Experience-Dependent Remodeling of Juvenile Brain Olfactory Sensory Neuron Synaptic Connectivity in an Early-Life Critical Period

Abstract

Early-life olfactory sensory experience induces dramatic synaptic glomeruli remodeling in the Drosophila juvenile brain, which is experientially dose-dependent, temporally restricted, and transiently reversible only in a short, well-defined critical period. The directionality of brain circuit synaptic connectivity remodeling is determined by the specific odorant acting on the respondent receptor class of olfactory sensory neurons. In general, each neuron class expresses only a single odorant receptor and innervates a single olfactory synaptic glomerulus. In the Drosophila genetic model, the full array of olfactory glomeruli has been precisely mapped by odorant responsiveness and behavioral output. Ethyl butyrate (EB) odorant activates Or42a receptor neurons innervating the VM7 glomerulus. During the early-life critical period, EB experience drives dose-dependent synapse elimination in the Or42a olfactory sensory neurons. Timed periods of dosed EB odorant exposure allow investigation of experience-dependent circuit connectivity pruning in juvenile brain. Confocal microscopy imaging of antennal lobe synaptic glomeruli is done with Or42a receptor-driven transgenic markers that provide quantification of synapse number and innervation volume. The sophisticated Drosophila genetic toolkit enables the systematic dissection of the cellular and molecular mechanisms mediating brain circuit remodeling.

Introduction

The remodeling of juvenile brain circuits during early life represents the last chance for large-scale synaptic connectivity changes to match the highly variable, unpredictable environment into which an animal is born. As the most abundant group of animals, insects share this evolutionarily conserved, foundational critical period remodeling mechanism¹. Critical periods open with the onset of sensory input, exhibit reversible circuit changes to optimize connectivity, and then close when stabilization forces resist further remodeling². Insects are particularly reliant on olfactory sensory information and show a well-defined olfactory critical period. Drosophila provides an excellent genetic model to investigate this experience-dependent critical period in the juvenile brain. Odorant experience during the first few days following eclosion drives striking circuit connectivity changes in individually identified synaptic glomeruli^3,4. The direction of remodeling is dependent on the specific input odorant experience. Some odorants cause an increase in the synaptic glomerulus volume for a couple of days post-eclosion (dpe)^3,5,6,7, whereas other odorants cause a rapid elimination of synapses during the 0-2 dpe critical period, resulting in decreased innervation volume^8,9,10. Specifically, ethyl butyrate (EB) odorant experience drives dose-dependent synaptic pruning of the Or42a olfactory receptor neurons only during this early-life critical period⁸. The synapse elimination is completely reversible by modulating EB odorant input within the critical period but becomes permanent following the closure of the critical period. This olfactory experience-dependent synaptic pruning provides a valuable experimental system to elucidate the temporally restricted mechanisms underlying juvenile brain circuit remodeling.

Here, we present a detailed protocol used to induce and analyze EB experience-dependent synaptic pruning of Or42a receptor olfactory sensory neurons during the early-life critical period. We show that Or42a synaptic terminals in the antennal lobe VM7 glomerulus can be specifically labeled by transgenically driving a membrane-tethered mCD8::GFP marker, either directly fused to the Or42a promoter (Or42a-mCD8::GFP)¹¹ or using the Gal4/UAS binary expression system (Or42a-Gal4 driving UAS-mCD8::GFP)¹². Individual Or42a neuron synapses can be similarly labeled using targeted transgenic expression of presynaptic active zone markers fused to an array of fluorescent tags (e.g., Bruchpilot::RFP)⁸ or an electron-dense signal for ultrastructural synapse analyses (e.g., miniSOG-mCherry)⁸. Or42a synaptic terminals can be imaged with a combination of laser-scanning confocal microscopy and transmission electron microscopy. We show that Or42a synaptic glomeruli pruning is EB dose-dependent, scaling to the concentration of the timed odorant experience. The percentage of EB odorant dissolved in mineral oil used as a vehicle can be varied, as can the timed duration of the odorant exposure in developmentally staged animals. Finally, we show the methods used to analyze the extent of synaptic glomeruli pruning by measuring the VM7 innervation fluorescence intensity and volume. Synapse number can also be quantified by counting labeled synaptic puncta and by measuring synaptic ultrastructure parameters using transmission electron microscopy⁸. Overall, the protocol shown here is a powerful approach that enables the systematic dissection of both cellular and molecular mechanisms mediating Drosophila olfactory circuit synaptic connectivity pruning during a juvenile critical period. The general odor exposure setup described in this study has been utilized in previous studies using other odors and assaying other glomeruli^3,7.

