Dissecting Cell-Autonomous Function of Fragile X Mental Retardation Protein in an Auditory Circuit by In Ovo Electroporation

Summary

Using in ovo electroporation, we devised a method to selectively transfect the auditory inner ear and cochlear nucleus in chicken embryos to achieve a cell-group-specific knockdown of fragile X mental retardation protein during discrete periods of circuit assembly.

Abstract

Fragile X mental retardation protein (FMRP) is an mRNA-binding protein that regulates local protein translation. FMRP loss or dysfunction leads to aberrant neuronal and synaptic activities in fragile X syndrome (FXS), which is characterized by intellectual disability, sensory abnormalities, and social communication problems. Studies of FMRP function and FXS pathogenesis have primarily been conducted with Fmr1 (the gene encoding FMRP) knockout in transgenic animals. Here we report an in vivo method for determining the cell-autonomous function of FMRP during the period of circuit assembly and synaptic formation using chicken embryos. This method employs stage-, site-, and direction-specific electroporation of a drug-inducible vector system containing Fmr1 small hairpin RNA (shRNA) and an EGFP reporter. With this method, we achieved selective FMRP knockdown in the auditory ganglion (AG) and in one of its brainstem targets, the nucleus magnocellularis (NM), thus providing a component-specific manipulation within the AG-NM circuit. Additionally, the mosaic pattern of the transfection allows within-animal controls and neighboring neuron/fiber comparisons for enhanced reliability and sensitivity in data analyzing. The inducible vector system provides temporal control of gene editing onset to minimize accumulating developmental effects. The combination of these strategies provides an innovative tool to dissect the cell-autonomous function of FMRP in synaptic and circuit development.

Introduction

Fragile X syndrome (FXS) is a neurodevelopmental disorder characterized by intellectual disability, sensory abnormalities, and autistic behaviors. In most cases, FXS is caused by a global loss of fragile X mental retardation protein (FMRP; encoded by Fmr1 gene) starting at early embryonic stages¹. FMRP is an RNA-binding protein that is normally expressed in most neurons and glial cells in the brain, as well as in sensory organs^2,3,4. In mammalian brains, FMRP is likely associated with hundreds of mRNAs that encode proteins that are important for various neural activities⁵. Studies of conventional Fmr1 knockout animals demonstrated that FMRP expression is particularly important to the assembly and plasticity of synaptic neurotransmission⁶. Several conditional and mosaic knockout models have further demonstrated that FMRP actions and signals vary across brain regions, cell types, and synaptic sites during several developmental events including axonal projection, dendritic patterning, and synaptic plasticity^{7,8,9,10,11,12,13,14}. Acute function of FMRP in regulating synaptic transmission was studied by intracellular delivery of inhibitory FMRP antibodies or FMRP itself in brain slices or cultured neurons^15,16,17,18. These methods, however, do not offer the ability to track FMRP misexpression-induced consequences during development. Thus, developing in vivo methods to investigate the cell-autonomous functions of FMRP is in great need, and expected to help determine whether the reported anomalies in FXS patients are direct consequences of FMRP loss in the associated neurons and circuits, or secondary consequences derived from network-wide changes during development¹⁹.

The auditory brainstem of chicken embryos offers a uniquely advantageous model for in-depth functional analyses of FMRP regulation in circuit and synapse development. The easy access to embryonic chicken brains and the well-established in ovo electroporation technique for genetic manipulation have contributed greatly to our understanding of brain development at early embryonic stages. In a recently published study, this technique was combined with advanced molecular tools that allow temporal control of FMRP misexpression^20,21. Here, the methodology is advanced to induce selective manipulations of presynaptic and postsynaptic neurons separately. This method was developed in the auditory brainstem circuit. Acoustic signal is detected by hair cells in the auditory inner ear and then conveyed to the auditory ganglion (AG; also called the spiral ganglion in mammals). Bipolar neurons in the AG innervate hair cells with their peripheral processes and in turn send a central projection (the auditory nerve) to the brainstem where they terminate in two primary cochlear nuclei, the nucleus magnocellularis (NM) and the nucleus angularis (NA). Neurons in the NM are structurally and functionally comparable to the spherical bushy cells of the mammalian anteroventral cochlear nucleus. Within the NM, auditory nerve fibers (ANFs) synapse on the somata of NM neurons via the large endbulb of Held terminals²². During development, NM neurons arise from rhombomeres 5 and 6 (r5/6) in the hindbrain²³, while AG neurons are derived from neuroblasts residing in the otocyst²⁴. Here, we describe the procedure to selectively knockdown FMRP expression in the presynaptic AG neurons and in the postsynaptic NM neurons separately.

Protocol

Eggs and chicken embryos were handled with care and respect in accordance with the animal protocols approved by the Jinan University Animal Care and Use Committee.

