Axonal Transport of Organelles in Motor Neuron Cultures using Microfluidic Chambers System

Summary

Axonal transport is a crucial mechanism for motor neuron health. In this protocol we provide a detailed method for tracking the axonal transport of acidic compartments and mitochondria in motor neuron axons using microfluidic chambers.

Abstract

Motor neurons (MNs) are highly polarized cells with very long axons. Axonal transport is a crucial mechanism for MN health, contributing to neuronal growth, development, and survival. We describe a detailed method for the use of microfluidic chambers (MFCs) for tracking axonal transport of fluorescently labeled organelles in MN axons. This method is rapid, relatively inexpensive, and allows for the monitoring of intracellular cues in space and time. We describe a step by step protocol for: 1) Fabrication of polydimethylsiloxane (PDMS) MFCs; 2) Plating of ventral spinal cord explants and MN dissociated culture in MFCs; 3) Labeling of mitochondria and acidic compartments followed by live confocal imagining; 4) Manual and semiautomated axonal transport analysis. Lastly, we demonstrate a difference in the transport of mitochondria and acidic compartments of HB9::GFP ventral spinal cord explant axons as a proof of the system validity. Altogether, this protocol provides an efficient tool for studying the axonal transport of various axonal components, as well as a simplified manual for MFC usage to help discover spatial experimental possibilities.

Introduction

MNs are highly polarized cells with long axons, reaching up to one meter long in adult humans. This phenomenon creates a critical challenge for the maintenance of MN connectivity and function. Consequently, MNs depend on proper transport of information, organelles, and materials along the axons from their cell body to the synapse and back. Various cellular components, such as proteins, RNA, and organelles are shuttled regularly through the axons. Mitochondria are important organelles that are routinely transported in MNs. Mitochondria are essential for proper activity and function of MNs, responsible for ATP provision, calcium buffering, and signaling processes^1,2. The axonal transport of mitochondria is a well-studied process^3,4. Interestingly, defects in mitochondrial transport were reported to be involved in several neurodegenerative diseases and specifically in MN diseases⁵. Acidic compartments serve as another example for intrinsic organelles that move along MN axons. Acidic compartments include lysosomes, endosomes, trans-Golgi apparatus, and certain secretory vesicles⁶. Defects in the axonal transport of acidic compartments were found in several neurodegenerative diseases as well⁷, and recent papers highlight their importance in MN diseases⁸.

To efficiently study axonal transport, microfluidic chambers that separate somatic and axonal compartments are frequently used^9,10. The two significant advantages of the microfluidic system, and the compartmentalization and the isolation of axons, render it ideal for the study of subcellular processes¹¹. The spatial separation between the neuronal cell bodies and axons can be used to manipulate the extracellular environments of different neuronal compartments (e.g., axons vs. soma). Biochemical, neuronal growth/degeneration, and immunofluorescence assays all benefit from this platform. MFCs can also assist in studying cell-to-cell communication by coculturing neurons with other cell types, such as skeletal muscles^12,13,14.

Here, we describe a simple yet precise protocol for monitoring mitochondria and acidic compartment transport in motor neurons. We further show the use of this method by comparing the relative percentage of retrograde and anterograde moving organelles, as well as the distribution of transport velocity.

Protocol

The care and treatment of animals in this protocol were performed under the supervision and approval of the Tel Aviv University Committee for Animal Ethics.

