dTBI: A paradigm for closed-head injury in Drosophila

Drosophila models have been instrumental in providing insights into molecular mechanisms of neurodegeneration, that are applicable to human disease. We have recently described a model of controlled head injury to ies, which remarkably parallels many of the physiological responses of humans to traumatic brain injury (TBI). This protocol describes the construction, calibration and use the of the Drosophila TBI (dTBI) device, a platform that employs a piezoelectric actuator to reproducibly deliver a force, which briey compresses the y head against a metal surface. The extent of head compression can be specied, allowing the operator to set different thresholds of injury. Using readily available components and tools, the device can be assembled and calibrated within two days, for a total cost of ~$700. The dTBI device can be used to harness the power of Drosophila genetics and perform large-scale genetic or pharmacological screens, using a 7-day post-injury survival curve to identify modiers of injury.


Introduction
INTRODUCTION Traumatic brain injury (TBI) is a complex disease process involving a multitude of pathophysiological processes that contribute to long-term changes in brain structure and function 1,2 . Mammalian models have uncovered many biomechanical, metabolic, biochemical and molecular events that contribute to the secondary injury response of TBI 3,4 . However, the relative contributions of these mechanisms to the progression of long-term injury, and the interactions of various pathways are di cult to study in mammalian models, given the long lifespans and relative di culty of genetic manipulations. In the last few years, non-penetrating TBI models have been developed in multiple small organisms such as Caenorhabditis elegans 5,6 , zebra sh 7,8 and Drosophila melanogaster [9][10][11] , which offer the advantages of short generation times and lifespans, tractable genetics and the potential to be adapted to high throughput applications like genetic or pharmacological screens 12 .

Development of the method
Drosophila has been instrumental in elucidating the mechanisms of nervous system development and complex behaviors such as learning, memory, circadian cycles and courtship 13,14 . Importantly, many CNS genes and pathways are conserved between ies and humans, with an estimated 75% of human diseasecausing genes also having functional y orthologs 15,16 . With sophisticated genetic toolkits that enable researchers to activate or inhibit speci c genes in de ned tissues with temporal control, Drosophila has provided mechanistic insight into many neurodegenerative diseases. These include models of axon and penetrative brain injury which selectively sever particular neurons like the olfactory nerve 17 and wing axons 18,19 , or pierce a needle through the head to cause damage to brain tissue 20 . However, these models do not re ect the diffuse, brain-wide injury that occurs during a closed head trauma. The rst model of non-penetrating, closed head TBI in Drosophila uses a spring to deliver a mechanical force to a cohort of ies in a plastic vial, causing whole body injury 9 . Another model employs a bead-mill homogenizer to deliver the force and similarly traumatize the ies 10 . Although these models demonstrate that a portion of body-injured ies have neurodegenerative de cits 9,10,21 , the injury is neither tissue-speci c nor uniform between individual ies, thus increasing the heterogeneity in the population response. A head-speci c model of TBI was described, that uses a pipette tip to immobilize ies and a ballistic impactor to strike the head 11 . Since the experimental setup involves loading and injuring ies individually, it could be challenging to perform high throughput screens.
We have recently developed a head-speci c model of Drosophila TBI, which we term dTBI, that is suitable for large-scale genetic or pharmacological screens (Saikumar et al., in press). The protocol uses a modi ed Heisenberg collar to immobilize the ies and a piezoelectric actuator to deliver a compressive injury to the y head within 250ms. We de ne 3 thresholds of injury -mild, moderate and severecaused by compressing the y head to 35%, 40% and 45% respectively. Head-injured ies display a remarkable similarity to mammalian TBI, and several phenotypes are observed in an injury severity dosedependent fashion (Saikumar et al., in press). Notably, neurological de cits such as immediate loss of righting re ex, locomotor de cits, spontaneous seizures, age-onset learning de cits, and a reduction in lifespan are observed as a consequence of dTBI. The brain undergoes injury severity-dependent vacuolization within a day of dTBI, which progresses with time. By contrast, cellular necrosis and blood brain barrier dysfunction occur early after injury and subside with time. The injured head also acutely upregulates proteasomal activity, markers of oxidative stress, and molecular chaperones. A full list of the features observed with dTBI along with the timeframes of their occurrence is presented in Fig 1. Overview of the method In this protocol, we describe the assembly of the dTBI circuit, construction and calibration of the device, the technique of collaring ies and administering dTBI, and discuss various factors that affect the reproducibility of the technique. The main components of the device are: the piezoelectric actuator mounted on a platform, the dTBI controller which houses the circuit controlling the piezoelectric, and a Heisenberg collar that holds the y with the head resting stably on the metal plates.
