Electroporation is an effective approach to deliver genes of interests into cells. By applying this approach in vivo on the neurons of adult mouse dorsal root ganglion (DRG), we describe a model to study axon regeneration in vivo.
Electroporation is an essential non-viral gene transfection approach to introduce DNA plasmids or small RNA molecules into cells. A sensory neuron in the dorsal root ganglion (DRGs) extends a single axon with two branches. One branch goes to the peripheral nerve (peripheral branch), and the other branch enters the spinal cord through the dorsal root (central branch). After the neural injury, the peripheral branch regenerates robustly whereas the central branch does not regenerate. Due to the high regenerative capacity, sensory axon regeneration has been widely used as a model system to study mammalian axon regeneration in both the peripheral nervous system (PNS) and the central nervous system (CNS). Here, we describe a previously established approach protocol to manipulate gene expression in mature sensory neurons in vivo via electroporation. Based on transfection with plasmids or small RNA oligos (siRNAs or microRNAs), the approach allows for both loss- and gain-of-function experiments to study the roles of genes-of-interests or microRNAs in regulation of axon regeneration in vivo. In addition, the manipulation of gene expression in vivo can be controlled both spatially and temporally within a relatively short time course. This model system provides a unique tool to investigate the molecular mechanisms by which mammalian axon regeneration is regulated in vivo.
Injuries in the nervous system caused by neural trauma or various neurodegenerative diseases usually result in defects in motor, sensory and cognitive functions. Recently, much effort has been devoted to regenerative potency re-establishment in adult neurons to restore the physiological functionsof injured neurons1,2,3. Sensory neurons in the DRG are a cluster of nerve cells that convey different sensory stimuli, such as pain, temperature, touch, or body posture, to the brain. Each of these neurons is pseudo-unipolar and contains a single axon that bifurcates with one branch extending toward the periphery and the other branch heading toward the spinal cord4. The adult sensory neurons in DRGs are among a few mature mammalian neurons known to regenerate their axons actively after injury. Hence, injuries of sensory axons have been extensively employed as a crucial model to study the mechanisms of axonal regeneration in vivo.
In vivo gene transfection techniques, which are usually less time-consuming to set up and more flexible than using transgenic animals, have been playing essential roles in studying the functions of genes and signaling pathways in the nervous system. The main techniques can be categorized into two approaches: instrument-based and virus-based5. Viral-based in vivo gene delivery in adult neurons can provide precise spatiotemporal manipulation of gene expression6. However, labor-intensive processes are involved in viral-based methods, such as the production and purification of viral particles containing the desired gene. In addition, many viral vectors could activate the immune system of the host, which may interfere with the data acquisition, data analysis and possibly mislead the interpretation of experimental results. Electroporation, a typical instrument-based transfection approach, uses an electrical pulse to increase the permeability of cell and nuclear membranes transiently, which favors the influx of gene vectors or small RNA oligos from space outside of the cells7. In vitro electroporation is widely recognized as a transient but highly efficient strategy for manipulating targeted gene expression in many cell types. Although in vivo electroporation only leads to transient gene expression with low transfecting efficiency compared to viral vectors, it has various advantages over viral approaches. For instance, it can be applied to almost all tissues and cells7,8,9. In addition, either plasmids encoding genes-of-interest or small RNA oligos (e.g., siRNAs, microRNAs) against certain transcripts can be injected into the target tissue directly and then electrically pulsed, which make the procedure less labor- and time- consuming. Moreover, transfecting multiple plasmids and RNA oligos simultaneously with single electroporation is possible.
We have established an in vivo electroporation approach to manipulate gene expression in adult mouse sensory neurons and successfully applied and validated such approach in numerous pioneer studies1,2,3,8,10. Here, we present a detailed protocol to facilitate the usage of this approach for future studies of mammalian axon regeneration.
All animal experiments were performed in accordance with the animal protocol approved by the Johns Hopkins Institutional Animal Care and Use Committee.
1. Materials and Reagents
2. Experimental Procedures
To quantify the cytotoxicity of the current protocol and to validate that transfection rate of in vivo DRG electroporation is high enough, we injected and electroporated fluorescently-tagged microRNA or siRNA into L4 and L5 DRGs. The detached DRGs were processed through cryo-sectioning and immunohistochemistry (Figure 1A-B). When estimating the cell survival rate after injection and electroporation, the intact DRGs from L4 and L5 were harvested and processed through cryo-sectioning and immunohistochemistry. The neuron densities, reflected with Tuj1 staining, did not show a significant difference when compared to the neuron densities of intact DRGs, which indicates that the electroporation did not induce neuronal cell death (Figure 1C). The transfection rate was calculated as the ratio between the number of tagged RNA oligo neurons and the number of Tuj1-positive neurons. The mean transfection rate of miRNA is 80.7 ± 4.3% and the siRNA 94.2 ± 0.3% (Figure 1D). When applying the current in vivo electroporation method to study axon regeneration, the sciatic nerve was flattened and imaged. Every regenerated axon with distinctive trajectory and recognizable distal axon end can be traced from the crush site indicated by the suture knot (Figure 2). Additionally, we electroporated the DRGs with EGFP plasmids and harvested the spinal cord from T7 to L2 along with dorsal roots of L4 and L5 after 7 days. We used a well-established tissue clearing protocol — uDISCO to clear the spinal cord11. The EGFP-labeled dorsal column axons in the spinal cord were imaged with a confocal microscope and then identified in both the longitudinal and the sagittal projecting images (Figure 3). Alternatively, the confocal images can be processed with microscopy visualization software to reconstruct a 3D image (Movie 1).