1. Egg and plasmid preparation

1. Egg preparation
   1. Obtain fresh fertilized chicken eggs (Gallus gallus) from the South China Agricultural University and store at 16 °C before incubation. For optimal viability, set all eggs for incubation within a week of arrival.
   2. Place eggs horizontally and incubate at 38 °C for 46-48 h until Hamburger and Hamilton (HH) stage 12²⁵ for neural tube transfection, or for 54-56 h until HH13 for otocyst transfection. Because the animal pole of the egg is lighter than the vegetal pole, horizontal placement positions the embryo on the top of the egg for easy manipulation. Use caution to keep the egg horizontal in this and all following steps.
   3. The location of the embryo appears as a dark area on the eggshell when casting light from the bottom of the egg using a flashlight. At the desired HH stages, use this flashlight method to mark the location of the embryo on the eggshell with pencil.
   4. Wipe the eggs with gauze containing 75% ethanol. Drill a hole in the pointy end of the eggs using the tip of scissors and remove 2 mL of albumin with an 18 G needle syringe. Ensure the hole is only large enough to allow needle insertion. Wipe away any leaking albumin with gauze and seal the hole with clear tape.
   5. To minimize cracks and to prevent falling shell debris during windowing, cover the top of eggs with clear tape, centering on the pencil-marked dark area.
2. Plasmid DNA extraction
   1. Clone chicken Fmr1 shRNA using a transposon-based Tet-on system as described previously²⁰. This system contains three plasmids: pCAGGS-T2TP, pT2K-CAGGS-rtTA-M2, and pT2K-BI-TRE-EGFP-Fmr1 shRNA, providing a drug-inducible vector system by which the expression of Fmr1 shRNA and EGFP is silenced after transfection and can be turned on by doxycycline (Dox) induction.
      ​NOTE: These plasmids are not currently available in a repository. The authors agree to provide the plasmids upon reasonable request.
   2. Extract plasmid DNAs using an endotoxin-free preparation kit according to manufacturer's instruction and precipitate by adding 1/15 volume of 7.5 M sodium acetate and 1 volume of 100% isopropanol.
   3. Precipitate plasmids at -20 °C overnight and centrifuge at 13,000 x g for 10 min. Dissolve the pellet with the Tris-EDTA buffer provided in the kit to a final concentration of 4-5 µg/µL. Plasmids can be stored at -20 °C for 12 months without reduced transfection efficiency.
   4. On the day of electroporation, thoroughly mix 2 µL of each plasmid in a sterile centrifuge tube using a pipette tip. The resultant 6 µL volume of the plasmid mixture is enough to electroporate three dozen eggs. To help visualize the plasmid during electroporation, add 1 µL of 0.1% fast green (prepared in sterile ddH₂O) for every 6 µL of DNA mixture²⁶, which turns the plasmid mixture blue.

2. In ovo electroporation

1. Egg windowing and embryo identification
   1. On the day of electroporation, remove the eggs from the incubator, one dozen eggs at a time. Keeping eggs out of their incubator for more than 1 h introduces developmental variations and reduces viability.
   2. Wipe all surgery tools with gauze containing 75% ethanol.
   3. Place each egg in a custom-made foam holder. Wipe the tape-covered area with 75% ethanol and cut a window (1-2 cm²) around the circumference of the pencil mark using a small pair of scissors (Figure 1A). When cutting, hold the scissors flat to avoid damaging the embryo underneath.
   4. Place the windowed egg under a stereomicroscope with a 10x eyepiece and 2x zoom and identify the embryo. Adjust the angle and brightness of the light source for better visualization.
      NOTE: It takes practice and experience to visualize the embryo and identify the neural tube at this stage.
      1. For beginners, use the following optional ink injection to facilitate embryo identification. Prepare 10% Indian ink solution in 0.01 M phosphate-buffered saline (PBS) and autoclave in advance. Fill a 1 mL syringe with the ink solution and then fit a 27G needle. Bend the needle to a 45° angle with forceps.
      2. Under the microscope, carefully poke from the edge of the area opaca and insert the needle beneath the embryo. Inject ~50 µL of ink, which will diffuse below the area pellucida for embryo visualization. The ink will form a dark background for a clear visualization of the embryo.