Protocol

1. Odorant exposure

1. Using a fine paintbrush, sort 40-50 developmentally-staged animals as pharate dark pupae (90+ h post-pupariation at 25 °C) into 25 mm x 95 mm polystyrene Drosophila vials containing standard cornmeal molasses food (Figure 1A).
2. Place fine stainless-steel wire mesh over the end of the Drosophila vials to contain the flies while also allowing good airflow. Secure the wire mesh caps with taped transparent film onto the side of the Drosophila vials.
3. In a 1.5 mL microcentrifuge tube, place 1 mL of 100% mineral oil (vehicle control) or dissolve the ethyl butyrate (EB) odorant in mineral oil in another tube to produce the desired concentration (e.g., 15% (150 µL) or 25% (250 µL)).
4. Place vehicle control or odorant microcentrifuge tubes upright within an air-tight 3700 mL Glasslock container. Using tape, anchor the tubes securely in the center of odorant chambers together with the Drosophila vials (Figure 1B).
5. Place the sealed vehicle control and EB odorant chambers containing the staged Drosophila pharate pupae vials into a humidified (70%), temperature-controlled (25 °C) incubator on a 12 h/12 h light/dark cycle.
6. Remove the newly-eclosed flies after 4 h of odorant chamber exposure and rapidly transfer 20-25 flies to fresh Drosophila vials in chambers with freshly made odorant (Figure 1A,B). Discard the uneclosed pupae.
7. Keep the flies in their sealed odorant chambers in the incubators for a total of 24 h. For longer odorant dosing, rapidly transfer flies to new Drosophila vials in freshly made odorant chambers every 24 h.
8. Anesthetize the developmentally-staged flies by submerging them in a dish of 70% ethanol for 1 min in preparation for immediate brain dissection and immunocytochemistry processing.

2. Brain dissection

1. Mark vials as vehicle control (100% mineral oil only) or EB exposed (% EB) by labeling and maintain these designations for the entire duration of brain dissection and subsequent processing. Process 10-20 animals for each genotype/odorant condition.
2. Using forceps, transfer a single anesthetized fly into a small dish of freshly made phosphate-buffered saline (PBS)¹³. Immerse the fly in PBS and continue full brain dissection with the fly fully submerged.
3. With fine (#5) sharpened forceps in both hands, place the fly ventral side up and grasp the upper thorax with one forceps and the head under the proboscis with the other forceps. Take care not to penetrate the brain.
4. Remove the head from the rest of the body by pulling gently in opposite directions with both hands. The head should easily detach from the thorax; leave the isolated head for dissection (Figure 1C).
5. Slide the forceps previously used to grasp the thorax under the opposite side of the proboscis. Begin to gently pull the exoskeleton cuticle in opposite directions so that it tears between the eyes to reveal the brain.
6. When pulling, the brain optic lobes may separate along with the exoskeleton. To prevent this, use a slow, steady pull with the two forceps while removing the exoskeleton cuticle (Figure 1C, middle).
7. Continue to remove the exoskeleton cuticle from the head. Ensure that all parts of the exoskeleton are removed. Remove any tissues attached to the brain, including the fat body and any projecting trachea.
8. To prevent sticking, rinse a P20 pipette tip with PBS + 0.2% Triton-X 100 (PBST). Immediately after dissection, transfer the brains into the fixing solution (4% paraformaldehyde (PFA) + 4% sucrose) in a capped tube (Figure 1C,D).