1. MFC preparation

1. PDMS casting in primary molds (Figure 1)
   1. Purchase or create primary molds (wafers) following a detailed protocol⁹.
   2. Use pressurized air to remove any type of dirt from the wafer platform before proceeding to the coating step. The surface of the wafers should look smooth and clear.
   3. Fill a container with 50 mL of liquid nitrogen. Prepare a 10 mL syringe and 23 G needle.
      NOTE: All procedures from this step forward must be performed in a chemical hood.
   4. In the chemical hood, use the syringe and needle to pool 2 mL of liquid nitrogen. Though it may seem like air was drawn, the syringe is filled with nitrogen (Figure 1A). Place the wafer-containing plate in a sealable container.
   5. Screw open a chlorotrimethylsilane bottle, pierce the rubber cap using the nitrogen-filled syringe, and inject the entire contents of the syringe into the bottle. Without pulling out the needle, turn the bottle upside down and draw back 2 mL of chlorotrimethylsilane.
      NOTE: Because of the syringe pressure, a small amount of chlorotrimethylsilane is sprayed out of the needle. To avoid a hazard, point the needle toward the inner wall of the hood (Figure 1B).
   6. Spread chlorotrimethylsilane uniformly in the container (from step 1.1.4), but not directly on the wafer or wafer-containing plate. Close the container and incubate for 5 min per wafer.
      NOTE: If this is the first time the wafer is coated with chlorotrimethylsilane, a 1 h incubation should be allowed for each wafer.
   7. Do not take the wafers and container out of the chemical hood for 30 min.
      CAUTION: Chlorotrimethylsilane is highly volatile.
   8. Weigh the PDMS base (see Table of Materials) in a 50 mL tube and add PDMS curing agent at a ratio of 16:1 respectively (e.g., 47.05 g of base and 2.95 g of curing agent). Mix for 10 min using a low speed rotator.
   9. Pour PDMS into each wafer-containing plate to the desired height (Figure 1C).
      NOTE: Using thin microfluidic chambers (up to 3-4 mm) improves adherence to the culture dish and prevents leakage.
   10. Place all plates together inside a vacuum desiccator for 2 h (Figure 1E). This process removes the air trapped within the PDMS, thus eliminating air bubbles and forming a clear, uniform mold.
   11. Place the plates inside an oven for 3 h (or overnight) at 70 °C (Figure 1E).
       NOTE: The plates should be level when placed in the oven.
2. PDMS casting in epoxy molds
   NOTE: Because wafer preparation is expensive, requires special equipment, and may damage the fragile wafers, it is possible to generate epoxy replicas of wafers. The replicas are cheaper, more durable, and can be used for mass production of microfluidic chambers.
   1. Cast and cure PDMS (as described in 1.1.8-1.1.11) into the original wafer.
   2. Remove and cut off excess parts of PDMS leaving only the microfluidic elements and the functional area required for processing them into microfluidic chambers.
   3. Immediately wrap the PDMS with thick, sticky tape to prevent it from accumulating dust.
   4. Choose a tissue culture grade plastic dish that fits the entire PDMS inside and leave room for epoxy around it. The distance from the PDMS to the plastic dish should be less than 5 mm.
   5. Prepare a small amount of PDMS mixed in a ratio of 10:1 (base:curing agent). The fresh liquid PDMS will be used to glue the solid PDMS onto the bottom of the plastic plate.
   6. Apply a minimal amount of the liquid PDMS to the center bottom of the plastic plate and then remove the sticky tape from the PDMS and adhere it to the plastic dish bottom. Make sure that the microfluidic elements are facing upwards.
   7. Let the PDMS cure for 30 min in a 70 °C oven.
   8. Prepare the epoxy resin by mixing the base and curing agent in a ratio of 100:45 respectively in a test tube. Different epoxy resins may have different mixing ratios. The required volume for a regular 100 mm plate is approximately 40 mL.
   9. Let the epoxy mix well for 10 min in a rotator until the mixture becomes visibly homogenous (i.e., there are no visible fiber-like artifacts in the liquid).
   10. Centrifuge the epoxy mixture at 400 x g for 5 min to remove air bubbles caught inside.