         NOTE: Expel the air from the syringe before insertion and injection. When skilled at visualizing, avoid the ink injection step as it reduces the survival rate.
2. Neural tube injection and electroporation
   NOTE: This procedure is performed at HH12 for transfecting NM neurons in the brainstem.
   1. Pull glass capillaries (~1 mm in diameter and 100 mm long) into pipettes using a pipette puller. Under a dissecting microscope, carefully open the tip of the capillary needle to 10-20 µm in diameter with forceps. Store the pipettes (usually 5-10 at one time) in a storage box until use.
   2. Fill a capillary pipette with 0.5-1 µL of the plasmid mixture by applying negative pressure through a rubber tube at the end of the pipette. This can be achieved by a syringe or air pump.
   3. Place the egg under the microscope so that the embryo is vertically oriented with the tail near to you (or horizontally for injecting capillary pipettes using a three-axis manipulator). Hold the capillary pipette with one hand or with the three-axis manipulator and drive the tip of the pipette to the r5/6 in a tail-to-head direction.
      NOTE: The most anterior hindbrain area is r1 and each subsequent bulge along the neural tube is an individual rhombomere. R5/6 is at the same anteroposterior level as the otocyst, which is a readily recognizable cup-like structure (Figure 1B-C).
   4. Gently poke the tip through the vitelline membrane and into the dorsal neural tube and then withdraw the pipette a bit so the tip is in the lumen of the neural tube. Inject the plasmid mixture by applying air pressure until the tinted plasmid diffuses fully into r5/6 and extends into r3 and r4.
      NOTE: It is not necessary to make a slot on the vitelline membrane for injection.
   5. Check for a successful injection, which is achieved when the blue plasmid solution rapidly diffuses down the neural tube without leaking (Figure 1B-C). When leaking occurs, the blue quickly fades.
   6. Immediately after the injection, place a platinum bipolar electrode on either side of the neural tube (Figure 1D). Deliver two pulses of 12 V and 50 ms duration with an electroporator. Observe air bubbles at the ends of the bipolar electrode, with more on the negative side.
   7. Check for successful electroporation, which is achieved when the tinted plasmid mixture enters the neural tube tissue near the positive side of the electrode. After electroporation, carefully remove the bipolar electrode.
   8. Cover the window on the eggshell with a piece of transparent film that is pre-cut into 2 in squares and sprayed with 75% ethanol. Place the egg back into its incubator.
   9. Clean the bipolar electrode by delivering 10-20 pulses of 12 V and 50 ms duration in saline before proceeding to the next egg.
3. Otocyst injection and electroporation
   NOTE: This procedure is performed at HH13 for transfecting hair cells and AG neurons in the inner ear.
   1. At HH13 (which is ~embryonic day 2.5), the embryo has turned so the right side of the head faces up. Under the microscope, place the egg so that the embryo is vertical, with the tail near to you. Hold the capillary pipette and gently poke the right otocyst in a dorsolateral direction (Figure 2A).
      NOTE: At this stage, the otocyst appears as a small circular structure on the top of the body at the same anteroposterior level as r5/6.
   2. Inject the plasmid mixture with air pressure until the otocyst is filled with blue solution²⁷. Check for a successful injection, which is achieved when the blue plasmid mixture is confined within the otocyst and doesn't leak.
   3. Immediately after the injection, place the bipolar electrode on the otocyst as depicted in Figure 2B. Position the positive and negative sides anterior and posterior to the otocyst, respectively. Deliver two pulses of 12 V and 50 ms duration with the electroporator.
   4. Check for successful electroporation, which is achieved when the blue plasmid mixture enters the tissue of the otocyst near the positive side of the electrode. After electroporation, carefully remove the bipolar electrode. Cover the window on the eggshell with a transparent film and return the egg to the incubator.

