Many regions of the brain are organized in such a way that they form an ordered representation of our sensory experiences. Similar physical stimuli –nearby points of light or sounds of similar frequency– are typically represented by physically adjacent neurons. This organization allows many computations in the brain to be performed locally, which reduces the length of required connections as well as the energy needed for communication. A commonly-encountered wiring theme is the “center-surround” receptive field. This concept is best illustrated in the retina: a bipolar cell is excited by the activity of photoreceptors in a small “center” region of the retinal map, but inhibited by receptors in the more distant “surround” region.
In the olfactory part of the brain, the first organized mapping of odors is a collection of glomeruli, densely packed on the surface of the olfactory bulb. Each glomerulus receives input from just one receptor type. Mitral cells, the output neurons of the olfactory bulb, combine signals within this map. Given their anatomy, it was long thought that they should present center-surround receptive fields. Early functional studies also suggested this kind of organization.
However, here we present results challenging this view. We found that the receptive field of a typical mitral cell consists of only a handful of glomeruli that are spatially dispersed on the bulb. Thus the mitral cell uses just ~2% of its potential synaptic connections. Furthermore, these glomeruli represent a broad range of odor sensitivities. We propose that each mitral cell performs a specific computation within the map of odors that combines a small and diverse set of glomerular inputs. Just one synapse beyond the receptor neurons, these computations seem more sophisticated than in the visual or auditory system, perhaps because the olfactory pathway is fairly short: mitral cells send their signals directly to the cortex of the brain.
Read more in Neuron