Protocol to visualize distinct motoneuron pools in adult zebrafish via injection of retrograde tracers

Summary In adult zebrafish, slow, intermediate, and fast muscle fibers occupy distinct regions of the axial muscle, allowing the use of retrograde tracers for selective targeting of the motoneurons (MNs) innervating them. Here, we describe a protocol to label distinct MN pools and tissue processing for visualization with confocal laser microscopy. We outline the different steps for selective labeling of primary and secondary MNs together with spinal cord fixation, isolation, mounting, and imaging. For complete details on the use and execution of this protocol, please refer to Pallucchi et al. (2022)1 and Ampatzis et al. (2013).2


BEFORE YOU BEGIN
In zebrafish, motoneurons can be divided into two main types, primary and secondary, depending on the time of their birth, soma size and position in the spinal cord. Primary motoneurons (pMN) are born first, they have large cell bodies and axons, and occupy a dorsal position in the motor column. Secondary motoneurons (sMN) are born later, they have comparatively smaller cell bodies and axons and reside in the ventral part of the motor column. pMNs are fast and innervate the large white muscles. In adult zebrafish, sMNs are heterogenous and comprise three functional subtypes each selectively innervating either fast, intermediate or slow muscles. This classification results in four distinct motoneuron pools. [2][3][4][5][6] These pools are organized somatotopically with a defined location in the motor column that relates to the muscle type they innervate. Slow motoneurons are located ventro-laterally, intermediate motoneurons have a ventro-medial location and fast motoneurons are located medio-dorsally in the spinal motor column.
Zebrafish transgenic lines labeling MNs are available, but these do not discriminate among the different MN types due to the lack of specific genetic markers for slow, intermediate and fast MNs. For the analysis of motor circuits, it is necessary to identify and access each MN pool separately, this can only be done through retrograde labeling from specific muscles.
In zebrafish, fast, intermediate and slow muscles are spatially segregated, in contrast to mammals where they are intermingled. The zebrafish slow red muscle fibers occupy a thin lateral strip, the intermediate (or pink) form a wedge-shaped area around the horizontal septum, and the white fast fibers represent the large medial portion of the myotome. [7][8][9] The spatial segregation of muscles fibers in adult zebrafish allows for precise targeting of each muscle type by injecting dextran tracers to retrogradely label the MNs innervating them. [2][3][4]6 Single or multiple MN pools can be labeled simultaneously over several segments in a single fish by targeting dye injections to the muscle they innervate. The specific labeling of the different MN pools enables their identification and allows for studying their activity, the synaptic input they receive, their morphology and their transmitter phenotypes using immunohistochemical protocols. 1,2 Institutional permissions Zebrafish (Danio rerio) were raised and housed in the Karolinska Institutet, Comparative Medicine Biomedicum (KM-B) animal facility according to established procedures. All experimental procedures followed the EU guidelines and were approved by the Animal Research Ethical Committee in Stockholm. Therefore, researchers must acquire authorization to perform animal work from their relevant institutions before using this protocol.
Prepare transgenic or wild-type zebrafish The procedure described here can be used in wild-type or transgenic zebrafish lines. When using a transgenic line with a fluorescent reporter, researchers should use a retrograde dye for the labeling of motoneurons that will not interfere with the fluorophores present in the line.
Raise zebrafish (Danio rerio) in the animal facility according to established procedures. Adult animals of either sex can be used, and we recommend using 8-12 week old fish (length: $15-20 mm) because the fish is big enough to distinguish the muscle fibers, and the spinal cord is not too thick for the whole-mount confocal imaging.

Prepare injection pin
Prepare the injection pipet by embedding a minutien pin (12.5 mm diameter) into a glass Pasteur pipet and seal with wax ( Figure 1A). The pin can be sharpened using polishing film for optimal injection with minimal tissue damage.

Stereo microscope
We use an Olympus SZX10 stereo microscope configured with an ergonomic tilting trinocular viewing head with WHN103/22 FOV eyepieces. The microscope features a 10:1 zoom range and

OPEN ACCESS
the dual position nosepiece has DF PLAPO 13 auxiliary lenses, therefore the magnification range is 6.33 -633. The focus mount provides coarse and fine focus adjustments.

Confocal microscope
We use a Zeiss LSM980-Airy2 confocal microscope equipped with a 203 air/dry objective (NA 0.8), a 403 water objective (NA 1.2) and an Airy detector2 for super-resolution capacity with three laser lines (488, 561 and 640).
Alternatives: a fluorescent microscope can be used to visualize labeled motoneurons. Confocal microscopy is not necessary, but it provides a better resolution and image quality.

Image analysis
We use ZEN (blue edition) from Zeiss to process and analyze the acquired confocal images, but other software can be used.