3. Administration of Dox to initiate and maintain plasmid transcription

1. Preparation of the Dox solution
   1. Measure 100 mg of Dox powder under a chemical hood and dissolve it in 100 mL of sterile PBS to make a 1 mg/mL of working solution. Filter the solution with a 0.22 µm filter and store 1 mL aliquots at -20 °C. Protect from light^20,28.
2. Administration of Dox
   1. For full-strength gene editing, at 24 h before the desired age, thaw a Dox aliquot on ice. Drop 50 µL of Dox directly on the chorioallantoic membrane of the egg using a syringe penetrating the transparent film. Seal the needle hole with transparent film or tape after injection.
   2. To maintain the knockdown effect, administer Dox every other day until tissue harvest at the desired developmental stage.

4. Tissue dissection and sectioning

1. Brainstem
   1. For embryos at embryonic stage 3 (E3) and E6, open the eggshell and cut the associated membrane with scissors. Spoon the embryo out and immerse it in 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer.
   2. For E9 embryos, open the eggshell, decapitate the embryo with scissors, and transfer the head to a silicon-bottomed plate filled with PBS. Pin the head down through the orbits and cut the top of the skull apart. Carefully place the forceps beneath the brain from both sides and elevate the entire brain from the skull with the shoulders of the forceps. Immerse the brain in 4% PFA.
   3. For E15 and E19 embryos, open the eggshell, decapitate the embryo with scissors, and place the head on a silicon-bottomed plate filled with PBS. Cut open the skin on top of the head to expose the skull.
   4. Incise through the skull vertically with a razor to separate the forebrain and the caudal brain. Immerse the caudal block in PBS, remove the top of the skull, and then take the brain off the skull.
   5. Pin the brain down through the optic lobe and separate the cerebellum and brainstem. Take care not to damage the auditory brainstem, which is located right below the cerebellum. Remove as much dura as possible from the brainstem and then immerse it in 4% PFA.
   6. Keep embryos and brain tissues in 4% PFA at 4 °C overnight except for E19 brainstems, which should be kept under the same condition for 24 h.
   7. Following fixation, dehydrate the tissue in PBS containing 30% sucrose until it settles, which usually takes 1-3 days, depending on the age.
   8. Section the brainstem (E15 and E19) at 30 µm at the coronal plane using a sliding microtome. Collect sections in PBS and store at 4 °C before immunostaining.
   9. For E3 and E6 embryos and E9 brainstems, section at 50 µm transversely. Collect sections in PBS and store at 4 °C. Mount the sections on gelatin-coated microscope slides before immunostaining. Collecting thicker sections for these ages facilitates tissue handling.
2. Auditory duct
   1. After removing the brainstems from E9, E15, and E19 embryos, the auditory duct, which is embedded in the temporal bone, is readily identifiable lying beneath the skull, and the temporal bone is easily separatable from the surrounding skull. For E9 embryos, fix the whole temporal bone in 4% PFA. For older embryos, remove the bony structure of the temporal bone to isolate the auditory duct and fix the auditory duct in 4% PFA.
   2. Keep all temporal bones and auditory ducts in 4% PFA at 4 °C overnight before tissue embedding.
   3. Perform tissue embedding as described. Prepare the embedding solution as follows: Soak 10% gelatin (from bovine skin) in cold water until the gelatin granules swell and settle (about 30 min). Heat the mixture to ~50 °C to dissolve the gelatin and add 20% sucrose into the gelatin solution and stir to dissolve. Store the gelatin-sucrose solution at 4 °C for up to a month.
   4. For use, warm the gelatin-sucrose solution at 37 °C. Soak the temporal bones/auditory ducts in the warm gelatin-sucrose solution on a 96-well plate at 37 °C until they settle, usually 30-60 min.
   5. Line the bottom of the wells of a 12-well plate with a strip of transparent film. Add a layer of warm gelatin-sucrose solution and wait until it solidifies. Transfer a temporal bone/auditory duct to each well.
   6. Implant the tissue with a second layer of warm gelatin-sucrose solution. Adjust the tissue's position so that it is in the center of the gel and the auditory duct is oriented horizontally. Carefully move the plate into a 4 °C refrigerator and wait until the gel is firm.
   7. Pull the transparent film strip to move the gel-tissue block out of the well. Trim extra gel into a square block and cut a corner of the block to identify the orientation. Wrap the gelatin block with a piece of foil, freeze on dry ice, and then store at -80 °C until sectioning.
   8. Section the block at 20 µm along the longitude of the auditory duct with a cryostat. Mount the sections directly on gelatin-coated microscope slides. Store the slides at -80 °C before immunostaining.
   9. Prior to immunostaining, immerse the slides in pre-warmed PBS (45 °C) for 5 min to dissolve the gelatin, and then wash the slides 3x with room temperature PBS to remove residual gelatin.

5. Immunostaining and microscope imaging

NOTE: Two types of immunostaining are performed depending on whether sections are mounted on slides or free-floating in PBS.

1. Immunostaining on slides
   1. Wash the slides 3x for 10 min each with PBS. Circle the region containing the sections with an oil pen to prevent liquid leaking during staining. Cover the sections with a blocking solution containing 5% normal goat serum in PBS and incubate for 30 min at room temperature.
   2. Dilute the primary antibodies (for the final concentration used refer to Table of Materials) in PBS containing 0.3% TritonX-100 and 5% normal goat serum. Remove the blocking solution and then cover the sections with the primary antibody solution. Incubate the slides at 4 °C overnight in a box containing a layer of distilled water at the bottom.
   3. Rinse the slides 3x for 10 min with PBS. Apply fluorescent secondary antibodies diluted at 1:500 in PBS containing 0.3% TritonX-100 on slides. Incubate the slides at room temperature for 4 h, shielded from light.
   4. Wash the slides 3x with PBS. Gently drop two drops of mounting medium on the slides and cover the sections with coverslips. Take care not to generate air bubbles between the coverslip and the slide.
2. Immunostaining for whole-mount embryos and free-floating sections
   1. Wash the embryos/sections with PBS 3x for 10 min each in a well plate. Dilute the primary antibodies in PBS containing 0.3% TritonX-100 and 5% normal goat serum in centrifugal tubes. Transfer each embryo/section into a tube with a glass hook and incubate on a shaker at 60 rpm overnight at 4 °C.
   2. Wash embryos/sections 3x with PBS for 10 min each in a well plate. Transfer each embryo/section into a dark centrifugal tube filled with fluorescent secondary antibody solution and incubate at room temperature for 4 h on the shaker.
   3. Wash the embryos/sections 3x with PBS in a well plate. Use a brush to mount the sections on gelatin-coated microscope slides in a 15 cm dish containing PBS. After the slides are dried slightly (~5 min), apply two drops of mounting medium and place a coverslip. Take care not to generate air bubbles between the coverslip and the slide. Keep the whole mount tissues in PBS for imaging.
3. Imaging
   1. Image the whole mount tissues in PBS in a dish with a silicon bottom containing carbon powder, which provides a black background. Capture and process images using a commercial image processing Olympus software package, as described previously²⁹.
   2. Image the sections on the slides with a confocal microscope as described previously²⁰. Image the sections at a single focal plane with 10x, 20x, and 63x objectives. Capture all images from the same animal using the same parameters. Take care to adjust the laser level and imaging acquisition settings to avoid signal saturation.
   3. Perform image processing using the Fiji software. For illustration, adjust image brightness, contrast, and gamma using a professional image editing software.

Representative Results

By performing in ovo electroporation at different sites and at different developmental stages, we achieved selective FMRP knockdown either in the auditory periphery or in the auditory brainstem.