Microcontrollers are commonly used to detect and respond to incoming electrical signals, often using these signals to gate, or control, events. In this design, a microcontroller is used to constantly check whether a switch is closed in the circuit. Upon closing, the microcontroller sends a separate signal to temporarily direct a voltage pulse into a voltage ampli er. The time duration of the voltage pulse sent to the ampli er is prescribed by a short program uploaded to the microcontroller. The amplitude of the voltage pulse is set using a potentiometer. The ampli ed voltage signal is used to brie y de ect the piezoelectric which compresses the y head against the metal plates of the Heisenberg collar. The actuator is mounted on a platform, which provides enough working space for mounting more than one piezoelectric, if desired. On the same platform, we have mounted a narrow piezoelectric used to hit one y at a time (low throughput) and a wide piezoelectric that can hit up to 6 ies simultaneously (high throughput). Once the circuit is assembled, the code is uploaded to the microcontroller through a computer. The device is calibrated to ensure that the displacement of the piezoelectric increases linearly with the input voltage within the working range of the piezoelectric. Once the dTBI device is constructed, the nal step prior to using it is the creation of a calibration curve to determine the relationship between voltage and the extent of head compression.

Experimental design
The overall experimental design is presented in Fig 2. The construction and full calibration of the device is a one-time setup. Thereafter, it is su cient to actively monitor the piezoelectric for signs of wear (reduction in % head compression for the same voltage), every 6 months. If the piezoelectric is replaced, it is advisable to verify that the new device yields similar mechanical and biological results. A spot-check should be done to verify that the head compression (using 6 ies) and 7d post-injury survival (using 50 ies) for severe dTBI are similar to the calibrated values. We typically use male ies for all experiments, and it is important to note that because of the difference in head sizes between sexes, the device has to be calibrated separately for male and female ies.
The device should be calibrated with the genotype that is expected to be most frequently used. We de ne severe dTBI as having a median post-injury survival of ~10d, moderate dTBI ~22d and mild dTBI ~43d. As a starting point, we recommend 45% head compression for severe dTBI, 40% for moderate and 35% for mild injury. When working with a new genotype, a pilot experiment with ~50 ies can be used to gauge the post-injury survival after a severe dTBI. If the ies are healthier or weaker than anticipated, the head compression can be increased or decreased by 5% to bring the post-severe injury median survival tõ 10d.
To estimate the number of ies needed when designing a new experiment, it is necessary to take into account the genotype (if anticipated to be weaker/healthier than the genotype used for calibration), the dTBI severity (mild vs. moderate vs. severe), the experimental readout (lifespan vs. molecular assays), the timepoint up to which ies are to be aged, and the controls (sham, vehicle control, genotype background control). We typically use 3d old male ies for dTBI experiments. The required number of animals are subjected to dTBI or sham injury and allowed to recover in food vials until the desired timepoint. When aging injured ies, we recommend ipping to fresh food vials at least every 2d. A number of assays can be used to study the effect of dTBI on brain health and longevity (Fig 1). In this protocol, we describe a 7d post-injury lifespan assay to quantify survival after dTBI. Certain steps require basic machining (drilling holes or cutouts into the enclosure, drilling and tapping holes into the platform), which can be accomplished using standard equipment or through assistance from a machine shop.
