In Vivo Imaging of Neural Activity in Unanesthetized Drosophila Adult Flies

We are trying to understand the molecular, cellular and circuit basis of natural memory forgetting, aiming to uncover how the brain actively erases or suppresses memories to maintain cognitive flexibility. Recent research has shown that forgetting is not merely a passive decay of memories, but rather a highly regulated active biological process that requires specific patterns of neuronal activity.

A major challenge is linking specific circuit manipulations to dynamic memory processes while integrating connectomic, genetic and behavioral data in a rigorous and interpretable way. We helped establish that forgetting is an active biologically regulated process. Our work identified specific dopaminergic neurons and molecular pathways required for normal forgetting in the Drosophila brain. Our protocol enables functional imaging in flies without anesthesia, preventing unwanted nonspecific effects by the anesthetics. We use this approach to investigate the neural correlates underlying memory formation and active memory forgetting.

[Narrator] To begin, use Dremel tools and a diamond saw blade to cut a 22-gauge hypodermic metal tubing to a length of approximately 10 centimeters. With a Dremel 420 cutoff wheel, buff both ends of the tubing to create a smooth and clean opening that can accommodate the proboscis of the fly. Wrap the cut tubing around a 15-milliliter centrifuge tube to form the desired curved shape. Then cut a 7-centimeter long piece of 12-gauge hypodermic metal tubing. Now use a razor blade to trim the end of a 2-microliter pipette tip to fit the 22-gauge metal tubing. Fit the 12-gauge tubing into the other end of the pipette tip. Next, mix a small amount of epoxy resin and hardener together. Apply the epoxy to the junctions where the small metal tubing meets the pipette tip and where the larger tubing connects to the other end. Allow the epoxy to fully cure overnight before connecting the assembly to a micro manipulator holder and adjusting the angle as necessary. To build a shock and odor delivery pipette, cut 1 milliliter off a 1 x 100 glass pipette at the 3 milliliter mark using a Dremel diamond tool. Then cut a small rectangular acrylic sheet measuring 24.5 millimeters x 8 millimeters with a thickness of 1/8th inch. Cut a copper shock grid to fit onto the rectangular acrylic piece. Solder two electrical wires to opposite ends of the copper grid. Now place the copper grid on the acrylic piece and bend it slightly to accommodate the fly's abdomen and legs. Use electrical tape to attach the copper grid to the acrylic piece. Then use a hot glue gun to attach the glass pipette to the shock grid, ensuring it is straight and centered. To build the recording chamber, take a glass microscope slide as the chamber base. Mix resin and epoxy glue together. Using the epoxy glue, attach neodymium magnets onto all four corners of a black acrylic chamber. Place an additional magnet on top of each glued magnet. Then glue the newly placed magnets to a glass slide using epoxy. Hold the assembly in place with paperclips while curing. Remove the 200-microliter pipette tip from the aspirator. Insert the aspirator into a vial containing the Drosophila and aspirate a single fly into the 1,000-microliter pipette tip. Replace the 200-microliter pipette tip back onto the aspirator. Then gently blow and flick the aspirator so the fly is immobilized headfirst at the top of the 200-microliter pipette tip. Next, place the dissection chamber onto the manipulator holder. Connect the vacuum to the fly holding tubing and adjust the flow rate to approximately 500 milliliters per minute. Now move the vacuum metal tubing to the center of the microscope's field of view. Gently aspirate the flies proboscis into the vacuum holder. Adjust the manipulator to align the fly's head with the chamber opening. Turn on the direct current power supply. Using platinum resistance wire, apply melted myristic acid to glue the eyes and thorax to the chamber. Once secured, disconnect the vacuum tubing. Remove the recording chamber from the vacuum connection using the manipulator and turn the chamber upside down. Then glue the proboscis from below using platinum resistance. When everything has been glued, turn off the direct current power supply. Then turn the chamber upright. Attach the chamber to the glass slide base. Cut a small piece of tape with scissors and place it in front and behind the fly's head. Rotate the chamber so the fly's head faces the experimenter at a 90 degree angle. With a dissecting needle, make vertical incisions along the sides of the eyes. Rotate the chamber horizontally. Then make a horizontal cut across the cuticle. Now add 100 microliters of saline to the top of the fly's head. Using sharp forceps, remove the cuticle window, then remove any remaining fat or trachea with the forceps. Place a prepared fly onto the microscope stage of a confocal microscope equipped with a laser and a water immersion objective. With a micro manipulator, adjust the position of the shock grid and odor pipette so the fly is correctly positioned on the shock grid. Use the course z adjustment knob to scan through the z-axis of the brain and locate the brain region of interest. Set the frame size to 512 x 512 pixels. Begin recording from the neuron of interest using a custom made or commercially available odor delivery system. Set the recording duration to 2 minutes. Initiate the training protocol using the odor delivery system 5 minutes after collecting pre-training responses. Then record post-training responses about 5-15 minutes after training. The calcium indicator GCaMP6f and the red fluorescent protein tdTomato were selectively expressed in the mushroom body output neuron with dendrites projecting into the gamma and alpha dash lobes of the mushroom body, and the neuron was visualized using the MB077C split-GAL4 driver line. Calcium responses in the mushroom body output neuron to 3-octanol were significantly reduced 5 minutes after reversive conditioning without anesthesia and remained suppressed at 15 minutes. In contrast, calcium responses to 4-methylcyclohexanol were significantly enhanced 5 minutes post-training and remained elevated at 15 minutes. Pseudocolor images demonstrated distinct fluorescence changes pre and post-training. In anesthetized flies, post-training calcium responses to CS+ were only partially reduced and responses to CS- were not significantly different from baseline. Quantitative analysis confirmed that CS+ response was significantly depressed post-training in anesthetized flies, but CS- responses remained statistically unchanged. Plasticity was significantly higher in non-anesthetized flies compared to anesthetized ones.