FMRP knockdown in NM
Small hairpin RNA (shRNA) against the chicken Fmr1 was designed and cloned into the Tet-On vector system as described previously²⁰. The setup for in ovo electroporation is shown in Figure 1A. The plasmid DNA tinted with fast green was injected into the neural tube at the r5/r6 level at HH12 (Figure 1B-C). A platinum bipolar electrode was placed on each side of the neural tube at the r5/6 level (Figure 1D). Two pulses at 12 V were delivered to drive the plasmid into the tissue near the positive side of the electrode. Dox was applied on the chorioallantoic membrane to trigger the expression of Fmr1 shRNA and EGFP. After 24 h of Dox treatment, EGFP was evident under a fluorescent stereomicroscope. Figure 1E-F shows an example at E3 when Dox was applied at E2. Transverse sections confirmed the location of EGFP expressing (EGFP+) cells on one side of the hindbrain (Figure 1G-H). Additionally, a group of transfected cells was found anterior to the otocyst; these were cranial neural crest cells migrating ventrolaterally from the neural tube (Figure 1F). Electroporated embryos can also be incubated longer to study circuit formation at later stages. At E15, EGFP was observed in the brainstem, but only on the transfected side (Figure 1I-J). Cross sections at the level of the dorsal brainstem demonstrated a handful of EGFP+ neurons that were scattered in the NM on the side of electroporation (Figure 1L). EGFP+ fibers were seen projecting to the target of NM neurons: the dorsal nucleus laminaris (NL) on the ipsilateral side (Figure 1L) and the ventral NL on the contralateral side (Figure 1K). The knockdown effect of Fmr1 shRNA was validated by marked reductions in FMRP immunoreactivity in transfected NM neurons as compared to neighboring non-transfected neurons (Figure 1M-N). In the auditory ganglion (AG), neurons were non-transfected (EGFP-), although some glial cells surrounding AG neurons were EGFP+ (Figure 1O-P), presumably derived from EGFP+ cranial neural crest cells (see Figure 1F). Thus, this electroporation strategy provides a selective transfection of the postsynaptic neurons in the AG-NM circuit.

FMRP knockdown in the auditory duct
The plasmid mixture was injected into the otocyst at HH13 (Figure 2A), followed by electroporation by placing the positive side of the electrode anterior to the otocyst and the negative side posterior to the otocyst (Figure 2B). Whole-head sections at E6 demonstrated strong EGFP fluorescence in the right otocyst, but not in the brainstem (Figure 2C). Isolated auditory ducts at E9 exhibited extensive EGFP fluorescence throughout the length of the duct (Figure 2D-E). On transverse sections, EGFP+ cells were confirmed in the basilar papilla (BP) along the wall of the cochlear duct and in the AG (Figure 2F-G). The knockdown effect of Fmr1 shRNA was validated by marked reductions in FMRP immunoreactivity in transfected hair cells (*, Figure 2H-J) as compared to neighboring non-transfected hair cells (•, Figure 2H-J). For supporting cells, FMRP immunoreactivity was weak in both transfected and non-transfected cells (Figure 2H-J). Similar to hair cells, FMRP immunoreactivity was largely diminished in transfected AG neurons as compared to neighboring non-transfected neurons (Figure 2K-M).

We next tracked the central projection of transfected AG neurons to the brainstem via the auditory nerve. The mosaic pattern of Fmr1 shRNA transfection in the AG was further confirmed at E6 using Islet-1 immunoreactivity (Figure 3A), a marker for the developing chicken inner ear³¹. When a part of the AG was extracted with the brainstem during dissection, isolated EGFP+ AG neurons and their fibers within the brainstem were visualized in situ in the same preparations (Figure 3B-C). On transverse sections at the level of the NM, EGFP+ ANFs had reached the NM at E9 (Figure 3D) and were continuously observed at E15 and E19 (Figure 3E-F). High-magnification images revealed growth cone-like endings of ANFs at E9 (Figure 3H), which formed the large palm-like endbulb at E15 before they grew into the characterized endbulb of Held morphology with complex branches at E19 (Figure 3I-J). Beyond the NM, extensive branches of ANFs were also seen in the NA, another primary cochlear nucleus (Figure 3E). This was expected, as the same ANFs bifurcate to innervate both the NM and the NA in chickens³². We also saw EGFP+ fibers entering the adjacent vestibular cell groups, including the descending vestibular nucleus (VeD) as shown in Figure 3E. Although we optimized the location of the electroporating electrode to preferentially target the anterior portion of the otocyst where AG is formed, some vestibular ganglion neurons were also transfected. Immunostaining for parvalbumin, a marker for ANFs in chickens, can be used to differentiate auditory vs. vestibular fibers in the brainstem (Figure 3F-G).

In summary, this electroporation strategy provides a selective transfection of the presynaptic neurons in the AG-NM circuit and can be used to track the development of endbulb of Held at individual terminal levels following genetic manipulations.

