Motevalli, D. | bioRxiv Science

bioRxiv2025

Genetic and Circuit Mechanisms Underlying Natural Variation in Value-Based Decision Making

How natural regulatory genetic variation shapes innate economic decision biases by modifying neural circuit structure and function remains poorly understood. Here, we trace this pathway using a value-based oviposition decision in Drosophila. While laboratory flies reject sucrose in favor of a plain substrate, a wild-caught African strain accepts sucrose. This behavioral divergence maps to three African-specific intronic SNPs in the gene pumilio (pum), encoding an RNA-binding translational repressor. These SNPs downregulate pum, derepressing its target - the voltage-gated sodium channel paralytic (para) - in a pair of GABAergic interneurons that encode option values. Increased para enhances excitability, compresses neuronal value-coding differences between sucrose and plain options, and promotes sucrose acceptance. Selectively reducing pum or overexpressing para in these neurons converts laboratory flies to the African phenotype at physiological and behavioral levels. Our findings provide a genome-to-circuit-to-behavior model, illustrating how subtle regulatory polymorphisms reshape neural computations to drive adaptive variation in economic decision-making.

neuroscience↗

bioRxiv2024

Spatial learning in feature-impoverished environments in Drosophila

The ability to return to memorized goal locations is essential for animal survival. While it is well documented that animals use visual landmarks to locate goals1,2, how they navigate spatial learning tasks in environments lacking such landmarks remains poorly understood. Here, using a high-throughput spatial learning task we developed to investigate this question in Drosophila, we found that Drosophila can simultaneously use self-generated olfactory cues and self-motion cues to learn a spatial goal under visually challenging conditions. Specifically, flies mark a rewarded goal location with self-deposited scents, to which they assign a positive value, and use these scents and their self-motion cues to guide them back to the goal. This learning process is mediated by the mushroom body (MB) - an olfactory learning center responsible for associating odors with reinforcement3 - and by PFN neurons, which encode egocentric translational velocity4,5, a self-motion cue. Intriguingly, when the environment is enriched with prominent external olfactory landmarks, flies shift to prioritizing landmarks over self-generated cues - a strategy adjustment reflected in both the critical circuit involved and an altered transcriptome in the brain. Our findings demonstrate that Drosophila can dynamically adapt to environmental complexities when solving spatial learning tasks by creating and integrating internal and external cues, revealing an unexpected level of sophistication in their cognitive capacities.

neuroscience↗