether the connections are randomly made (or some other pattern). The cited study shows that the smell connections are not random, it’s topologically mapped, so that “patterns” of smell, if there is such a thing, would be available to the brain.
The paper also identifies the cellular mapping mechanism that enforces the mapping (during development). Since there’s a mechanism enforcing the mapping, it’s probably a survival trait and leads me to wonder what sorts of “patterns” of smell can be detected.
(Note: AI researcher, not a biochemist, explanation might have inaccuracies. But at least I’m trying to add to the discussion.)
Interesting. What i know is, that visual memories are also represented as the same shape in neurons. Maybe smells too?