The dTBI protocol is simple to master and only requires basic y pushing expertise. A beginner should start with learning to collar 5-6 ies in under 1 min without any mortality (see dTBI Procedure below).
Collar the anesthetized ies, wait for the last collared y to wake up, then remove the ies and return them to fresh y food. Maintain and observe these ies for the next 2-3d, since any mortality associated with rough handling should be evident by this time. Practice the technique until able to rapidly collar ies without any adverse effects.
The severe dTBI 7d post-injury survival curve is the most e cient way to assess the dTBI technique.
Practice until able to hit a cohort of 100 dTBI ies within 60 min with the low throughout device, or 30 minutes with the high throughput device. With the strain of w 1118 ies in our lab, median lifespan for severe dTBI is ~10d; it is important to note that this may vary with different strains across different labs.
There are multiple factors in addition to genotype that in uence post-injury survival, such as the gap between the head and piezoelectric since it determines extent of head compression, and the positioning of the piezoelectric above the head. It is important to keep these parameters consistent between experiments and between different operators to achieve reproducible results.

Applications and limitations of the method
The dTBI paradigm recapitulates key mammalian phenotypes of injury, making it an ideal platform in which to conduct large-scale genetic or small molecule screens, for the identi cation of key molecular pathways and interventions that ameliorate the effects of TBI. The initial experiments were done with a low throughput version of the device that injures a single y at a time (Saikumar et al, in press). Here, we also describe a high throughput version that can injure multiple ies simultaneously in order to scale up the injury process (Movie 1). The severity of dTBI can be modulated either through the extent of head compression or through repetitive injury. We observe that post-injury survival and the vacuolization of the brain are both excellent early measures of the organism's response to dTBI, capable of distinguishing between different injury severities and can be used to identify genetic or environmental modi ers of dTBI (Saikumar et al., in press). For severe dTBI, a large portion of the mortality occurs within the rst week of injury, making the 7d post-injury survival curve a quick and e cient screening tool. Techniques for mass histology of Drosophila have been described for assessing vacuole pathology in para n-embedded sections 22 , allowing rapid evaluation of the brain morphology after dTBI as a complementary approach.
A potential limitation with the paradigm is that the response is dependent on the genetic background, so

Procedure
The circuit diagram of the device is presented in Fig 3a,  3. On the power supply (A in Fig 4, S1), examine the red button that allows you to adjust the voltage output from the supply. Use a at bladed screwdriver to rotate this voltage setting to 12V. From the available power supply plugs that came with the power supply, choose the largest diameter plug (5.5mm diameter) and connect this plug to the end of the black wire extending from the power supply. This is the male DC power supply plug (A.1 in Fig 4, S1) that will attach to the female jack of the panel mount in the enclosure.
4. Use wirestrippers to strip the black wire and red wire from the panel mount, exposing approximately ¼" of the covering from each wire. If you want to avoid soldering, use a male Dupont connector to attach metal pins to the red and black wire. Otherwise, solder a single header pin to each end of the red and black wire.  Fig 4, S2) in a helping hands tool, which uses clips to hold electronics components safely while working on the component. Make sure the bottom of the buck converter is exposed. Locate the blue potentiometer on the buck converter ( Fig S2) and identify the three pins that connect this potentiometer to the buck converter. Turn on a soldering iron, wait until it reaches the working temperature, and then touch the tip of the soldering iron to one of the exposed potentiometer pins. You will soon melt the solder that connects the pin to the buck converter. Use this to remove the solder from each of the three pins of the potentiometer. Once complete, you can remove the potentiometer from the buck converter.
Page 10/38 CAUTION: The removal of the potentiometer from the buck converter must be done carefully or you risk damaging other components of the converter. Place the tip on a pin for a few seconds, remove it, and then place it on the pin for a few seconds more. Eventually, you will heat and melt the solder.
NOTE: The potentiometer that is supplied with the buck converter does not have an appropriate working range that is suitable with the operating range of the piezoelectric.