   11. During centrifugation, spread a thin layer of silicone grease around the walls and all other exposed plastic parts of the culture dish. This will prevent the epoxy from polymerizing with the dish plastic and will enable removal of the cured epoxy easily at the end of the protocol.
   12. Pour the epoxy slowly into the dish until it completely covers the PDMS and goes beyond it by at least 5 mm. Prevent the formation of any bubbles within the epoxy by keeping zero distance between the tube and the plate. Place the plate in a secure place so it will not be moved for the next 48 h.
   13. After 48 h the epoxy should be completely cured. Insert the plate into a preheated oven at 80 °C for 3 h for final curing.
   14. Remove the cured epoxy from the plate and the original PDMS mold by gently yanking the plastic wall of the plate until it breaks. It should then easily separate from the epoxy and peel off.
   15. Once extracted, wipe the remaining grease off the new epoxy replica and inset it upside down (i.e., with the replicated microfluidic elements facing up) into a new culture dish. The epoxy replica is now ready for PDMS casting.
   16. Use pressurized air or N₂ to blow any remains of PDMS or dirt off the epoxy mold and rinse it 2x with isopropanol. Fill it a third time and incubate for 10 min on an orbital shaker plate. Rinse the mold again 3x with isopropanol and discard the remaining liquid. Blow dry with air or N₂ or place in a 70 °C oven until dry.
       NOTE: Follow safety procedures when working with and discarding isopropanol.
   17. Keep the mold plates closed until casting. Follow steps 1.1.8.-1.1.11.
3. Punching and sculpting the PDMS into an MFC (Figure 2)
   1. Cut and remove the PDMS mold from the plate by following the (+) marks on the wafers using a scalpel. Do not use force, as the molds are fragile (Figure 2A).
   2. Follow the instructions drawn on the sketch to punch and cut the chambers depending on the experimental setup (Figure 2B-F).
      1. For spinal cord explant culture (Figure 2C,E), punch two 7 mm wells in the distal side of a the large MFC. Locate the wells in a way that they will overlap with the channel edges. On the proximal side, punch one 7 mm well in the middle of the channel, with minimal overlap so that sufficient space will be left for the explants. Punch two additional 1 mm holes in the two edges of the proximal channel. Turn the MFC with the microfluidic elements facing upwards, and using a 20 G needle, carve three small explant caves on the punched 7 mm well.
      2. For dissociated MN culture (Figure 2D,F), punch four 6 mm wells in the edges of the two channels of a small MFC.
4. Sterilizing the MFC for tissue culture use
   1. Spread 50 cm long sticky tape bands on the bench. Press and pull back the chamber to face the sticky tape (both upper and lower faces) and remove crude dirt. Place the clean chambers in a new 15 cm plate.
      NOTE: Do not press directly on the microfluidic elements when these are facing upwards.
   2. Incubate the chambers in analytical grade 70% ethanol for 10 min on an orbital shaker.
   3. Dispose of the ethanol and dry the chambers in a tissue culture hood or in an oven at 70 °C.
5. Placing the MFC on a glass bottom dish
   1. Place the chamber in the center of a tissue culture grade 35 mm/50 mm glass bottom dish and apply minor force on the edges to make the PDMS and dish bottom bind. To avoid breaking the glass bottom, always apply force on top of a solid surface.
   2. Incubate 10 min in 70 °C. Press the chambers to strengthen adherence to the plate.
   3. Incubate under UV light for 10 min.
6. Coating and culturing
   1. Add 1.5 ng/mL poly-L-ornithine (PLO) to both compartments. Make sure the PLO is running through the channels by pipetting the coating media a few times directly in the channel entrance.
   2. Examine the microfluidic chamber under a light microscope with 10x magnification to check for the presence of air bubbles. If air bubbles are blocking the microgrooves, place the MFC in a vacuum desiccator for 2 min. Later, remove the excess air that got caught in the channels by pipetting the coating media through them. Incubate overnight.
   3. Replace PLO with laminin (3 µg/mL in DDW) for overnight incubation in the same manner.
   4. Prior to plating, wash the laminin with neuronal culture medium.

