Search

Drosophila melanogaster can learn to associate rewards or punishments to odor cues. The goal of this project was to carefully characterize how learning alters fly behavior. We designed and set up a Closed-Loop Virtual Reality arena where the fly controls a rotating odor nozzle by walking on a floating ball which we used to test three experimental paradigms: i) behavioral responses to innately attractive and repulsive odors ii) responses to optogenetically stimulating attractive and repulsive signaling MB Output neurons and iii) behavioral responses to flies trained to form Pavlovian associations with odor. We find that flies responded to innately appetitive odor by tracking upwind and the results with aversive odor were not clear but seem to suggest that cross- or downwind avoidance may be a strategy. We also find that direct activation of MBONs elicits similar behavior, and suggest a search-like behavior may occur after stimulation offset. Finally, we find that Training flies with different fictive rewards may have different tendencies to enhance the upwind tracking of an odor.

Navigating the world requires an animal to make choices in a dynamic and uncertain world. Therefore, animals can benefit by adapting their behavior to past experiences, but the exact nature of the computations performed and their neural implementations are currently unclear. Extensive prior knowledge about fruit flies (D. melanogaster) provides a unique opportunity to explore the mechanistic basis of cognitive factors underlying decision-making. However, to do this, we require a large number of choice trajectories from single flies. We, therefore, expand and calibrate a Y-maze olfactory choice assay to run 16 flies in parallel to allow us to build and test better models using behavioral perturbation methods such as choice engineering. We take two complementary approaches to explore various learning rules that the fly may use - a model-fitting approach and a novel de-novo learning rule synthesis approach. Firstly, we fit increasingly complex reinforcement learning rules to explain behavior. We find that approximating perseverance/habits better explains and predicts individual choice outcomes. Next, we develop a flexible framework using small neural networks to infer learning rules and predict choices. We find that small neural networks with less than < 5 neurons trained to estimate odor values can accurately predict decisions across flies better than the best reinforcement learning models. We analyze the behavior of these networks to reveal underlying dynamics that reiterate the presence of perseverance behavior. We successfully reproduce most of our observations across different behavioral setups. Our results suggest that habit-forming tendencies beyond naive reward-seeking may influence flies’ choices.

The GENIE Project Team at HHMI Janelia Research Campus have developed a multiple genetically-encoded fluorescent indicators including the jGCaMP8 Calcium indicator family, the Voltron and Positron voltage indicators, and animal models incorporating these indicators. This poster provides updates on developments in each tool and how to obtain them. Chemigenetic voltage indicator improvement: Voltron2 showed a 65% increase in 1AP sensitivity and 3x better subthreshold membrane potential detection. Mutations in the voltage-sensing domain (Ace2N) were swapped between various types of voltage indicators and compared with patch-clamp. A new positive-going variant, Positron2, showed a 300% improvement in sensitivity. jGCaMP8s and 8m transgenic mice: TetO-jGCaMP8s mouse has ~5X increased SNR compared to TetO-GCaMP6s. Sensitivity and fast kinetics match AAV performance. TIGRE jGCaMP8s/m knock-in mice stably expressed in targeted neuronal subpopulations with reduced tetTA toxicity effects. Mice available soon from JAX.

Data and analysis scriptsfor the manuscript entitled"Idiosyncratic neural coding and neuromodulation of olfactory individuality in Drosophila".&nbsp

Drosophila melanogaster can learn to associate rewards or punishments to odor cues. The goal of this project was to carefully characterize how learning alters fly behavior. We designed and set up a Closed-Loop Virtual Reality arena where the fly controls a rotating odor nozzle by walking on a floating ball which we used to test three experimental paradigms: i) behavioral responses to innately attractive and repulsive odors ii) responses to optogenetically stimulating attractive and repulsive signaling MB Output neurons and iii) behavioral responses to flies trained to form Pavlovian associations with odor. We find that flies responded to innately appetitive odor by tracking upwind and the results with aversive odor were not clear but seem to suggest that cross- or downwind avoidance may be a strategy. We also find that direct activation of MBONs elicits similar behavior, and suggest a search-like behavior may occur after stimulation offset. Finally, we find that Training flies with different fictive rewards may have different tendencies to enhance the upwind tracking of an odor.

Recent progress in intracellular calcium sensors and other fluorophores has promoted the widespread adoption of functional optical imaging in the life sciences. Home-built multiphoton microscopes are easy to build, highly customizable, and cost effective. For many imaging applications a 3-axis motorized stage is critical, but commercially available motorization hardware (motorized translators, controller boxes, etc) are often very expensive. Furthermore, the firmware on commercial motor controllers cannot easily be altered and is not usually designed with a microscope stage in mind. Here we describe an open-source motorization solution that is simple to construct, yet far cheaper and more customizable than commercial offerings. The cost of the controller and motorization hardware are under $1000. Hardware costs are kept low by replacing linear actuators with high quality stepper motors. Electronics are assembled from commonly available hobby components, which are easy to work with. Here we describe assembly of the system and quantify the positioning accuracy of all three axes. We obtain positioning repeatability of the order of in X/Y and in Z. A hand-held control-pad allows the user to direct stage motion precisely over a wide range of speeds ( to ), rapidly store and return to different locations, and execute “jumps� of a fixed size. In addition, the system can be controlled from a PC serial port. Our “OpenStage� controller is sufficiently flexible that it could be used to drive other devices, such as micro-manipulators, with minimal modifications.