8. Once the potentiometer is removed from the buck converter, place the buck converter in the enclosure, locating it in the bottom left quadrant. Orient the buck converter so the +OUT and -OUT terminals point to the right. Make a mark on the enclosure base to identify the location of the two mounting holes (B.1 in Fig  4, S2) for the converter. Remove the converter, drill the pilot holes into the enclosure, and install nylon standoffs to mount the buck converter. Cut and strip a green wire at both ends, soldering one end to the +OUT on the buck converter and connecting the remaining end to position D10 on the breadboard (connecting to terminal #4 of the relay). Connect a new black wire, also stripped at both ends, from GND rail to position G13 of the breadboard (connecting to terminal #8 of the relay).
12. Strip the ends of the 3 colored wires (red, black and white) from the digital display (E in Fig 4, S3). Connect the end of the red wire to the +12V voltage rail of the breadboard, and the black wire to the GND rail. Connect the end of the white wire (colored green in Fig 4, S3 for easy visualization) into the breadboard at position C10 (connecting to terminal #4 of the relay).
CRITICAL STEP: At this point, you have assembled a circuit that will allow you to adjust the voltage input to the buck converter, reading out voltage supplied to the voltage ampli er. This is a good point to check that the circuit is working properly. Plug in the power supply to a wall outlet, connect the male DC plug into the female mount and turn the knob on the potentiometer. You should see the voltage on the display change as you turn the potentiometer.
For Steps 13-16, refer Fig S4   13. Place the Arduino Microcontroller (F in Fig 4, S4) in the top right quadrant of the enclosure. Position the Arduino to orient the USB connect towards the back wall of the enclosure, touching the USB connection to the wall. Mark the four mounting holes for the Arduino (F.1 in Fig 4, S4), remove the microcontroller and drill the appropriate holes in the enclosure base. Place the Arduino back in position and mark the opening needed for the USB connection (approximately 14mm x 14 mm -F.2 in Fig 4, S4). Make sure to take into account the extra vertical height from the standoff. Cut the opening for the USB connection and mount the Arduino using nylon standoffs, just as you mounted the buck converter earlier.
14. Cut and strip a piece of red wire, connecting one end to the Vin terminal on the Arduino and the other end into the +12V voltage rail of the breadboard. Similarly cut and strip a piece of black wire, connecting it from one of the GND terminals of the microcontroller to the GND rail of the breadboard. 15. Cut and strip a red wire, soldering one end into a post of the pushbutton switch (G in Fig 4, S4). Connect the remaining end of this red wire into the +5V terminal on the Arduino microcontroller. Using a second red wire, solder one end to the remaining terminal of the pushbutton switch. Connect the remaining end of the red wire to pin #7 of the microcontroller. 16. Connect a new red wire, stripped at both ends, from pin #13 of the Arduino microcontroller to position G10 of the breadboard (connecting to terminal #5 of the SPST relay). 18. Cut and strip the ends from one half of the electrical terminal connector (I in Fig 4, S5). If you want to avoid soldering, use a male Dupont connector to connect the electrical terminals into the breadboard circuit. Otherwise, solder header pins to each end of the wire, connecting the pin from the red wire to position I20 (the +OUTPUT terminal of the proportional voltage booster). Next, connect the pin from the black wire to position I10 (the -OUTPUT terminal of the proportional voltage booster).
19. At this point, the circuit is complete. Drill a hole in the top of the enclosure to mount the potentiometer.
In addition, drill a hole in the side for the electrical connector, tying a loop in the wiring to prevent it from pulling out from the circuit. Mount the digital display and pushbutton to the top of the enclosure. 24. At this point, you can test the functionality of your circuit. With the power supply plugged into the wall, press the pushbutton switch. If the code is working correctly, you should see an LED on the microcontroller blink momentarily. In addition, you can connect the output from the proportional voltage booster to a multimeter, adjust the voltage on the digital display, and see the output from the proportional voltage booster increase temporarily when you press the pushbutton switch.