2. Neuronal culture plating

1. Dissociated motor neuron culture
   1. Using straight scissors and fine forceps, dissect a spinal cord out of an E12.5 ICR-HB9::GFP mouse embryo. Work in an 1X HBSS solution with 1% penicillin-streptomycin (P/S) (Figure 3A-C).
   2. Using microdissection scissors, remove the meninges and the dorsal horns (Figure 3D).
   3. Collect the spinal cord pieces and transfer to tube (#1) with 1 mL HBSS + 1% P/S.
   4. Cut the spinal cords to small pieces using curved scissors and wait for the pieces to settle.
   5. Add 10 µL of trypsin 2.5% and place in a 37 °C water bath for 10 min. After 5 min, mix by tapping the tube. The pieces should form a helix-like clump.
   6. Transfer the clump into a new tube (#2) containing 800 µL of prewarmed L-15, 100 µL of BSA 4%, and 100 µL of 10 mg/mL DNase. Grind 2x, then wait for 2 min to let undissociated pieces settle. Transfer the supernatant to a new tube (#3).
   7. Add 100 µL of BSA 4%, 20 µL of 10 mg/mL DNase, and 900 µL of complete neurobasal medium (CNB, see Table of Materials and Table 1). Grind 8x and wait 2 min. Collect supernatant to tube #3.
   8. Repeat step 2.1.7 and grind 10x. Collect supernatant to tube #3. A small amount of tissue should be left at the bottom of the tube.
      NOTE: If a large clump still remains at the bottom of tube #2, repeat step 2.1.8.
   9. Add 1 mL of BSA 4% cushion to the bottom of tube #3.
   10. Centrifuge at 400 x g for 5 min. Discard the supernatant.
   11. Resuspend the cell pellet by gently tapping the tube, then add 1 mL of CNB medium. Pipette 6x and add 20 µL of 10 mg/mL DNase.
   12. Supplement with an additional 5 mL of CNB medium and transfer 3 mL to a new tube (#4).
   13. Add 1 mL of 10.4% density gradient medium (see Table 2) to the bottom of each tube (#3 and #4). A sharp phase separation between the two interfaces should appear.
   14. Centrifuge at 775 x g for 20 min at room temperature (RT). Centrifuge deceleration should be set to a low level to avoid breakdown of the phase separation.
   15. Cells should be floating, appearing as a cloudy interphase between the media. Collect the cells from both tubes into a new tube (#5) already containing prewarmed 1 mL CNB medium.
   16. Add an additional 4-6 mL of CNB medium.
   17. Add 1 mL of BSA 4% cushion to the bottom of the tube.
   18. Centrifuge at 400 x g for 5 min at RT. Discard the supernatant and resuspend the pellet gently with 1 mL of CNB medium.
   19. Count the cells. A yield of 0.75-1 x 10⁶ MN per spinal cord is expected.
       NOTE: Ventral spinal cord also contains other neuronal subtypes besides motor neurons (e.g., interneurons). MN purity depends mostly on the removal of dorsal areas during dissection and the ability to reach MN enriched rostral areas. To ensure the imaged neurons are MNs, use of a mice strain with an endogenous MN marker, such as HB9::GFP mice, is recommended. To achieve pure MN culture (but with decreased cell yield), use of FACS purification¹⁵ is possible.
   20. Plating dissociated MN culture in the MFC
       1. Concentrate 150,000 MNs per chamber by centrifuging at 400 x g for 5 min.
       2. Aspirate the supernatant and gently resuspend the cells in rich neurobasal medium (RNB) at 4 µL per MFC. RNB is CNB supplemented with additional 2% B27 and 25 ng/mL BDNF.
       3. Remove the medium from both compartments, leaving a low volume equivalent to ~10 µL in the wells of the distal compartment. It appears as a thin ring of medium in the well perimeter.
       4. Slowly load 4 µL of cells into the channel. Take out 4 µL of the well in the other side of the channel, and slowly load them back directly into the channel to reverse the current flow and maximize cell density in the channel.
       5. Verify that the cells have entered the channel using a 10x light microscope and place the chamber in the incubator for 30 min without adding more media.
       6. Slowly add ~10-15 µL of RNB into the proximal and distal wells and place the chambers in the incubator for another 15 min.
       7. Following this incubation, slowly add ~75-80 µL of RNB into each well.
   21. Dissociated MN culture maintenance
       1. One day after plating (DIV1), replace medium with RNB supplemented with 1 µM cytosine arabinoside (Ara-C) to inhibit glial growth.
       2. Two days after Ara-C application (DIV3) replace medium with fresh CNB medium in the proximal compartment (without Ara-C).
       3. In order to enhance crossing of axons through the microgrooves, apply RNB supplemented with 25 ng/mL of glial cell-derived neurotrophic factor (GDNF) and 25 ng/mL of brain-derived neurotrophic factor (BDNF) only to the distal compartment. Maintain a volume gradient of at least 10 µL per well between the axonal distal wells (higher volume) and the proximal wells.
       4. Refresh the medium every 2 days. It can take the axons up to 4-6 days to cross distally.
2. Spinal cord explant culture
   1. Using straight scissors and fine forceps, dissect a spinal cord out of an E12.5 ICR-HB9::GFP mouse embryo. Work in an 1X HBSS solution with 1% penicillin-streptomycin (P/S) (Figure 3A-C).
   2. Using microdissection scissors, remove the meninges and the dorsal horns (Figure 3D).
   3. Cut the spinal cord into 1 mm thick transverse sections (Figure 3E). Dispose of all medium from the proximal compartment of the MFC.
   4. Pick up a single spinal cord explant with a pipette in a total volume of 4 µL. Inject the explant as close as possible to the cave and draw out any excessive liquid from the proximal well via the lateral outlets (1 mm punches). The explants should be sucked into the proximal channel.
   5. Slowly add 150 µL of spinal cord explant medium (SCEX, see Table of Materials and Table 3) to the proximal well.
   6. Spinal cord explant culture maintenance
      1. Add SCEX medium in the proximal compartment, and rich SCEX medium (SCEX with 50 ng/mL of BDNF and GDNF) in the distal compartment. Maintain a volume gradient of at least 15 µL per well between the distal wells (higher volume) and the proximal well.
      2. Refresh the medium every 2 days. It can take the axons up to 3-5 days to cross distally.

