Synergistic olfactory processing for social plasticity in desert locusts

Introduction

Locusts were one of the Biblical plagues of Egypt—since the advent of human agriculture, fear of these insects and their voracious appetites has echoed throughout religion and storytelling. They have the ability to aggregate into dense clouds that destroy whole harvests. In spite of our ancient enmity with these insects, however, it is only recently that we’ve begun to understand how and why they form such imposing swarms.

Locusts (in this study, Schistocerca gregaria) come in two forms: a solitary and a gregarious form. Solitary locusts are bright green in color and do not aggregate. Gregarious locusts, by contrast, have a bold yellow-black appearance and do form large swarms. Locusts that are reared in high densities become gregarious, while those reared at low densities become solitary. Little is known about differences in sensory abilities between these two forms.

Petelski, Günzel and colleagues wanted to understand the differences between these two locust phenotypes and in particular the way in which they both process odors in the brain. To do so, they turned to calcium imaging, a technique that makes it possible to visualize neuronal activity by adding proteins that fluoresce in the presence of calcium (which flows into a neuron when it fires). Compared with fruit flies, calcium imaging in the locust smell-processing center, the antennal lobe, is especially tricky: in fruit flies, neurons that respond to a given odor converge on a single spherical bundle of synapses called a glomerulus. Fruit flies have 51 of these olfactory glomeruli (Bates et al., 2020). By contrast, the glomeruli in locusts are organized in a radial pattern made up of over 1,000 microglomeruli (Fig. 2C).

The authors first established that odor plays a role in the aggregation of gregarious locusts; then, they observed gregarious and solitary locust brain activity when exposed to various biologically relevant odors. Finally, they attempted to wrangle sense out of the 1,000 plus little blobs of activity that came out of their calcium imaging data.

Fig. 2: calcium imaging of the locust antennal lobe. A-Civ Images of the locust antennal lobe, with Ci-iv showing different Z-stacks of projection neurons in green and olfactory sensory neurons. D shows an overview of the authors’ analysis pipeline. E-Gii show calcium image traces from gregarious animals on the left and solitary animals on the right.

Highlighted results

To establish that olfaction plays a more prominent role in gregarious locust behavior, the authors first conducted a behavioral assay where a single locust in an arena could freely choose between going to a container with leaves, a container with locusts, or a container with both. Importantly, these containers could have holes to allow the scent to dissipate, be transparent to let the trial locusts see its compatriots, or both. The authors found that scent played an especially prominent role in the decision-making process of the gregarious locusts. Using a GC-MS analysis of the leaves coupled with calcium imaging, the authors concluded that leaf alcohol acetate is the primary odor driving locust attraction to leaves and produces similar calcium traces to whole leaf odor.

Petelski, Günzel and colleagues then used calcium dye backfills to image the olfactory  projection neurons, capturing activity in the dendrites and somata when exposed to leaf alcohol acetate odor, locust odor, or both odors together, resulting in few key conclusions:

1. Calcium imaging responses were consistent across trials within the same animal and across several animals. Individual response units (called granules in the paper) responded in a stable combinatorial way to the applied odors. Olfactory responses appear to function in a combinatorial code in locusts.
2. Gregarious, but not solitary animals, showed higher calcium responses to the food cues (leaf/leaf alcohol acetate) when the social odor cue (locust smell) was in the mix. Gregarious animals have a higher proportion of mixture-specific olfactory units than solitary animals. Gregarious locusts respond to the coincidence of both social and food odors.
3. The responses of projection neuron somata can be divided into different response motifs. An analysis of these motifs shows that there are more synergistic interactions in gregarious locusts. Projection neuron somata show a higher degree of response overlap and integration between social and food odor in gregarious locusts.
4. The authors created a model that has 92% accuracy in determining if a locust was solitary or gregarious based on their projection neuron somata response motifs. Projection neuron response motifs can predict the locust phenotype.

Taken together, these results suggest that changes in locust gregarious olfactory responses versus solitary locust olfactory responses are due to a subtle reweighting of existing neuronal pathways and dynamics. The interaction between social and food odor encoding in the antennal lobe could play a role in shaping collective foraging decisions in locusts. Teasing apart these subtle interactions in locusts and in other insects will no doubt be enriching for the broader field of sensory processing.

Why we like this study

Calcium imaging is a challenging technique: preparations require a lot of effort, limiting sample sizes and magnifying the impacts of noise on the data. We therefore appreciated the authors’ attempts to wrangle meaning out of a tricky dataset, further complicated by the biological structure of the locust antennal lobe. We especially liked seeing the inclusion of an insect with a different antennal lobe Bauplan than many of the more commonly imaged insects.

 

doi: https://doi.org/10.1242/prelights.37050

Read preprint (No Ratings Yet)