3. Brain immunocytochemistry

1. Fix the brain for 30 min at room temperature (RT) with end-over-end rotation. While keeping the brains in the tube, pipette off and properly dispose of the hazardous fixative. Quickly wash fixed brains 3x with PBS (Figure 1D).
2. Brains can often be found stuck in the cap of the tubes after rotation. To prevent accidental loss of brains, which is a constant hazard in all transfers, use a dissection microscope while pipetting liquids.
3. In preparation for antibody labeling, block the fixed brains for 1 h at RT in PBST + 1% bovine serum albumin (BSA) + 0.5% normal goat serum (NGS) with constant end-over-end rotation (Figure 1D).
4. Remove the block and incubate brains with the selected primary antibody diluted appropriately in PBST + 0.2% BSA + 0.1% NGS. Incubate the brains overnight (14-16 h) at 4 °C with constant rotation.
   NOTE: Example antibodies used: rat anti-CadN (dilution 1:50)¹⁴ to label synaptic glomeruli and chicken anti-GFP (1:1000)⁸ to label Or42a-mCD8::GFP (Figure 1D).
5. Pipette off the primary antibody. Wash brains 3x for 20 min each with PBST. Incubate brains with secondary antibodies diluted in PBST with 0.2% BSA + 0.1% NGS for 2 h at RT with constant rotation.
   NOTE: Alternatively, incubate brains overnight (14-16 h) at 4 °C. Dilute antibodies in PBST with 0.2% BSA + 0.1% NGS; secondary antibodies used here: 488 goat anti-chicken (1:250) and 546 goat anti-rat (1:250).
6. Pipette off the secondary antibodies. Wash brains 3x for 20 min each with PBST. Finish by washing brains for 20 min with PBS (Figure 1D).
7. Prepare glass microscope slides (75 mm x 25 mm) by adding two thin strips of double-sided adhesive tape to the slide ~25 mm apart from one another. This provides a spacer to avoid crushing the brains.
8. Pipette ~10-15 µL of mounting medium between the two strips of tape to mount the brains. With a P20 pipette tip pre-rinsed with PBST, transfer the labeled brains onto the microscope slide (Figure 1E).
9. Once the brains have been transferred, use a fine paintbrush to align them, ensuring the antennal lobes are facing upwards. Look for the side that has the arched hump, which will contain the antennal lobes.
10. Once the brains are properly oriented, cover them with a glass coverslip (No. 1.5H), ensuring the coverslip is secure on the tape. Then, fill in the sides of the coverslip with additional mounting medium (Figure 1E).
11. Seal the edges of the coverslip with clear nail polish. Allow the slide to dry thoroughly, and then store it in the refrigerator for subsequent imaging.

4. Confocal imaging

1. Blind all brain slides to both genotype and experience conditions prior to imaging by marking the slide with a coded label for later decoding.
2. Use a laser-scanning confocal microscope with a 63x oil-immersion objective. We use a Zeiss 510 META microscope in all images shown here (Figure 1F).
3. Use appropriate laser lines for the fluorophores employed. Here, use an Argon 488 and HeNe 543 laser for the antennal lobe synaptic glomeruli and Or42a olfactory sensory neuron imaging (Figure 2).
4. Determine the optimal gain and offset for both channels, ideally keeping the gain for both channels <750 and offset ≈0. This is done to ensure the signal-to-noise is optimal while limiting the background.
5. Position imaging to the center of the brain. When imaging, the VM7 glomeruli reside proximally to the hole left in the middle of the brain following removal of the esophagus during dissection (Figure 2A).
   NOTE: The VM7 glomeruli will always be close to the opening of the esophagus. However, the exact location and size vary slightly based on brain dissection and mounting differences, so take this into consideration when imaging.
6. Select the imaging resolution and optical slice thickness. Here, 1024 x 1024 resolution with a Z-stack slice thickness of 0.37 µm is used.
7. Take an entire confocal Z-stack projection through the antennal lobe, ensuring the capture of the full Or42a neuron innervation of the VM7 glomeruli.

5. Synaptic measurements

1. Load the genotype/condition-blinded image into Fiji (1.54f). Split the laser line channels by clicking Image > Color > Split Channels.
2. In the Or42a olfactory sensory channel, determine which slices contain the Or42a innervation by scrolling through the entirety of the Z-stack, identifying where the fluorescence begins and ends.
3. Create a sum slices projection including only slices containing Or42a neuron innervation by clicking Image > Stacks > Z Project > Sum Slices and entering the desired range.
4. In Fiji, click the Lasso tool on the top bar to trace the outline of the Or42a neuron innervation in the VM7 glomerulus, treating each brain side glomerulus independently (Figure 3).
5. Multiply the circumference by the number of Z-stack slices and the thickness of each slice to obtain the VM7 synaptic glomerulus innervation volume. To quantify synapse number, the find maxima tool can be used along with the find maxima stacks macro to count synapse puncta in the outlined region^8,15.