3. Axonal transport (Figure 4A)

1. Labeling of mitochondria and acidic compartments
   1. Prepare fresh SCEX medium (or CNB for dissociated MN) containing 100 nM Mitotracker Deep Red FM and 100 nM LysoTracker Red. Incubate for 30-60 min at 37 °C. Other colors can be used as long as their fluorophores do not overlap.
   2. Wash 3x with warm CNB/SCEX medium. The plates are ready for imaging.
2. Live imaging
   1. Acquire 100 time-lapse image series of axonal transport at 3 s intervals, with a total of 5 min per movie.
      NOTE: The imaging system used in this study included an inverted microscope equipped with spinning disc confocal, controlled via propriety cell imaging software, 60x oil lens, NA = 1.4, and an EMCCD camera. Movies were acquired in a controlled environment at 37 °C and 5% CO₂.
      NOTE: Longer or shorter time-lapse movies can be imaged, dependent on the experiment. Even overnight movies can be recorded if needed. However, it is critical to try and reduce the exposure time and laser power, as well as the number of total images, to decrease phototoxicity and bleaching during movie acquisition.

4. Image analysis (Figures 4-5)

1. Analysis of particle transport distribution and density using kymograph analysis
   1. Open the file in FIJI. Separate channels pressing Image | Stacks | Tools | Deinterleave.
   2. Set image properties by pressing Image | Properties.
   3. Choose the Segmented Line Tool by right-clicking the Line Icon. Set Width to 8-10 by double-clicking the Line Icon. Be consistent with the same line width throughout the entire analysis.
   4. Mark a segmented line following the axon path from distal to proximal. Double-click to stop the line marking.
   5. Click t to add a new line region of interest (ROI) to the ROI Manager. Add this to a spreadsheet analysis table.
   6. Click m to measure area and length of the axon. Add this to the analysis table.
   7. Generate a kymograph by clicking Plugins | KymoToolBox | Draw Kymo. Alternatively, other kymograph generation plugins available can be used.
   8. Manually count moving (retrograde or anterograde) and nonmoving particles and add them to the table in the correct column.
      NOTE: Particles are classified as moving anterograde (i.e., moving left in the kymograph) or retrograde (i.e., moving right) if their displacement is higher than 10 µm in the specific direction. It is possible to measure the displacement simply by marking a horizontal line with the Line Icon and pressing m. Immobile particles or those that do not meet the displacement criteria are defined as nonmoving (Figure 4B).
2. Single particle tracking: Manual tracking
   1. Download the manual tracking plugin for FIJI software (developed by Fabrice P. Cordelière) from http://rsb.info.nih.gov/ij/plugins/track/track.html
   2. Open the file in FIJI/ImageJ. Use the Rotate option to align the MFC grooves horizontally.
   3. To improve the signal-to-noise ratio if needed, click Process | Subtract Background.
   4. Open the Manual Tracking plugin. Set the parameters (e.g., pixel size, time interval, etc.) according to the specific microscope used for imaging. For the results shown here, the microscope and lens the ratio was 0.239 µm/pixel and the frame interval was 3 s.
   5. Obtain the tracks X and Y coordinates and save the results by copying the text to spreadsheet.
   6. Analyze multiple-channel movies by clicking Image | Color | Merge Channels to merge the channels and then track only colocalized puncta.
3. Single particle tracking: Semi-automated tracking (Figure 5)
   1. Open the analysis software. This study used Bitplane Imaris software version 8.4.1.
   2. Switch to Surpass in the top menu.
   3. Click Image Processing | Swap Time and Z | Ok.
   4. Click Edit | Image Properties (Ctrl+I) | Geometry | Voxel Size Row. Set Image Properties according to the microscopy setup used.
      NOTE: For the data displayed here, the microscope and lens ratio was 0.239 µm/pixel.
   5. Click All Equidistant and change the Time Interval. For example, use 3 s as the interval. Click the Reset button on the right bottom or click Ctrl+B.
   6. Add a layer of Spots in the top left by clicking an Icon of Small Yellow Spots. At the bottom left a new Menu for Editing the Spots is opened.
   7. Press the Right Blue Arrow until the spot detection starts.
      NOTE: It is important to filter out some of the dots using the filter on the bottom left of the window. Check the movie a few times to see that a sufficient number of dots is selected.
   8. Verify or configure the parameters to fit the experimental needs. For example, Max Distance = 12 µm (The maximal allowed distance between two distinct spots to still include them in the same single track); Max Gap Size = 1 (The number of frames that a track is allowed to miss and still considered one track).
   9. Click Settings and then Track Style = Off, Points = Sphere.
   10. Using the Filter Bar, choose different filters for adjustment. For example, in the data supplied here, Track Duration = 9 removed all tracks with fewer than 3 frames. When all the parameters are set, click the Right Green Arrow. Further editing is not possible after this step.
   11. Click the Small Pencil with Dots to manually edit all the tracks.
   12. View the movie (Figure 5B). If an error occurs, there are several possible options:
       1. To disconnect a track, click the Object option and choose the two spots that need to be disconnected holding Ctrl, and choose Disconnect.
       2. To connect a track, click the Object option, choose the two spots that need to be connected holding Ctrl and choose Connect.
       3. To delete a track or spot, with the right option (Track/Object), switch to the screen with the Regular Pencil Icon, and choose Delete.
       4. To add spots manually, switch to the Regular Pencil Icon Screen. At the bottom of the screen there is a Manual Tracking Mark. Make sure the Auto-Connect Checkbox is V. On the movie itself, in order to add a spot, hold the Shift button and Left-Click.
   13. To add a spot to an existing track, choose a desired track (yellow) and frame, switch to the Regular Pencil Icon Screen and add a spot manually. When the entire movie is finished, switch to the Icon that Looks Like a Red Graph (Statistics). It is possible to edit the analyzed parameters later. To edit, on the bottom left of the screen, press the Swedish Key Icon. For example: Position X, Position Y.
   14. Press on the Icon that Looks Like Several Floppy Disks, Export All Statistics.
       NOTE: The spreadsheet output can be either handled directly or further analyzed using the published code⁹ used for this analysis, which will be shared upon demand. The following parameters are extracted from the analysis: Speed, Track Displacement, Run Length, Velocity (including Directionality), Stop Count, Average Stop Duration, Alpha, Direction Changes, and Instantaneous Velocity. A detailed explanation of each parameter is described in Gluska et al.⁹.