Construction of the dTBI apparatus
For Steps 25-28, refer Fig S6   25. Using the remaining half of the electrical terminal connector, cut, strip and solder a small spade terminal onto each wire (J in Fig 4, S6). Fig 4, S6) to mount the piezoelectric actuator (L in Fig 4, S6). Ensure that the gap between the drilled holes corresponds to the gap between the holes provided on the piezoelectric mount. Use thin washers (circle in J in Fig 4, S6), sized for the 4-40 mounting screws (hexagon in J in Fig 4, S6), to adjust the height of the piezoelectric actuator relative to the polycarbonate sheet. You will want a height that allows you to insert the ies, immobilized in the Heisenberg collar (M in Fig 4, S6), under the actuator without contact (See next section Collars for exact details).

Drill and tap two holes in the polycarbonate sheet (K in
27. Insert the screw through one of the holes on the piezoelectric mount, the washers and spade terminals to attach the actuator to the polycarbonate sheet (Fig S6). Repeat for the second screw. At this point, the actuator will be securely mounted to the sheet.
28. Solder the wiring from the mounted actuator to the respective electrical spade terminals (red to red; black to black) connecting the piezoelectric to the control circuit.
CRITICAL STEP: Con rm that the control circuit works. Attach the components together, plugging in the power supply to the wall outlet and connecting the electrical output from the control circuit to the piezoelectric actuator. You should see the piezoelectric de ect when the pushbutton is activated. 30. Multiple piezoelectric actuators can be mounted on the same platform and used with the same voltage control circuit (Fig 3b). When switching between different actuators, attach the half of the electrical terminal connector from the voltage control circuit to the other half that is connected to the piezoelectric you wish to use. The low throughput piezoelectric (A in Fig 5a) can be used to injure a single y head at a time, while the high throughput piezoelectric (B in Fig 5a) can injure up to 6 simultaneously. 32. Importantly, the space between the metal plates needs to be precisely set to 125µm, which we nd is the optimal gap that allows ies to slide through easily, while still providing the stable bottom surface against which the head is compressed (C in Fig 5a).
33. The distance between the top of the head and the piezoelectric is also important, since it is one of the factors determining the exact magnitude of head compression. When mounting the piezoelectric to the platform, it is important to use the appropriate number of washers so that the gap between the surface of the collar and the piezoelectric is 393±13 µm. This allows for y heads to easily slide under, and ne adjustments can then be made by tightening the mounting screw of the piezoelectric. For our w 1118 male ies, the average height of the y head was 319±8 µm (as measured from the base of the head to the highest point of the eye). All the calibrations and subsequent experiments were performed with a gap of 67±13 µm between the piezoelectric and the y head (as measured between the highest point of the eye to the piezoelectric) (Fig 5b, Movie 2).
CRITICAL STEP: It is important to ensure that the gap is similar between calibration and the actual experiments. If the gap is reduced between calibration and the nal experiment, the same voltage will cause a larger magnitude of compression. Whereas if the gap is increased, the resulting compression will be smaller for the same voltage setting.
High throughput dTBI device 34. The collars to be used with the multi-dTBI device must have at-head screws on the top plate to ensure that the collar can slide underneath the piezoelectric actuator (D in Fig 5a). Note that commercially bought screws may need to be attened further to avoid the piezoelectric striking them. Attempts at machining a longer collar that could t under the piezoelectric were unsuccessful because it was challenging to keep the metal plates absolutely at throughout the length of the collar. This resulted in unequal head compression across a single cohort of ies.
CRITICAL STEP: As far as possible, the metal plates must be parallel to the piezoelectric such that the gap between collar and piezoelectric is uniform throughout the length of the collar. Additional screws throughout the metal plate may be used to keep it perfectly at.

Calibration (Timing: ~1d)
Generation of voltage vs. displacement graph 35. Set up the dTBI device under the Leica Z16 APO macroscope. Place a collar underneath the piezoelectric and move the 45˚-angled mirror up against the collar (Fig 5c). Note that in this case, the collar is only used to ensure that the placement of the mirror is consistent (between multiple video rounds, or between calibrations with and without y head compression).