KEY RESOURCES TABLE
CRITICAL: Paraformaldehyde has an acute toxicity by inhalation, ingestion, and dermal exposure. It is both corrosive to the skin and eyes and is a suspected carcinogen. Wear a lab coat, nitrile exam gloves and handle PFA under a ventilated hood. REAGENT

MATERIALS AND EQUIPMENT
CRITICAL: overexposure to MS-222 in humans may cause skin, eye and respiratory irritation. Wear a lab coat, nitrile exam gloves and safety glasses when handling powder.
Donkey serum Prepare 0.5 mL aliquots and store at À20 C. Stable for years.

Primary and secondary antibody mix
Prepare the antibodies mix in Triton X-100 in PBS following the manufacturer's recommended dilutions. To be prepared and used fresh.

STEP-BY-STEP METHOD DETAILS
The whole protocol must be performed within 10 days to ensure the best staining. The fluorescence should be detectable at least for a month, but it loses intensity over time because dextran dyes degrade.

OPEN ACCESS
This section describes how to retrogradely label motoneurons by injecting dextran dyes into zebrafish muscles.
1. Prepare the injection pin. a. Place some crystals (approx. 0.1 mg) of the fluorescent dextran dye on a clean glass slide. b. Add 2 mL of distilled water to the dye, mix well and then let the mix dry for some minutes. c. Dip the tungsten pin when the dye mix has the consistency of a paste. Avoid letting the mix desiccate and add some microliters of water if needed. 2. Prepare the fish (Methods video S1).
a. Prepare the anesthetic solution by diluting the MS-222 stock solution to a final concentration of 0.03% in fish tank water. b. Place the fish in the anesthetic solution for about 30-60 s until deep anesthesia is reached.
Note: In zebrafish, deep anesthesia is characterized by a lack of movement, few or no opercular movement, and a lack of response to a soft touch.
c. Place the fish laterally on a wet paper tissue under a stereo microscope. d. Secure the fish position by covering the head and the tail with similar paper tissue ( Figure 1B).

Inject the dye in selective muscle fibers (Methods video S2).
a. Dry the surface of the fish and scrape away a few scales using forceps to gain better access to the targeted muscle. b. Inject the dye-soaked pin in the target muscle fibers (Figures 1C and 1D). If needed, soak the pins in the dye multiple times and repeat the injection. c. Let the fish recover from anesthesia in fresh fish tank water. d. Keep the animal in a regular fish tank for at least 2 h or overnight (14-20 h) to allow for retrograde transport of the tracer. 4. Repeat the steps 1-3 on the desired number of fish depending on the experimental design.
CRITICAL: Commonly used dyes are rhodamine-dextran 3,000 MW (Thermo Fisher) or Alexa Fluor 488-dextran 10,000 MW (Thermo Fisher). Dextran dyes are retrogradely transported from the severed axon to the cell soma and the speed of their transport is related to their MW. Both 3,000 and 10,000 MW dextrans reach the MN soma after less than 2 h. CRITICAL: Make sure to carry out the back labeling procedure within 5-10 min, since prolonged anesthesia may be lethal for the animal. For this reason, it is recommended to carry out the injections on one fish at a time. To help the recovery, fish water can be kept at a temperature of 26 C-28 C. Note: High MS-222 concentration will keep the fish under deep anesthesia during dissection.

Tissue fixation and spinal cord isolation
b. Place the fish on a paper towel for a few seconds to remove the excess of water. c. Glue the fish laterally on a glass Petri dish using superglue. Cover the head and the tail with the glue and immediately submerge the fish in 13 PBS. d. Using surgical tools, remove internal organs, the skin first and then the muscles to expose the vertebrae (Figure 2A). e. Discard the 13 PBS and add 4% PFA in the dish and keep it at 4 C 2-24 h to allow for optimal fixation. 6. Spinal cord isolation.
a. Discard the 4% PFA, rinse once with 13 PBS and fill the dish with 13 PBS. b. Perform the spinal cord isolation using forceps under a stereomicroscope. i. Remove the neural spines (ns) and vertebrae arches to facilitate the dissection ( Figure 2B). ii. Remove the hemal spines (hs) and the centrum (c) to expose the spinal cord ( Figure 2C).
Note: Once the vertebrae are removed ( Figure 2D), the spinal cord can be transferred to a multiwell plate filled with 13 PBS and stored at 4 C until further processing. It is recommended to perform the next steps within 48 h to avoid degradation of dextran dyes.