Representative Results

Following the described protocol, mouse embryonic HB9::GFP spinal cord explants were cultured in MFC (Figure 4A). Explants were grown for 7 days, when axons fully crossed into the distal compartment. Mitotracker Deep Red and Lysotracker Red dyes were added to the distal and proximal compartments in order to label the mitochondria and acidic compartments (Figure 4C). Axons in the distal grooves were imaged, and the movies were analyzed as follows: First, we compared the general movement distribution using kymograph analysis (Figures 4B,D). This analysis revealed a bias in the retrograde direction only in acidic compartments (nonmoving = 77.1 ± 9.5%; retrograde = 16.9 ± 8.3%, anterograde = 6 ± 5%; Figure 4E) but not in mitochondrial transport (nonmoving = 83.4 ± 6.8%; retrograde = 10.5 ± 6.9%; anterograde = 6.8 ± 5.1%; Figure 4F). Kymograph analysis was used to quantify the particle density, revealing a higher number of mitochondrial particles compared to acidic compartments in HB9::GFP spinal cord explant axons (Mitochondria = 0.46 ± 0.13; Acidic compartments = 0.3 ± 0.07 particles/µm axon, Figure 4G).

Next, single particle transport analysis was conducted using semiautomated software followed by in-house code (Figure 5A-B). This analysis revealed that despite having similar particle velocity in general (Figure 5C), when observing the distribution of velocities (Figure 5D) only acidic compartments but not mitochondria displayed a bias towards retrograde movement.

Figure 1: Silicone mold preparation. Schematic drawing describing the procedure of chlorotrimethylsilane wafer cleansing. (A) First, 50 mL of liquid nitrogen were added to an appropriate container. Working in a chemical hood, a syringe and needle were used to draw 8 mL of liquid nitrogen. The entire content of the syringe was injected into the chlorotrimethylsilane bottle. The bottle was turned with the cap facing down and 8 mL of chlorotrimethylsilane were drawn back. (B) Chlorotrimethylsilane spread in the container (not directly on the wafer). The container needs to be closed, followed by 5 min incubation for each mold. (C) Liquid PDMS was poured into each wafer up to the desired height. (D) All plates were placed together inside a vacuum desiccator for 2 h, followed by 3 h-overnight in a 70 °C oven. Please click here to view a larger version of this figure.

Figure 2: MFC specialized design. (A) Polymerized PDMS template taken out of the mold using a metal scalpel. (B) Depending on the experimental setup either 6 mm, 7 mm, or 1 mm punchers were used for punching the PDMS templates. (C) For explant culture in the MFC, 7 mm and 1 mm punchers were used, and a 20 G syringe was utilized for making "caves" for easy explant insertion. (D) For dissociated MN culture MFC, a 6 mm puncher was used to create four wells at the channel edges. (E-F) Illustrations of the final MFC shapes described in C and D, respectively. Please click here to view a larger version of this figure.

Figure 3: Neuronal culture. (A) E12.5 mouse embryo was placed in position after the head, tail, and skin were removed in order to expose the neural tube. (B) Dissection of the whole spinal cord. (C) Using gentle forceps, the meninges was peeled away from the spinal cord. (D) Left panel: Removal of the spinal cord lateral segments from the ventral spinal cord to yield better MN purification. Right Panel: Representative image of dissociated MN culture in the MFC. HB9::GFP axons crossed to the distal compartment (green). Hoechst staining indicates neuronal nuclei (blue). (E) Spinal cord explants generated by dissecting 1 mm thick transverse sections of the ventral spinal cord. Representative image of HB9::GFP (green) spinal cord explant axons in an MFC. Please click here to view a larger version of this figure.

Figure 4: Axonal transport of mitochondria and acidic compartment in MNs. (A) Illustration of the axonal transport essay. Lysotracker Red and Mitotracker Deep Red were added to both the proximal and distal compartments of the MFC, containing HB9::GFP ventral spinal cord explant. (B) Kymograph analysis. Moving particles were defined as moving anterograde or retrograde following displacement of more than 10 µm in that direction. Rotating or immobile particles were counted as nonmoving. Scale bar = 10 µm. (C) First frame of a time-lapse movie displaying primary HB9::GFP mouse spinal cord explant axons dyed with Lysotracker red to tag acidic compartments and Mitotracker Deep Red to tag mitochondria. Scale bar = 10 µm. (D) Representative kymographs displaying a typical axonal movement of acidic compartments and mitochondria. Scale bar = 10 µm. (E) Kymograph analysis of mitochondrial axonal transport, ****p < 0.0001, Anova with Holm-Sidak correction (n = 77 axons). Scale bar = 10 µm (F) Kymograph analysis of acidic compartment axonal transport, ** p < 0.01, ****p < 0.0001, Anova with Holm-Sidak correction (n = 77 axons). (G) Axonal particle density analysis of mitochondria and acidic compartments, ****p < 0.0001, Student's t-test (n = 77 axons). Error bars represent values with SD. Please click here to view a larger version of this figure.

Figure 5: Semi-automated single particle analysis to measure organelle velocity. (A) Schematic workflow for semiautomated single particle transport analysis. (B) Ventral spinal cord explant axons were analyzed for single particle tracking. The analysis software is capable of tracking single axonal particles in time-lapse movies, as indicated for mitochondria (yellow dots, upper panel) and acidic compartments (green dots, lower panel). (C) The average velocity did not change between mitochondria and acidic compartments, Mann Whitney test (n = 417 mitochondria, n = 371 acidic compartments). Error bars represent values with SD. (D) Distribution of mitochondrial and acidic compartments retrograde and anterograde velocities. Please click here to view a larger version of this figure.

