Experience-dependent plasticity of a highly specific olfactory circuit in Drosophila melanogaster

Introduction (how flies get used to the smell of death)

There are certain smells that set off alarm bells because they signal potential dangers. For human beings, one such smell is that of dead, putrefying carcasses. It is hard to ignore, and it just makes you want to escape. For fruit flies, an equivalent scent is the odorant geosmin. This chemical emanates from toxic molds, proving lethal to both flies and their larvae—so, when flies catch a whiff of it, they know it is time to make a quick escape.

Given the significance of this odorant to flies, extensive research has focused on the neural circuit governing its detection in Drosophila melanogaster: olfactory sensory neurons of a single type detect this chemical via the narrowly tuned odorant receptor Or56a. All Or56a-expressing olfactory sensory neurons converge in a singular spherical sub-region, known as a glomerulus, within the antennal lobe—the brain’s odor-processing region.

It was initially thought that the antennal lobe and the glomeruli in it were highly stereotyped and not altered by experience, in part because even drastic interventions such as removing all the neural inputs to this region did not result in equally drastic changes. Since then, however, numerous studies have shown that the fly’s experience can alter the volume of different glomeruli, suggesting that either the number of neurons or the number of synaptic connections they make with each other could change. Fabian and colleagues were therefore interested in understanding the effects of chronic geosmin exposure on the Or56a glomerulus, also known as the DA2 glomerulus (Fig. 1). For an odorant so critical to the fly’s survival, is it possible that their neural circuit is “soft-wired” instead of hard-wired—that is, might this glomerulus also change its volume or show other effects from odorant exposure?

Fig. 1. Example image of the geosmin-sensing glomerulus (DA2) in the antennal lobe (AL) and projection neurons going to higher brain centers, the mushroom body (MB) and the lateral horn (LH). Images were created with D. melanogaster lines that express photoactivatable GFP in the projection neurons.

Highlighted results

First, the researchers confirmed that prolonged exposure to geosmin does indeed change the volume of the Or56a glomerulus. Adult flies were exposed to one-minute-long pulses of geosmin odor every five minutes over a span of four days—a considerable exposure and amount of time for fruit flies. These flies were expressing photoactivatable GFP, which made it possible to induce green fluorescence in specific neurons in the brain. The authors thus were able to show that the Or56a glomerulus gets larger in flies exposed to geosmin relative to control flies exposed to mineral oil.

The authors found the following circuit-level changes leading to this increase in volume:

1. Projection neurons: the number of neurons going from the Or56a glomerulus to higher brain centers remained the same. These projection neurons did not show any detectable differences in their structure in higher brain centers, e.g., the mushroom body or the lateral horn. The number of axon terminals, boutons, and axon length all did not show any changes in the Or56a glomerulus projection neurons in higher brain centers. However, it appeared that projection neurons did expand their dendritic fields in the glomerulus itself within the antennal lobe.
2. Local interneurons: the neurons that connect the Or56a glomerulus to other glomeruli in the antennal lobe showed an increase in volume also, but only for one out of the four local interneuron sub-populations examined. The increase observed in this line was largely due to an increase in local interneuron boutons, i.e., swellings along the neuronal branches with high synaptic densities. Other subpopulations showed varying effects. These results demonstrate that long term geosmin exposure had different effects on specific local interneuron subpopulations at the synaptic and morphological level.
3. Olfactory sensory neurons and glial cells: The authors did not see an increase in olfactory sensory neuron or glial cell volume. However, the authors suspected that heightened activation of the olfactory sensory neurons might lead to mitochondrial changes in those neurons. Using a fly line with GFP tagged to mitochondria, the authors found evidence that, in olfactory sensory neurons, the number of mitochondria increases while their overall volume remains the same, suggesting higher levels of mitochondrial fission.

The authors also looked at other glomeruli that do not respond to geosmin and found that they did not change in the same way that the Or56a glomerulus did. Thus, only circuits that are activated by the exposure were modulated. Taken together with prior findings in the literature, these results suggest that chronic exposure to an odorant can cause that odorant’s glomerulus to get larger, or it could sometimes have the opposite effect. There seem to be general patterns and similarities in plasticity effects, yet no firm rules seem to exist (as far as we can tell right now).

Importantly, Fabian and colleagues demonstrated that the circuit changes they observed also translate to behavioral changes: on average, flies exposed long-term to geosmin are not as repelled by the odor, going so far as to tolerate it for feeding and oviposition—even though calcium imaging data suggests that these flies can still detect geosmin as well as control flies.

Why I liked this study

Given that the insect brain is often viewed as stereotyped due to early studies of the antennal lobe, it is nice to see a more recent antennal lobe study showing that there is greater plasticity and complexity in this brain region than previously imagined. This paper adds to a growing corpus of plasticity work. Although no firm rules regarding mechanisms for neural plasticity have been established yet, this type of study provides a necessary foundation for eventually developing them. It is also fascinating to see D. melanogaster flies acclimate to an odor that should be a major repellent. This gives me hope that I may yet get used to even the nastiest smells that I frequently encounter.

Tags: behavior, chemical ecology, fruit fly, green fluorescent protein, neural imaging, neuroanatomy, neuroethology, neuron

doi: https://doi.org/10.1242/prelights.36557

Read preprint (1 votes)