36. Adjust the brightness, zoom and focus to capture the re ection of the piezoelectric in the angled mirror.
37. Using a frame rate of at least 10 fps, capture 3 replicate videos of the piezoelectric de ection events in 5V steps, starting from 35V till 80V.
38. Analyze the videos in FIJI, using the Manual Tracking plugin to track a single pixel on corner of the piezoelectric to measure the y-displacement.
39. Generate a graph between voltage and y-displacement to ensure that the piezoelectric responds linearly to the voltage.
Generation of voltage vs. head compression graph 40. Set up the dTBI device and angled mirror under the Leica Z16 APO macroscope as above (Fig 5c). 41. Collar 5 ies (see dTBI Procedure for detailed instructions) and once they are awake, position the last collared y underneath the piezoelectric (see dTBI Procedure for detailed instructions , Fig 5d).
42. Place the angled mirror against the collar and adjust the brightness, zoom and focus to capture the re ection of the piezoelectric and the y head in the mirror.
43. Capture videos of the compression event in 5V steps, starting from 35V till 80V. Do not reuse the same y for multiple videos; instead use a fresh y for each compression recording. 44. Obtain 3-6 replicate videos for each voltage step, using ies of the same genotype obtained from different bottles to control for natural variation in head size.
45. Analyze the videos in FIJI and obtain the % head compression by comparing the y head heights between frames of no compression and maximum compression (Fig 5b, Movie 2). 46. Use the equation of the line obtained from linear regression analyses to calculate the voltage required for 35% (mild), 40% (moderate) and 45% (severe) compression.
47. Perform a survival analysis on sham, mild, moderate and severe injury to identify the median lifespan.
CRITICAL STEP: Since the y background is one of the factors important to survival post-injury, it is possible that 45% head compression may be too severe for some genotypes. It is important to de ne "mild", "moderate" and "severe" thresholds of head compression according to the survival response. In our hands, the median lifespan post-injury is 10d for severe dTBI, 22d for moderate, and 43d for mild. TROUBLESHOOTING 48. After the initial calibration, the device needs to be assessed every 6 months for piezoelectric wear and tear. Use 6 ies to measure the % head compression at the voltage calibrated for 45%, and verify that the compression is not signi cantly different from what was calculated during calibration. If the newly measured compression is lower than the calibrated compression (with all other factors being unchanged since calibration), the piezoelectric may need to be replaced. After replacement, re-measure the % head compression for 45% to ensure that it is now comparable to the calibrated measurement. Additionally, use the 7d post-injury survival to verify that the biological response of the new piezoelectric is similar to the old one.

TROUBLESHOOTING
High throughput dTBI device 49. The maximum number of ies that can be simultaneously compressed depends on the maximum voltage rating of the piezoelectric and must be empirically determined. For the piezoelectric used in our device (Q220-A4BR-2513YB from piezo.com), the severe 45% compression injury was reached at 62V when 6 ies were used, and using more than 6 ies caused the 45% compression to occur beyond the maximum voltage rating of the piezoelectric. 50. Before calibration, verify that the gap between the head and piezoelectric is similar across all 6 ies to ensure that all heads are uniformly compressed. 54. We recommend using the 7d post-injury survival assay on severe dTBI ies to ensure that the head compression is even across all 6 ies. When returning ies to food vials during the dTBI process (see dTBI Procedure), split the injured ies according to their position on the collar (last 2 ies on the left can be combined as one group "L", middle two ies as group "M", right 2 ies as group "R"). Track survival of the 3 groups separately and ensure that the 3 lifespan curves are not signi cantly different from each other.
dTBI procedure (Timing: 3.5 -5 min for a cohort of 6 sham, 6 dTBI ies) Low throughput dTBI device 55. Determine the number of ies needed for the experiment taking into consideration dTBI severity, nal experimental readout and the timepoint until when ies need to be aged. Collect the required number of ies in fresh food vials in groups of 30 or less, and age them to 3d. 56. When the ies are 3d old, brie y anesthetize a vial of ies and tip them on to a CO 2 pad. Only tip the number of ies necessary for a single cohort, i.e. 5-6 sham and 5-6 TBI ies.