Figure 1. FMRP knockdown in the chicken NM. (A) Image of electroporation setup. (B) Schematic drawing showing the microinjection site at the rhombomere r5/6 level of an HH12 chicken embryo. Abbreviations: NT = neural tube; SO = somite. (C) For enhanced visibility, Indian ink (black) was injected beneath the embryo with a syringe. Plasmid mixture tinted with fast green was then injected into the lumen of the neural tube. (D) Positioning of the bipolar electrode for electroporation. + and - indicate the positive and negative sides, respectively. Note, the blue plasmid mixture diffused into the neural tube on the positive side after electroporation. (E-F) Bright field (BF) image (E) merged with EGFP channel (F) of an electroporated chicken embryo at embryonic day 3 (E3), showing transfection in the neural tube at the same anteroposterior level of the otocyst. Dox was applied at E2, 6 h following the electroporation. Migrating cranial neural crest (CNC) cells are indicated by the yellow arrow and magnified in the inset in F. Scale bar = 500 µm in E, applies to E and F; 100 µm in the inset. (G-H) Transfected cells with Fmr1 shRNA and EGFP on the cross section of a transfected E3 embryo. EGFP+ cells were confined on one side of the neural tube. The section was double stained with FMRP immunoreactivity (gray). Scale bar = 40 µm in G, applies to G and H. (I-J) BF image (I) merged with EGFP channel (J) of an E15 brainstem following Dox treatment at E8. Scale bar = 1 mm in I, applies to I and J. (K-L) Cross section of an E15 brainstem with EGFP+ NM neurons on the transfected (trans) side only. On the contralateral (contra) side (K), EGFP+ axons were seen in the ventral portion of the nucleus laminaris (NL). (M-N) High magnification images of the NM at E19 with double labeling of FMRP (red) and SNAP25 (blue) immunostaining. SNAP25 is a presynaptic marker that labels the endbulb of Held. Note the marked reduction of FMRP immunoreactivity in the transfected (green; *) neurons as compared to neighboring non-transfected neurons (•). (O-P) High magnification images of the auditory ganglion (AG) at E19 with FMRP immunoreactivity. AG neurons (AGN) were not transfected (EGFP-). Some glial cells derived from CNC cells indicated in F were transfected. Scale bar = 10 µm in M, applies to M-P. Please click here to view a larger version of this figure.

Figure 2. FMRP knockdown in the chicken auditory duct. (A) Bright field (BF) image showing the microinjection of plasmids into the otocyst at HH13. Indian ink (black) was injected below the embryo for enhanced visibility. (B) Positioning of the bipolar electrode for electroporation. The positive side (labeled with a plus symbol, +) was positioned anterior to the otocyst, and the negative side (labeled with a minus symbol, -) was placed posterior to the otocyst. (C) Transverse section of a chicken head at E6 was immunostained for FMRP (red) and neurofilament (NF; blue). Dox was applied at E3 following the electroporation. EGFP fluorescence was detected in the otocyst but not in the brainstem. Scale bar = 100 µm. (D-E) BF image (D) merged with EGFP channel (E) of an auditory duct at E9. Scale bar = 500 µm. (F-G) Transverse section of an E9 auditory duct with FMRP (red) and NF (blue) immunostaining. EGFP+ cells were seen in the cochlear duct wall, basilar papilla (BP), and auditory ganglion (AG). Scale bar = 50 µm in F, applies to F and G. (H-J) High magnification images of an E9 BP showing largely diminished FMRP immunoreactivity in transfected hair cells (HCs; *) compared to non-transfected HCs (•). NF-positive fibers (blue) were the periphery innervation from AG neurons. (K-M) High magnification images of AG neurons at E9 with FMRP immunostaining (red). FMRP immunoreactivity was strong in non-transfected neurons (•) and not detectable in transfected neurons (EGFP+; *). Scale bar = 50 µm in H, and applies to H-M. Please click here to view a larger version of this figure.

Figure 3. Axonal tracking of transfected AG neurons in the brainstem. Embryos were electroporated at HH13 for otocyst transfection of Fmr1 shRNA. Dox application started at E2. (A) Transverse sections of an E6 chicken head with Islet-1 immunostaining and DAPI counterstain. EGFP+ cells were evident in the auditory ganglion (AG). Stars* indicate several transfected AG neurons in the inset. Scale bar = 100 µm in A and 10 µm in inset. (B-C) Bright field (BF) image (B) merged with EGFP channel (C) of an E9 brainstem. In this case, a portion of the AG (white arrows) was left connected to the auditory brainstem. Yellow arrow indicates transfected AG neurons (AGN) in the inset. EGFP+ fibers were evident on the transfected side of the brainstem. Scale bar = 1 mm in B, applies to B and C; 100 µm in the inset. (D) Cross section of the E9 brainstem at the level indicated in C by the dashed line. EGFP+ auditory nerve fibers (ANF; green) extended along the dorsal margin of the brainstem and approached the nucleus magnocellularis (NM). Scale bar = 100 µm. (E-G) EGFP+ fibers at E15 (E) and E19 (F-G). Some fibers were also seen in the nucleus angularis (NA) and the descending vestibular nucleus (VeD). E19 sections were stained for parvalbumin (PV) immunoreactivity as a marker for ANFs. The dashed box in F is magnified in the insets, showing an EGFP+ ANF, which was PV immunoreactive. Scale bar = 100 µm in E; 50 µm in F, applies to F and G; 2 µm in the insets. (H-J) High magnification images of ANF terminals in NM at E9, E15, and E19, showing the formation and progressive maturation of the endbulb of Held. Scale bar = 5 µm in H, applies to H-J. Please click here to view a larger version of this figure.

