How experience regulates the birth of specific olfactory sensory neuron subtypes
A major focus of research within our lab is to elucidate how postnatal olfactory sensory neurogenesis is regulated, with the long-term objective that it may be manipulated to enhance human health. To achieve this, we believe that an important related goal is to understand why olfactory sensory neurogenesis occurs throughout life.
The birth and incorporation of new neurons occurs prominently throughout life in only a few regions of the nervous system, including the dentate gyrus of the hippocampus, the sub-ventricular zone of the lateral ventricle, and the olfactory epithelium. While postnatal neurogenesis in the hippocampus and lateral ventricle is known to play important adaptive roles by facilitating learning and memory, persistent neurogenesis in the olfactory epithelium has long been assumed to play only the reparative role of replacing neurons that have died.
This assumption is based largely on the widely accepted model of neurogenesis within the olfactory epithelium, according to which post-mitotic neural precursors each randomly ‘choose’ a single odorant receptor gene from among hundreds within the genome. This choice defines the mature neuron’s ‘subtype,’ so most mammals have hundreds of different olfactory sensory neuron subtypes, each defined by the expression of a different receptor. Due to the random nature of receptor choice, it has long been assumed that the birth of olfactory sensory neurons is also random with respect to subtype. However, recent studies from our lab have revealed that neurogenesis is not random with respect to subtype, but rather that the birthrates of a fraction of subtypes depend on odor stimulation. These findings came initially from experiments using unilateral naris occlusion, a procedure in which one nostril is closed to reduce stimulation on one side of the olfactory epithelium, and EdU-birthdating. From these experiments, we found that olfactory deprivation reduces the birthrates of some subtypes (e.g., Olfr827), but does not affect the birthrates of most others (e.g., Olfr1414)
Olfactory deprivation selectively reduces the birthrates of specific neuron subtypes. In the experiment shown, a section of olfactory epithelium from a unilaterally naris occluded and EdU-injected mouse were stained for EdU (red in image) and Olfr827 mRNA (green). Greater numbers of newborn Olfr827 neurons (arrows in image; left graph) were observed on the open side relative to the closed, indicating a higher rate of neurogenesis for this subtype in the presence of olfactory stimulation. By contrast, the birthrates of most neuron subtypes, such as those expressing Olfr1414 (right graph), were not significantly affected by olfactory deprivation.
As summarized in the schematic below, our findings indicate that the birthrates of a fraction of subtypes are selectively accelerated in the presence of olfactory stimulation compared to the absence.
Olfactory stimuli selectively accelerate the birthrates of a faction of neuron subtypes. Our initial findings are summarized in this schematic, which depicts a section of olfactory epithelium from a unilaterally naris occluded mouse. Progenitors are depicted at the bottom, immature neurons in the middle, and mature neurons at the top. The birthrates of a fraction of subtypes (e.g., OR B; orange) are selectively accelerated in the presence of olfactory stimulation (open, left) compared to the absence (closed, right).
These findings are surprising and intriguing to us because they appear to challenge the widely accepted model that neurogenesis is random with respect to subtype identity. As such, they present an exciting opportunity to elucidate new insights into how olfactory sensory neurogenesis is regulated. They also suggest the possibility that persistent olfactory neurogenesis may serve an unknown adaptive function in addition to a reparative one.
To investigate how (and why) olfactory stimuli can selectively accelerate the birthrates of specific neuron subtypes, we are addressing the following specific questions (referred to in the schematic above):