Complete Neurobasal Medium – for 50mL
Ingredient	Volume	Concentration
Neurobasal	47mL
B27	1 mL	2%
Horse serum	1 mL	2%
P/S	0.5 mL	1%
L-Glutamine (Glutamax)	0.5 mL	1%
Beta-Mercaptoethanol (50mM)	25 µL	25µM
BDNF (10ug/mL)	5 µL	1ng/mL
GDNF (10ug/mL)	5 µL	1ng/mL
CNTF (10ug/mL)	2.5 µL	0.5ng/mL

Table 1: Recipe for preparation of complete neurobasal (CNB) solution.

Optiprep Solution – for 10mL
Ingredient	Volume	Concentration
DDW	5.27 mL
Density Gradient Medium (Optiprep) 60%	1.73 mL	10.4% (w/v)
Tricine 100mM	1 mL	2%
Glucose 20% (w/v)	2 mL	2%

Table 2: Recipe for preparation of density gradient medium solution.

Spinal Cord Explant Medium (SCX) – for 20mL
Ingredient	Volume	Concentration
Neurobasal	19.5 mL
B27	200 µL	2%
P/S	100 µL	1%
L-Glutamine (Glutamax)	100 µL	1%
BDNF	50 µL	25ng/mL

Table 3: Recipe for preparation of spinal cord explant (SCEX) solution.

MN Culture	Spinal Cord Explants
Longer Procedure prior to plating	Short procedure – Dissect & Plate
Extra caution needed for plating in MFC	Easier to plate in the MFC
High concentration of motor neurons with no glial cells – more accurate	Presence of glial cells and other neuronal types – more physiological
Unlimited manipulation possibilities on both Soma and axons	Limited manipulation possibilities on cell bodies
Easy immunostaining for both cell bodies and axons	Low efficiency during immunostaining of cell bodies within the explant.
High efficiency of viral infection	Very low efficiency of viral infection

Table 4: Comparison between spinal cord explants and dissociated MN culture based on perimeters of speed, feasibility, glial presence, manipulation possibilities, immunostaining, and viral infection.

Discussion

In this protocol, we describe a system to track axonal transport of mitochondria and acidic compartments in motor neurons. This simplified in vitro platform allows precise control, monitoring, and manipulation of subcellular neuronal compartments, enabling experimental analysis of motor neuron local functions. This protocol can be useful for studying MN diseases such as ALS, to focus on understanding the underlying mechanism of axonal transport dysfunction in the disease^10,16. Moreover, this system can also be applied for studying transport of trophic factors^9,16, microRNA¹⁰, mRNA⁸, and labeled proteins in healthy and diseased MNs or in other neurons, such as sensory⁹ or sympathetic axons¹⁴. A similar method can also be applied to study organelle transport in a coculture system, such as MNs cultured with skeletal muscle cells¹² or sympathetic neurons cocultured with cardiomyocytes¹⁴. The MFC system can be utilized to generate active NMJs^17,18 and study the effect of synapse formation on axonal transport¹⁶.

This protocol has several advantages compared to other protocols for culturing of neurons in MFC and using live imaging to analyze axonal transport: 1) MFCs are commercially available by several manufacturers. However, self-manufacturing of the MFCs is extremely cost effective compared to the commercial alternative. A single PDMS wafer yields four large MFCs or nine small ones at the minimal expense of several US dollars for the PDMS resin itself. 2) The self-manufactured MFCs can be modified to answer specific experimental needs (e.g., changing the size and the location of the wells, increasing the thickness of the MFC PDMS). 3) The MFCs can be irreversibly attached to a plate (via plasma bonding) but can also be recycled multiple times to reduce the possibility of contamination. 4) The PDMS MFCs are transparent, making them ideal for live imaging by having reduced background, which is critical for axonal transport assays where signal-to-noise distinction could be a limiting factor. 5) Spinal cord explant culture is very efficient and fast. One embryonic spinal cord can yield up to 30 explants, enough for 10 MFCs. This can help to save time and materials, and ensures that even a pregnant mouse with few embryos can produce a successful experiment. See a detailed comparison of spinal cord explant and dissociated MN culture in Table 4.

This protocol is relatively simple and inexpensive. However, it requires expertise in several technical matters. The manufacturing and handling of the MFC needs to be accurate and gentle, to avoid structural defects and creation of air bubbles in the obligatory vacuum step (1.1.10), for example. During the optional vacuum step (1.6.2) it is important to clear all the air, or it will block axonal crossing, but also keep the vacuum short to avoid detachment of the MFC from the glass. Pay close attention to the dissection protocol steps, as it is important to properly remove the meninges and dorsal horns, as well as to try and cut the pieces to the right size in order to insert them to the MFC's cave without using physical force. Any excessive physical force applied to the MFC can easily detach the grooves or the entire MFC, thus the procedure should be done gently. Culturing of MNs and plating them in the MFC needs to be relatively swift, as MNs tend to aggregate and lose viability quickly under stress conditions such as the low medium volume of the plating step. The plating in the MFC should be done in low volume to allow proper attachment of the neurons in the MFC channel, and after no more than 30 min warmed medium should be added very gently to prevent detachment of the MNs.