CAUTION: Maintain the ies on the lightest possible CO 2 anesthesia. Excessive CO 2 will cause the wings to fold upward making it di cult to collar them.
57. Place a collar on the CO 2 pad with the opening on the right side.
58. For steps 58-62, see Movie 3. Under a stereo microscope, select a single y and use a pair of bluntended forceps to pick it up by both wings. Manipulate the y on to its right side, with the straight wings on the right side of the y body and the legs on the left. Ensure that the forceps grasp both wings, close to the y body. This gives better motor control to precisely manipulate the y when collaring.
59. Bring the y to the opening of the collar and gently thread the neck through the gap between the metal plates. Make sure that the head is stably resting on the metal plates before proceeding.
60. Flip the collar over and push the y to the far end. Hold the forceps closed and rest them vertically on the metal groove against the right side of the y body. Push against the y body gently so that it moves smoothly along the collar.
CRITICAL STEP: Ensure that the forceps are closed so as to not injure the y body. It is important to move the y by sliding the forceps along the groove made by the metal plates rather than pushing against the body directly, because the latter can lead to decapitation. TROUBLESHOOTING 61. A slower, but safer alternative is to move the y by using a paintbrush to nudge the head with short pushes.
62. Collar the rest of the ies to have a total of 5-6 ies per collar. Collar the ies for the dTBI group rst, then the sham, so that that dTBI ies are awake by the time the sham ies are collared. CRITICAL STEP: This ensures that the head slides in easily without bumping against the piezoelectric (which can lead to decapitation). Take great care not to let your hand or the screws on the collar graze against the piezoelectric, which can damage it. 65. Viewing through the stereo microscope, position the head so that the piezoelectric is above the third antennal segment. Make sure that the neighboring y is far enough away that it will not be damaged by the piezoelectric as it de ects. · dTBI procedure o Low throughput dTBI device (sham and dTBI) -5 minutes per cohort of 6 ies each (2 minutes to collar 12 ies; 2 minutes to hit the ies; 1 minute to return all ies to food vials) o High throughput dTBI device (sham and dTBI) -3.5 minutes per cohort of 6 ies each (2 minutes to collar 12 ies; 30 seconds to hit the ies; 1 minute to return all ies to food vials)

Anticipated Results
The variation in head compression with increasing voltage should be linear for both the low throughput and high throughput device (Fig 6a, b)  post-injury survival curves show that severe dTBI causes a sharp and early mortality using both the low throughput and high throughput devices (Fig 6c, d). Moderate dTBI causes a mild decrease in survival and mild dTBI has no signi cant effect compared to sham in the early period after injury (Fig 6c). These responses indicate that severe dTBI is an ideal injury setting to quickly assess the e cacy of genetic, environmental or pharmacological interventions aimed at identifying key factors driving brain injury (Saikumar et al., in press). With the evolution of dTBI and other paradigms, we anticipate that Drosophila will eventually be a valuable contributor to our understanding of the basic biology of neural injury mechanisms associated with TBI.  Overview of experimental work ow for dTBI paradigm: The construction and calibration of the device is a one-time setup, with regular monitoring every 6 months for maintenance. The dTBI work ow section is to be repeated for every new experiment.   Representative results for low throughput and high throughput dTBI devices: a, b. Graph indicating that head compression increases linearly with increasing voltage for both devices. Dotted lines indicate mild (35%), moderate (40%) and severe (45%) head compression, and the equation of the line is given below.
Number of ies: 3 per voltage in the low throughput device, 6 per voltage in the high throughput device. c. 7d post-injury survival curve using the low throughput device for sham, mild, moderate and severe injury indicating the severe and moderate injury are signi cantly different from sham in this period. Number of animals: 45 for sham, mild, moderate, 75 for severe injury d. 7d post-injury survival curve using the high throughput device for severe injury. Number of animals: 150 Genotype for all gures: w1118 male.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.