Figure 1: Immunohistochemistry of DRG cryo-section after the electroporation of fluorescently-tagged non-specific microRNA or siRNA in vivo. (A) Representative cryo-section of DRG tissue injected and electroporated with fluorescently-tagged (Dy547) non-target microRNAs. Left image: the image of the fluorescent signal (red color) of fluorescently-tagged (Dy547) non-target microRNAs. Middle image: immuno-staining (green color) of Tuj1 with an Alexa488 fluorophore on the secondary antibody. Right image: the merged image of the previous two channels. Scale bar = 100 µm.(B) Representative cryo-section of DRG tissue injected and electroporated with fluorescently-tagged (Cy3) non-target siRNAs. Left image: the image of the fluorescent signal (red color) of fluorescently-tagged (Cy3) non-target siRNAs. Middle image: immuno-staining (green color) of Tuj1 with an Alexa488 fluorophore on the secondary antibody. Right image: the merged image of the previous two channels. The scale bar represents 100 µm. (C) The neuron density of an intact DRG cryo-section is 809.6 ± 14.2 cells/mm2 (N = 6) and an injected DRG cryo-section is 801.6 ± 27.4 cells/mm2 (N = 6), Mean ± SEM, Student's t-test, NS: no significance. Three mice were performed in vivo DRG electroporation on the left L4 and L5 DRGs with tagged siRNA. Both left and right L4 and L5 DRGs are harvested after 48 h and processed through cryo-section and immunohistochemistry. Each DRG has been sectioned into approximate 60 slices and the thickness of the slice is 10 μm. Three slices of each DRG are selected by 200 μm and averaged. (D) The transfection rate of the tagged miRNA is 80.7 ± 4.3% (N = 4) and the tagged siRNA is 94.2 ± 0.3% (N = 6), Mean ± SEM. Three mice were performed in vivo DRG electroporation on the left L4 and L5 DRGs with tagged siRNA and two mice with tagged miRNA. The DRGs are harvested after 48 h and processed through cryo-section and immunohistochemistry. Each DRG has been sectioned into approximate 60 slices and the thickness of the slice is 10 μm. Three slices of each DRG are selected by 200 μm and averaged. Please click here to view a larger version of this figure.
Figure 2: Ectopically expressed EGFP displays regenerated neuronal axons in sciatic nerve after crushing injury. The axons in a flattened sciatic nerve suffered from crush injury are labeled with EGFP (green fluorescence) transported from the somas to axons. The red line indicates the position of the crushing injury site, which was originally surgically marked by a suturing knot. The white arrowhead, arrow, and star indicate three distinctive axon ends, all of which extend from the crush site. Scale bar = 500 µm. Please click here to view a larger version of this figure.
Figure 3: Projection images of EGFP-expressing axons in the spinal cord after in vivo electroporation of DRGs. (A) Longitudinal projection of the EGFP-labeled axons within the dorsal column and the L4/L5 dorsal roots. Inset images A1 and A2 show magnified views of the two rectangular areas marked in panel (A). A1 exhibits a detailed view of axons in the dorsal column. A2 exhibits a detailed view of axons in the dorsal root entry zone (DREZ). Scale bar = 200 µm. (B) Sagittal projection of the EGFP-labeled axons within the dorsal column. Scale bar = 50 µm. Please click here to view a larger version of this figure.
Movie 1: 3D reconstruction of EGFP-expressing axons in the spinal cord after in vivo electroporation of DRGs. Please click here to view this video. (Right-click to download.)