Representative Results

Figure 1 shows the workflow for the olfactory experience-dependent critical period odorant exposure and brain imaging methods. The protocol starts with the age-matching of pharate dark pupae immediately prior to eclosion (Figure 1A). The pupae are placed into odorant chambers for 4 h, and then newly-eclosed adults are flipped into fresh vials in either the vehicle control or dosed EB odorant chambers (Figure 1B). We typically expose to the odorant for 24 h, although any duration is possible. After odorant exposure, the brain dissection begins by removing the head from the thorax and then removing the cuticle exoskeleton with fine forceps (Figure 1C). The isolated brain is cleaned of any associated tissues that could interfere with imaging and then immediately fixed in 4% paraformaldehyde and 4% sucrose in PBS. A short fix of 30 min is sufficient for good integrity of the brain tissue. The immunocytochemical protocol begins with a 1 h block (e.g., BSA, NGS), overnight incubation in primary antibodies (e.g., anti-GFP, -CadN), and then a brief 2 h incubation in secondary antibodies (Figure 1D). The labeled brains are mounted on glass slides under sealed coverslips with the brain dorsal surface up for best imaging of the antennal lobes (Figure 1E). Note that all the samples are blinded to both genotype and experience conditions by marking the slide with a coded designation for later decoding. A laser-scanning confocal microscope with a 63x oil-immersion objective is used to image the central brain antennal lobes (Figure 1F). Laser lines optimal for the fluorophores used are selected (e.g., Argon 488 and HeNe 543) to image the Or42a olfactory sensory neuron innervation of VM7 in the context of surrounding glomeruli.

It is important to select the optimal gain, offset, imaging resolution, and optical slice thickness to capture the entire Z-stack projection of the Or42a neuron innervation in the VM7 glomeruli. All genotypes and experience conditions to be compared are imaged with identical microscope settings. The Z-stacks are processed using ImageJ software with experience-dependent changes assessed primarily by innervation volume^7,14. There are two ways in which volumetric measurements have been computed. One calculates the innervation area slice-by-slice using the ImageJ lasso function and then multiplies each value by the thickness of each slice (µm)^8,9. The other uses sum slice Z-projections and multiplies the number of Z-stack slices by slice thickness to determine the average volume. Both methods give consistent results. In addition to volume measurements, fluorescence intensity can also be assessed with maximum intensity projection of VM7 glomeruli¹⁶. The weighted sum of all pixels is assessed via ImageJ sum slice max projection, with the pixel number at each intensity level summed to calculate the overall intensity^8,9. The background can be subtracted by quantifying an extra-glomerular region of identical size. This subtraction method is more time-consuming but more accurate, as it accounts for varying levels of background antibody labeling in each brain sample to produce a more representative quantification of the innervation fluorescence intensity values. Intensity measurements are made in a typical range of 18-30 optical slices (0.37 µm/slice) within a Z-stack, averaged for each brain¹⁶. In nearly all cases, the innervation fluorescence intensity and volume measurements give consistent results.

Across all measurement modalities, the representative results remain unwavering; early-life EB odorant experience results in a significant loss of Or42a neuron innervation. The VM7 glomerulus can be reliably identified by size, shape, and location among other precisely mapped antennal lobe olfactory glomeruli (Figure 2A) and by targeted expression of the Or42a-mCD8::GFP membrane label coupled to glomerular labeling with anti-neural cadherin (CadN; Figure 2B). Timed exposure to the odorant vehicle control (mineral oil) results in Or42a neuron innervation indistinguishable from an untreated, normal animal. Still, EB odorant experience in the first few days post-eclosion (dpe) causes synaptic glomeruli pruning (Figure 2C). This pruning mechanism is entirely restricted to the brief early-life critical period^8,10. We typically now use 24 h odorant exposures to assess VM7 glomeruli synaptic pruning and the resulting decrease in innervation volume in the juvenile Drosophila brain. A 24 h exposure (0-1 dpe) to the odorant vehicle control (mineral oil) maintains the normal, dense Or42a-mCD8::GFP innervation (green) in the VM7 synaptic glomeruli (white dashed circles) in both brain hemisphere antennal lobes (Figure 2D, left). In striking contrast, 24 h exposure (0-1 dpe) to 25% EB odorant dissolved in the mineral oil results in strong pruning of Or42a-mCD8::GFP innervation (green) and a loss of the synaptic glomerular volume (Figure 2D, right). These results are both visually obvious and consistently clear. One of the greatest attributes of this odorant experience protocol is the clarity of the brain imaging results, which are also extremely amenable to straightforward innervation fluorescence intensity and volume quantifications.