Tissue labeling and mounting
Timing: 4-5 days (for step 7) This section describes the tissue immunolabeling (optional) and the spinal cord mounting on microscope slides.
7. Optional: Immunohistochemistry. Researchers can combine retrograde labeling of MNs with immunohistochemistry, which can be carried out in a multiwell plate (48-well for example) as follows. a. Wash the tissue extensively with 13 PBS (3 3 15 min). Note: Primary antibody incubation is variable. Overnight (14-20 h) is enough for highly expressed proteins (like reporters), but a range of 36-96 h works better for endogenous proteins.
Note: The secondary antibody incubation can be shortened to 2-6 h at room temperature (18 C-25 C); however, this is not recommended.
8. Tissue mounting. a. Cut a piece of parafilm to fit the microscope glass slide and cut out the middle of the parafilm where the spinal cord will be placed. b. Press the parafilm on the slide and warm it up gently on a heating plate to avoid that it detaches from the glass.
Note: The parafilm frame on the glass slide is necessary to preserve the spinal cord shape and avoid compression of the tissue.
c. Let the slide cool down before placing the spinal cord ( Figure 2E). d. Position gently the spinal cord in the desired orientation on the slide using forceps ( Figure 2F). e. Add non-hardening mounting media. Coverslip the sample.
Note: While most microscope mounting media are suitable, hardening media are not recommended because they will alter the shape of the tissue.

Confocal microscopy
Timing: 1-4 h (for step 9) This section describe the image acquisition process at the confocal microscope.
9. Acquisition of confocal images. a. Prepare the Zeiss LSM980-Airy2 confocal microscope equipped with a 203 (NA 1.1) objective. b. Focus on the spinal cord using the fluorescence or the bright field. c. Move to the acquisition mode and set up the desired channels. d. Acquire the part of the spinal cord where MNs were back labeled. If the protocol has been successful, images as those shown in Figure 3A should be obtained. and 3D-3F). Slow secondary MNs are located ventro-laterally, they have soma size and dendritic arborization comparable to intermediate MNs ( Figures 3B and 3D-3F).

EXPECTED OUTCOMES
The procedure described here allows to gain access to distinct MN pools innervating specific muscle types. This enables comparing their physiological, morphological and transmitter features. 10 The retrograde labeling of MNs also affords their electrophysiological characterization in live tissue. This procedure combined with electrophysiological recordings allows for gaining detailed information on the intrinsic properties of the different motoneuron pools, their connectivity with premotor interneurons as well as the extent of their dendritic arborizations. In addition, this procedure can be used in mutant or transgenic animals modeling motoneuron diseases.

LIMITATIONS
In general, it is difficult to label all motoneurons innervating a given muscle type and therefore multiple injections should be performed at different segments and repeated in different animals. Moreover, MN staining intensity is proportional to the quantity of dye incorporated. MNs with larger axons will be strongly labeled. This protocol is applicable only to juvenile-adult zebrafish, as larvae

OPEN ACCESS
have different muscle composition and the pins used in this study are not suitable for injections in larvae.

Problem 1
Zebrafish do not recover from anesthesia after back labeling of motoneurons (related to steps 1-2-3).

Potential solution
Adjust the anesthetic concentration. Make sure to prepare fresh MS-222 at the right concentration. The solution can be stored at 4 C for at least one month but should be reprepared as necessary.
Do not keep zebrafish under deep anesthesia for more than 5-10 min. Make sure you inject the dye only in the muscle to avoid organ lesions.

Problem 2
Muscle fibers are not easily distinguishable (related to step 3).

Potential solution
Please use Figure 1 as a reference and gently remove some scales to expose the muscle fibers. Use the recommended size and age of zebrafish to facilitate the identification of the different muscle fiber types. Illumination at an appropriate angle usually helps identify the horizontal septum.

Problem 3
Unsuccessful spinal cord isolation (related to steps 5 and 6).

Potential solution
The spinal cord may break during the dissection, so researchers should make sure that the tissue has been properly fixed. Generally, a 2-h fixation with 4% PFA is enough, but it is recommended to use a longer fixation to facilitate the dissection. The procedure requires some training for achieving optimal results.

Problem 4
No or very few motoneurons are visible (related to step 9).

Potential solution
The concentration of the dye is important when injecting it into muscles. Too low concentration results in a low fluorescent signal and too high concentration could damage MNs. Tissue should also be processed according to the time scale described because dextran dyes deteriorate over time.

Potential solution
Adjust the concentration of antibodies according to the manufacturer's recommendation. Increase the incubation time since whole spinal cord requires a longer time for reaching optimal antibody penetration. For primary antibodies, overnight incubation (14-20 h) is enough for highly expressed proteins (like reporters), but a range of 36-96 h works better for endogenous proteins. For secondary antibodies, overnight incubation (14-20 h) is sufficient. Some primary antibodies may require a short tissue fixation in order to bind to the antigens. In this case, perform a 2-h spinal cord fixation with 4% PFA. Alternative fixation methods can be used.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Abdel El Manira (abdel.elmanira@ki.se).

Materials availability
This study did not generate new unique reagents.

Data and code availability
This study did not generate/analyze datasets or code.