Several surgical steps require particular attention. The L4 and L5 DRGs (location of somas), which dominate the sciatic nerve, need to be correctly identified and injected with gene constructs. Otherwise, the GFP-labeling will be absent in sciatic nerve axons. The iliac crests can be viewed as useful anatomical landmarks to pinpoint L4 and L5 DRGs. In most mice, the facet joint between L5 and L6 vertebrae is proximate to iliac crests12. Alternatively, L3 DRG can be chosen instead of L5, especially for assays such as immunohistochemistry or western blot13, as surgical exposure of L5 DRG usually meets great difficulties due to a deep location and rich blood supply. Also, the whole surgery should avoid damaging DRG surrounding structures, including the spinal cord and the nerve root. The second critical operational step worth noting is the injection. The Fast Green dye was mixed with the solution of gene constructs to visualize the solution ejected from the needle and diffuse in the DRG. A successful injection should show a clear round-shape DRG outlined by the fluorescent dye. If the solution with fluorescent dye leaks out from the DRG while injecting, a back-and-forth fine-tuning on the depth of the tip of micro-needle can ensure complete injection of all solution into the DRG capsule. Avoid impaling the DRG at more than three different sites. Further, mild bleeding is usually inevitable after DRG injection, as DRGs are tightly encapsulated with rete venosum13. It is important to stop bleeding so that there will not be blood insulating the surface of DRG from electrodes during electroporation. Successful electroporation depends on electric current transduction, and PBS solution can be applied on the electrode. Finally, the site on the sciatic nerve where is crushed with forceps should be marked with micro-suturing, because the injury site is invisible under the microscope and needs to be indicated as the starting point of axon regeneration for image analysis later. If the epineural suture knot at the crush site falls off after animal perfusion and dissection, it is critical to re-label the same site with a suture immediately on the PFA-fixed nerve when the crush-induced indentation is still identifiable.
Among transfection techniques, electroporation has a higher transfection rate. According to our study, the transfection rate of the DRG neuron for microRNAs or siRNAs after in vivo electroporation is close to 90%. More importantly, in vivo electroporation is far less time-consuming than virus-based methods, which require virus packaging of desired gene constructs. As far as we know, even though the lipofectamine-based in vivo transfection has less toxicity than electroporation, the lipofectamine does not work on DRG neurons particularly, neither in vivo nor in vitro. Noteworthy, the current methodology has several technical limitations. First, the duration of siRNA efficacy to knocking down gene-of-interest is shorter than the virus-based delivery of plasmids. Therefore, the whole procedure from electroporation to animal sacrifice has to be finished in 4–6 days14. Additionally, the surgery performed under the microscope requires practice and some manual dexterity. For a novice, the surgery duration is often longer than expected. The anesthetic dose administrated to the mouse has to be carefully controlled to prevent unwanted mortality. Finally, flattening the sciatic nerve causes overlapping of axons when conducting epifluorescent imaging. Imaging the unflattened nerves with a confocal microscope is an alternative option.
Anatomically, the DRG sensory neurons have two axonal branches – the peripheral descending branch and the ascending central branch projecting into the dorsal column of the spinal cord10. The current methodology also shows distinctive labeling of the dorsal column of the spinal cord. Thus, a similar methodology can be used as a model to investigate sensory axon regeneration after spinal cord injury. Combined with tissue-clearing techniques11, conventional confocal microscopy or light-sheet microscopy can be employed on the cleared spinal cord sample to build 3D reconstructed images of sensory axons within the dorsal column of the spinal cord.
The authors have nothing to disclose.
The study was funded (awarded to F-Q.Z.) by NIH (R01NS064288, R01NS085176, R01GM111514, R01EY027347), the Craig H. Neilsen Foundation and the BrightFocus Foundation.
ECM 830 Square wave electroporation system | BTX Harvard Apparatus | 45-0052 | For in vivo electroporation |
Tweezertrodes electrodes | BTX Harvard Apparatus | 45-0524 | For in vivo electroporation, 1 mm flat |
Picospritzer III | Parker Instrumentation | 1096 | Intracellular Microinjection Dispense Systems |
Glass Capillary Puller | NARISHIGE | PC-10 | |
Borosilicate Glass Capillaries | World Precision Instruments, Inc. | 1902328 | |
Stereo Dissection Microscope | Leica | M80 | |
Microsurgery Rongeur | F.S.T | 16221-14 | |
Microsurgery Forceps | FST by DUMONT, Switzerland | 11255-20 | Only for sciatic nerve crush |
Glass Capillary | World Precision Instruments, Inc. | TW100-4 | 10 cm, standard wall |
Tape | Fisherbrand | 15-901-30 | For fixing the mouse on the corkboard |
2, 2, 2-Tribromoethanol (Avertin) | Sigma-Aldrich | T48402 | Avertin stock solution |
2-methyl-2-butanol | Sigma-Aldrich | 152463 | Avertin stock solution |
siRNA Fluorescent Universal Negative Control #1 | Sigma-Aldrich | SIC003 | Non-target siRNA with fluorescence |
microRNA Mimic Transfection Control with Dy547 | Dharmacon | CP-004500-01-05 | Non-target microRNA with fluorescence |
Plasmids preparation kit | Invitrogen Purelink | K210016 | GFP-coding plasmid preparation |
Fast Green Dye | Millipore-Sigma | F7252 | For better visualization of the DRG outline during injection |
Ketamine | Putney, Inc | NDC 26637-731-51 | Anesthesia induction |
Xylazine | AnaSed | NDC 59399-110-20 | Anesthesia induction |
Acetaminophen | McNeil Consumer Healthcare | NDC 50580-449-36 | Post-surgical pain relief |