Discussion

To determine the cell-autonomous function of FMRP, manipulating its expression in individual cell groups or cell types is necessary. Since one of the major functions of FMRP is to regulate synaptic formation and plasticity, selectively manipulating each synaptic component of a certain circuit is prerequisite for a full understanding of FMRP mechanism in synaptic communication. Using in ovo electroporation of chicken embryos, we demonstrated a method to target FMRP expression in the AG-NM circuit, either in presynaptic AG neurons or postsynaptic NM neurons. To achieve this, choosing appropriate developmental stages for electroporation and accurately positioning the electroporation electrode on the embryo are critical steps. NM neurons arise from neural progenitors in the neural tube at the r5/6 level at HH12²³. At the same stage, cranial neural crest cells migrate from the dorsal tip of the neural tube to the cochleovestibular ganglion to envelop future AG neurons³³. Thus, both the NM neurons and the glia cells in the AG are neural tube derived. In contrast, hair cells and AG neurons originate from the ectoderm²⁴. The ectoderm adjacent to the r5/6 level thickens at HH10 and then invaginates to form the otic cup at HH13. The otic cup then closes and separates from the ectoderm to form the otocyst, with the anterior portion becoming the auditory sensory organ and the posterior portion contributing to the vestibular system. In both the auditory and vestibular portions, the neuroblast cells residing in the ventral part of the otic cup delaminate and migrate to form the cochleovestibular ganglion. Building upon these fate mapping studies, we transfected the progenitors of NM neurons by electroporating the neural tube at HH12, a strategy that has been established previously^26,28,34. Although this strategy leads to additional transfection of glial cells in the brainstem and the AG, none of the AG neurons were transfected. Neuronal-specific transfection can be achieved by employing an Atoh1 promotor-driven strategy, as what we did previously²¹. For selective transfection of AG neurons, we electroporated the otocyst at HH13. Plasmids injected into the otic cup predominantly entered the anterior portion of the otocyst for hair cell and AG neuron transfection simultaneously when positioning electrodes as shown in Figure 2B. To preferentially target AG neurons over hair cells, it is feasible to inject the plasmid solution straight into the region adjacent to the anterior otocyst, where the delaminated neuroblasts locate³⁵. However, this method has a lower efficiency of AG neuron transfection as compared to otocyst injection, as the plasmids diffuse quickly after injection. This method may also lead to additional transfection of the surrounding head mesenchymal cells. One method to preferentially target hair cells over AG neurons is to delay the otocyst injection and electroporation to HH18 (which is ~embryonic day 3), as the majority of neuroblasts (AG neuron precursors) have migrated out from the otocyst at this stage²⁴.

With modifications, this method can be extended to study the vestibular system. For vestibular ganglion transfection, the direction of the electrodes shown in Figure 2B must be reversed to transfect the posterior portion of the otocyst due to the distinct location of auditory and vestibular ganglion precursors.

In ovo electroporation of chicken embryos is a well-established technique for investigating early stages of development (within the first week of incubation). For long-term studies, survival rate is a major concern. When starting Dox treatment at E8, the survival rate found here is ~60% at E9, but drops to 30% at E15, and even lower at E19. Hatchlings from electroporated eggs are possible but rare and can only survive for several days in most cases. To increase the survival rate, the following should be considered: 1) use an electroporator that provides square waves instead of sharp waves; 2) apply one or two drops of sterile PBS to the top of the embryo if the albumin layer is not thick enough to promote electric conductivity (this will help reduce electrical trauma); 3) opening of the vitelline membrane to expose the embryo at the injection site is not required for a successful transfection and may harm the embryo; 4) to reduce the risk of contamination, use a syringe to poke the parafilm and administer Dox instead of reopening the window. After injection, wipe the transparent film with 75% ethanol and cover the needle opening with tape; 5) for otocyst transfection, inject plasmids into the otocyst dorsolaterally with a 45° angled pipette to avoid possible puncture of the aorta dorsalis underneath. If the aorta dorsalis is damaged, bleeding will flush out the plasmids and even cause death; 6) if possible, avoid using Indian ink to visualize the embryo. Less handling and fewer procedures will result in a higher survival rate.

In addition to the ability to selectively transfect AG or NM neurons, in ovo electroporation provides a unique system that allows neighboring comparisons for data analyses. After electroporation, only a subset of cells in either the NM or the AG are transfected, allowing the surrounding non-transfected cells in the same cell group to serve as an ideal control²⁰. If potential interaction among neighboring neurons is a concern, neuron counterparts from the non-transfected side of the ear and brain can serve as an additional control. For histological and imaging studies, this outcome enables data analyses with high efficiency at the single cell or single fiber level. Molecular analyses are also feasible if combined with fluorescent-based cell sorting and large-scale protein and RNA analyses such as mass spectrometry and next-generation sequencing. However, using in ovo electroporation as a means of gene editing may be challenging for functional and behavior studies because the percentage of transfected cells within a cell group may not be large enough to result in functional consequences.