The major advantage of this odorant experience protocol is the ability to harness it for the study of temporal restriction and odorant dose-dependent mechanisms during critical period synaptic pruning. For EB-driven Or42a neuron synapse elimination in the VM7 glomerulus, the remodeling mechanism is primarily restricted to the early juvenile (0-2 dpe), with no detectable experience-dependent synaptic pruning at maturity (7-9 dpe) except minor modulation at the highest EB exposure levels⁸. Such temporal restriction defines a critical period, and the understanding of this timing mechanism is an important objective. Another attribute of this protocol is that experience-dependent synaptic pruning is EB dose-dependent (Figure 3). Timed exposure to the odorant vehicle control (mineral oil) with 0% EB odorant results in normal Or42a neuron innervation of the VM7 synaptic glomerulus, which is indistinguishable from an untreated animal. Calculation of the Or42a neuron innervation 3-dimensional volume shows a size typical of a normal juvenile brain (2260 µm³; Figure 3A, left). Representative results show that increasing EB odorant concentration causes progressively greater synaptic glomeruli pruning. For example, a 24 h exposure (0-1 dpe) to 15% EB odorant results in clear and consistent innervation loss (Figure 3A, middle). Quantification of the 3D innervation volume shows a striking reduction compared to vehicle controls (760 µm³ vs. 2260 µm³). A 24 h exposure (0-1 dpe) to 25% EB odorant causes even more innervation loss (281 µm³; Figure 3A, right). Representative quantifications for this range of EB odorant concentrations show consistent results for fluorescence intensity (Figure 3B) and innervation volume (Figure 3C). This dose dependence allows for the study of the means to prevent synaptic pruning as well as enhance pruning with lower odorant exposures to demonstrate increased synaptic glomeruli remodeling. Some prior studies have shown the opposite effect of increased glomerular volume using other odorants with this protocol¹⁰. EB odorant exposure in the critical period always results in striking and easily detectable synaptic glomeruli pruning.

Figure 1: Flowchart of odorant exposure studies in the Drosophila juvenile brain. (A) Pharate dark pupae (90+ h post-pupariation at 25 °C) are collected and placed in odorant chambers. Under odorant conditions, 4 h are allowed for eclosion, and then juvenile flies are transferred to a fresh vial. (B) Age-matched flies are divided into either the vehicle control (100% mineral oil) or the ethyl butyrate (EB; % v/v in mineral oil) experience conditions. A fine wire mesh (inset) allows odorant access. Flies are typically exposed for an additional 20 h for a total of 24 h of exposure, 0-1 day post-eclosion (dpe). (C) Flies are dissected by first removing the head, and then fine forceps are used to delicately remove the cuticle exoskeleton and isolate the juvenile brain. The isolated brain is cleared of all surrounding tissues and protruding trachea. (D) Isolated brains are fixed, blocked, and then sequentially incubated in primary and secondary antibodies. The range of antibodies used typically includes a general membrane probe to reveal the antennal lobe synaptic glomeruli and a specific membrane probe for the Or42a olfactory sensory neurons. (E) Labeled brains are placed in a mounting medium on a glass microscope slide with tape spacers under a sealed coverslip. Correct brain orientation (dorsal surface up) is critical for optimal imaging of the antennal lobes. (F) The brains are imaged using a laser-scanning confocal fluorescent microscope. All samples are blinded to genotype and experience conditions prior to imaging and subsequent quantified analyses. Please click here to view a larger version of this figure.

Figure 2: Odorant experience-dependent critical period synaptic glomeruli pruning. (A) The antennal lobe (AL) olfactory glomeruli are precisely mapped. A few representative synaptic glomeruli are shown. The Or42a receptor olfactory sensory neuron (OSN) class specifically innervates the VM7 glomeruli. (B) Or42a-mCD8::GFP tagged neurons (green) densely innervate the paired VM7 glomeruli on the left (L) and right (R) sides of the brain, with N-Cadherin (CadN) antibody labeling of all the AL synaptic glomeruli (magenta). The fly line is w¹¹¹⁸; Or42a-mCD8::GFP/+. (C) Schematic graphical representation of the Or42a OSN innervation of the VM7 synaptic glomeruli in the vehicle control condition (100% mineral oil; left) and ethyl butyrate exposure (% EB v/v in mineral oil; right) during the juvenile critical period. Experience-dependent innervation pruning only occurs with early-life EB exposure, with synapses eliminated and axons lost, resulting in reduced innervation volume. (D) Representative confocal images of Or42a OSN innervation of the paired VM7 synaptic glomeruli (dashed white outline circles; Or42a-mCD8::GFP, green) following oil vehicle control (left) and 25% EB exposure (right) for 24 h (0-1 dpe). The EB odorant experience drives synaptic pruning and the loss of innervation volume. This experience-dependent mechanism is temporally restricted and transiently reversible only within a well-defined olfactory critical period in the first few days post-eclosion. Please click here to view a larger version of this figure.