After several days in culture, and once axons have extended to the distal compartment, the cultures are ready to undergo organelle staining followed by acquisition of axonal transport movies and finally a specific procedure for image analysis. The analysis can be performed either by single particle tracking over time, or by using an automated or semiautomated tracking algorithm. In our experience, automated methods are less time-consuming, but have several disadvantages. Mainly, automated tracking reliability is decreased, with crowded axons that cross paths. Furthermore, automated algorithms have a decreased ability to distinguish between overlapping particles. Consequently, performing manual or semiautomated tracking is recommended. In general, semiautomated tracking is preferable as it is less time-consuming, but manual tracking can be a good option when there is no available paid software. A thorough comparison between manual and automatic analysis can be found in Gluska et al.⁹.

In conclusion, adopting this simple protocol can yield important data for the study of axonal transport of various cellular components. It can be adapted to fit a diversity of imaging setups and neuronal subtypes, as well as to cocultures of neurons with other cells. It also allows distinct spatial pharmacological manipulation of either cell bodies or axons in order to understand basic mechanisms of neuronal health and to improve drug development for diseases with altered axonal transport.

Materials

Name	Company	Catalog Number	Comments
35mm Fluodish – glass bottom dish	World Precision Instruments WPI	FD35-100
50mm Fluodish – glass bottom dish	World Precision Instruments WPI	FD5040-100
Andor iXon DU-897 EMCCD camera	Andor
ARA-C (Cytosine β-D-arabinofuranoside)	Sigma-Aldrich	C1768	stock of 2mM in filtered DDW
B-27 Supplement (50X)	Thermo Fisher	17504044
BDNF	Alomone Labs	B-250	Dilute to 10 µg/mL in filtered ddw with 0.01% BSA)
Biopsy punch 1.25mm	World Precision Instruments WPI	504530	For preperation of large MFC
Biopsy punch 6mm	World Precision Instruments WPI	504533	For preperation of small MFC
Biopsy punch 7mm	World Precision Instruments WPI	504534	For preperation of large MFC
Bitplane Imaris software - version 8.4.1	Imaris
Bovine Serum Albumine (BSA)	Sigma-Aldrich	#A3311-100G	5% w/v in ddw
Chlorotrimetylsilane	Sigma-Aldrich	#386529-100ML
CNTF	Alomone Labs	C-240	Dilute to 10 µg/mL in filtered ddw with 0.01% BSA)
Density Gradient Medium - Optiprep	Sigma-Aldrich	D1556
Deoxyribonuclease I (DNAse) from bovine pancreas	Sigma-Aldrich	DN-25	stock 10mg/mL in neurobasal
Dow Corning High-vacuum silicone grease	Sigma-Aldrich	Z273554-1EA	For epoxy mold preperation
DPBS 10X	Thermo Fisher	#14200-067	dilute 1:10 in ddw
Dumont fine forceps #55 0.05 × 0.02 mm	F.S.T	1125520
Epoxy Hardener	Trias Chem S.R.L	IPE 743	For epoxy mold preperation
Epoxy Resin	Trias Chem S.R.L	RP 026UV	For epoxy mold preperation
FIJI software	ImageJ
GDNF	Alomone Labs	G-240	Dilute to 10 µg/mL in filtered ddw with 0.01% BSA)
Glutamax 100X	Thermo Fisher	#35050-038
HB9:GFP mice strain	Jackson Laboratories	005029
HBSS 10X	Thermo Fisher	#14185-045	Dilute 1:10 in ddw with addition of 1% P/S and filter
iQ software	Andor
Iris scissors, curved, 10 cm	AS Medizintechnik	11-441-10
Iris scissors, straight, 9 cm	AS Medizintechnik	11-440-09
Laminin	Sigma-Aldrich	#L-2020
Leibovitz's L-15 Medium	Thermo Fisher	11415064
LysoTracker Red	Thermo Fisher	L7528
Mitotracker Deep-Red FM	Thermo Fisher	M22426
Neurobasal medium	Thermo Fisher	21103049
Nikon Eclipse Ti micorscope	Nikon
Penicillin-Streptomycin (P/S) Solution	Biological Industries	03-031-1
Poly-L-Ornithin (PLO)	Sigma-Aldrich	#P8638	Dilute 1:1000 in flitered 1X PBS
Sylgard 184 silicone elastomer kit	DOW Corning Corporation	#3097358-1004
Trypsin from bovine pancreas	Sigma-Aldrich	T1426	stock 25 mg/mL in 1XPBS
Vannas spring microdissection scissors, 3 mm blade	F.S.T	15000-00
Yokogawa CSU X-1	Yokogawa