It is essential to perform two verification analyses when adopting the in ovo electroporation approach. First, the effect of gene editing on the protein of interest should be confirmed. In ovo electroporation transfects only a subset of neurons in the NM or AG (i.e., a mosaic pattern). Performing western blot on homogenized tissues that contain a mixture of transfected and non-transfected neurons is not sensitive enough to accurately reflect the knockdown effect. Thus, the validation must be conducted at the individual cell level using methods such as immunocytochemistry. For this purpose, we previously generated an antibody that recognizes chicken FMRP. The specificity of this antibody was verified by immunocytochemistry and western blot in a previous study³⁰. The knockdown effect of Fmr1 shRNA was validated quantitatively by observing significant reductions in the intensity of FMRP immunoreactivity in transfected NM neurons as compared to neighboring non-transfected neurons. Second, control groups using scrambled RNAs, or other control constructs, should be employed to confirm the specificity of observed cellular effects of FMRP misexpression^20,21. To achieve this goal, we designed a scrambled shRNA, cloned it into the same Tet-On vector system, and electroporated it into NM precursors, as described previously²⁰. FMRP expression in the NM neurons transfected with the scrambled shRNA was not affected, as evidenced by the unchanged intensity of FMRP immunoreactivity as compared to neighboring non-transfected neurons. We have also identified a number of effects of Fmr1 shRNA transfection on axonal growth, dendritic pruning, terminal formation, and neurotransmission (reviewed in Curnow and Wang, 2022)³⁶. We concluded that these effects were induced specifically by cell autonomous loss of FMRP because the scrambled shRNA failed to induce similar changes.

In conclusion, the in ovo electroporation technique enables selective Fmr1 gene editing with temporal control and component-specificity in the auditory ear-brainstem circuits and can be modified to manipulate the vestibular system. Development of this advanced technology in the chicken embryo, a well-established model in developmental biology, has and will continue to contribute to our understanding of brain development under normal and abnormal conditions such as FXS.

Materials

Name	Company	Catalog Number	Comments
Egg incubation
16 °C refrigerator	MAGAT		Used for fertilized egg storage.
Egg incubator	SHANGHAI BOXUN	GZX-9240MBE
Fertilized eggs	Farm of South China Agricultural University		Eggs must be used in one week for optimal viability.
Plasmid preparation
Centrifuge	Sigma	10016
Fast green	Solarbio	G1661	Make 0.1% working solution in distilled water and autoclave.
Plasmid Maxi-prep kit	QIAGEN	12162	Dissolve plasmid DNA in Tris-EDTA (TE) buffer; endotoxin-free preparation kit
Sodium Acetate	Sigma-Aldrich	S2889	Make 7.5M working solution in nuclase-free water.
Electroporation and Doxycycline Administration
Electroporator	BTX	ECM399
1 mL / 5 mL Syringe	GUANGZHOU KANGFULAI
Dissecting microscope	CNOPTEC	SZM-42
Doxcycline	Sigma-Aldrich	D9891	Use fresh aliquots for each dose and store at -20 °C.
Glass capillary	BEIBOBOMEI	RD0910	0.9-1.1 mm*100 mm
Laboratory parafilm	PARAFILM	PM996	transparent film
Pipette puller	CHENGDU INSTRUMENT FACTORY	WD-2	Pulling condition: 500 °C for 15 s
Platinum elctrodes	Home made		0.5 mm diameter, 1.5 mm interval.
Rubber tube	Sigma-Aldrich	A5177
Tissue Dissection and Fixation
Forceps	RWD	F11020-11	Tip size: 0.05*0.01 mm
Other surgery tools	RWD
Paraformaldehyde	Sigma-Aldrich	158127	Freshly made 4% PFA solution in phosphate-buffered saline can be stored in 4 °C for up to 1 week.
SYLGARD 184 Silicone Elastomer Kit	DOW	01673921	For black background plates, food-grade carbon powder is applied.
Sectioning
Cryostat	LEICA	CM1850
Gelatin	Sigma-Aldrich	G9391	From bovine skin.
Sliding microtome	LEICA	SM2010
Immunostaining
Alexa Fluor 488 goat anti-Mouse	Abcam	ab150113	1:500 dilution, RRID: AB_2576208
Alexa Fluor 555 goat anti-rabbit	Abcam	ab150078	1:500 dilution, RRID: AB_2722519
DAPI	Abcam	ab285390	1: 1000 dilution
Fluoromount-G mounting medium	Southern Biotech	Sb-0100-01
FMRP antibody	Y. Wang, Florida State University	#8263	1:1000 dilution, RRID: AB_2861242
Islet-1 antibody	DSHB	39.3F7	1:100 dilution, RRID: AB_1157901
Netwell plate	Corning	3478
Neurofilament antibody	Sigma-Aldrich	N4142	1:1000 dilution, RRID: AB_477272
Parvalbumin antibody	Sigma-Aldrich	P3088	1:10000 dilution, RRID: AB_477329
SNAP25 antibody	Abcam	ab66066	1:1000 dilution, RRID: AB_2192052
Imaging
Adobe photoshop	ADOBE		image editing software
Confocal microscope	LEICA	SP8
Fluorescent stereomicroscope	OLYMPUS	MVX10
Olympus Image-Pro Plus 7.0	OlYMPUS		commercial image processing software package