Figure 3: Quantifying EB dose-dependent synaptic glomeruli innervation fluorescence intensity and volume. (A) Or42a-mCD8::GFP labeled innervation in a VM7 synaptic glomerulus (green) exposed from 0-1 dpe to the odorant vehicle mineral oil control (0% EB, left), 15% EB (middle), or 25% EB (right). Below: The same images with glomerular innervation outlined (white circles) and the 3-dimensional innervation volume calculated. The vehicle control has VM7 synaptic glomerulus innervation by Or42a neurons that is indistinguishable from the untreated animals. Critical period experience with 15% EB odorant exposure results in obvious synaptic glomerulus pruning, with a 66% decrease in the 3-dimensional innervation volume in this case. Exposure to the higher level of 25% EB results in greater pruning, with a nearly 90% loss of synaptic glomerulus volume in this case. Thus, one can image dose-dependent pruning as the odorant experience increases. (B) Representative results of the Or42a-mCD8::GFP labeled fluorescence intensity within VM7 synaptic glomeruli exposed from 0-1 dpe to the odorant vehicle control (oil, green), 15% EB (middle, orange), or 25% EB (right, black). (C) Representative results of the VM7 glomeruli 3-dimensional innervation volumes under the same critical period odorant exposure conditions. The scatterplots show all the individual data points with the mean ± SEM. The significance is indicated as p≤0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***), and p ≤ 0.0001 (****). Please click here to view a larger version of this figure.

Discussion

The odorant exposure and brain imaging protocol presented here can be used to reliably induce and quantify experience-dependent olfactory sensory neuron synaptic glomeruli pruning during an early-life critical period. Earlier studies utilizing this treatment paradigm to explore olfactory circuit remodeling began odorant exposure on the 2^nd day after eclosion^3,4,5. In contrast, we begin odorant exposure in pharate pupae prior to adult eclosion and find that the vast majority of remodeling occurs within the first 24 h post-eclosion^8,9. Thus, an essential step in this odorant exposure protocol is to place staged dark pupae into odor chambers prior to eclosion (Figure 1). This is not for pupal odorant exposure per se but rather to ensure odorant exposure immediately upon adult eclosion. We selected eclosed juvenile adults 4 h following staging to synchronize ages within a tight temporal window and to avoid differences in odorant exposure times. This earlier experience interval is in line with recent studies demonstrating EB olfaction habituation closes after just 2 days post-eclosion¹⁰. However, the consequences of odorant exposure timing differences within the critical period (e.g., 0-1 vs. 1-2 days) have not been well characterized. Or42a olfactory sensory neurons selectively innervate the VM7 glomerulus in the antennal lobe, which is easily identified based on the precise glomerular maps and targeted innervation labeling (e.g., Or42a-mCD8::GFP; Figure 2). EB odorant exposure during the critical period does not alter Or42a promoter expression or reporter levels⁸. Timed odorant exposure causes sequential synapse elimination and axon innervation loss from Or42a receptor neurons. This pruning is completely reversible within the early-life critical period (0-2 dpe) but not at adult maturity (7-9 dpe) when changes in the odorant environment no longer alter innervation. The tight temporal restriction of this sensory experience-dependent mechanism is a major outstanding question in the field. Many attempts are being made to re-open critical period-like plasticity at maturity. The protocol given here is a powerful avenue for investigation.

The ease of the odorant exposure and brain imaging protocol described here allows for wide accessibility and broad adaptability. The odorant exposure experience onset, termination, and duration are all completely at the investigator's discretion to allow targeted studies of critical period properties, remodeling reversibility, and the re-opening of critical period-like plasticity. Likewise, the strength of the odorant experience can be easily manipulated for the study of dose-dependent synaptic glomeruli pruning (Figure 3). Importantly, studies that use lower EB concentrations (e.g., 15% EB) can be used to test manipulations that could enhance the rate or extent of synaptic glomeruli remodeling. For example, results could show synaptic pruning at lower EB levels (e.g., 15%) to be indistinguishable from higher EB levels (e.g., 25%). Across different manipulations, outcomes can be compared to consistent innervation of odorant vehicle controls and the robust synaptic glomeruli pruning occurring at 25% EB (Figure 3). One limitation of this method is that the odorant exposure entirely relies upon the volatility of the dissolved odorant. Variable factors such as temperature and relative humidity affect the odorant exposure effectiveness⁸. Fly density in the odorant chambers could also potentially impact the odorant exposure, for example, through the production of fly-derived odorants. A density of no more than 20-25 flies per vial is recommended. The odorant concentration required to induce very strong synaptic glomeruli pruning is likely higher than what is typically experienced in nature^10,17. Much higher resolution imaging techniques would be required to assay synapse elimination at lower odorant concentrations. However, the current protocol allows consistent and reliable studies for a range of manipulations.

There are immense benefits to this odorant exposure and brain imaging protocol. Drosophila is an enormously advantageous animal model founded on an extensive and sophisticated genetic toolkit, which can be systematically exploited for the broad study of the gene-environment interactions in critical period olfactory circuit remodeling. Examples shown here include the Or42a-Gal4 transgenic driver line, UAS-mCD8::GFP membrane marker, and Or42a-mCD8::GFP direct fusion. There are no distinguishable differences between synaptic glomeruli pruning with the Or42a direct fusion versus Gal4/UAS indirect labeling approaches^8,9. The Or42a-Gal4 driver provides the ability to both manipulate the targeted olfactory receptor neurons and visualize experience-dependent changes in VM7 innervation, whereas the direct fusion line frees up the binary UAS-Gal4 transgenic system for other cell-targeted manipulations. Many other mutant and transgenic lines from the extensive Drosophila toolkit can be employed to test temporal-restriction and experience-dependent mechanisms, including identifying cells mediating synaptic pruning, the intercellular signaling pathways relaying olfactory experience information, and the intracellular machinery required for targeted synapse elimination. Overall, this protocol allows the systematic interrogation of many mechanisms driving olfactory experience-dependent brain circuit remodeling. This odorant exposure and brain imaging protocol can also be used to investigate critical period-specific olfactory experience behavioral modifications, genetic and pharmaceutical manipulations that may enable the re-opening of critical period-like plasticity at maturity and other aspects of olfactory circuit plasticity beyond synaptic glomeruli pruning.

Materials

Name	Company	Catalog Number	Comments
For Odor Exposure
Drosophila vials	Genesee Scientific	32-110
Ethyl butyrate	Sigma Aldrich	E15701
Microcentrifuge tubes	Fisher Scientific	05-408-129
Mineral oil	Sigma Aldrich	M3516
Odor chambers	Glasslock
Paint brushes	Winsor & Newton	Series 233
Parafilm	Thermofisher	S37440
Wire mesh	Scienceware	378460000
Brain Dissection
Ethanol, 190 proof	Decon Labs	2801	Diluted to 70%
Forceps	Fine Science Tools	11251-30	Dumont #5
Paraformaldehyde	Electron Microscope Sciences	157-8	Diluted to 4%
Petri dishes	Fisher Scientific	08-757-100B
Phosphate-buffered saline	Thermo Fisher Scientific	70011-044	Diluted to 1x
Sucrose	Fisher Scientific	BP220-1
Sylgard	Electron Microscope Sciences	24236-10
Triton-X 100	Fisher Scientific	BP151-100
Brain Immunocytochemistry
488 goat anti-chicken	Invitrogen	A11039
546 goat anti-rat	Invitrogen	A11081
Bovine serum albumin	Sigma Aldrich	A9647
Chicken anti-GFP	Abcam	13970
Coverslips	Avantor	48366-067	25 x 25 mm
Double-sided tape	Scotch	34-8724-5228-8
Fluoromount-G	Electron Microscope Sciences	17984-25
Microscope slides	Fisher Scientific	12-544-2	75 x 25 mm
Nail polish	Sally Hansen	109	Xtreme Wear, Invisible
Normal goat serum	Sigma Aldrich	G9023
Rat anti-CadN	Developmental Studies Hybridoma Bank	AB_528121
Confocal/Analysis
Any computer/laptop
Confocal microscope	Carl Zeiss		Zeiss 510 META
Fiji software	Fiji		Version 2.14.0/1.54f