Direct Gene Knock-out of Axolotl Spinal Cord Neural Stem Cells via Electroporation of CAS9 Protein-gRNA Complexes
Summary
Presented here is a protocol to perform time- and space-restricted gene knock-out in axolotl spinal cords by injecting CAS9-gRNA complex into the spinal cord central canal followed by electroporation.
Abstract
The axolotl has the unique ability to fully regenerate its spinal cord. This is largely due to the ependymal cells remaining as neural stem cells (NSCs) throughout life, which proliferate to reform the ependymal tube and differentiate into lost neurons after spinal cord injury. Deciphering how these NSCs retain pluripotency post-development and proliferate upon spinal cord injury to reform the exact pre-injury structure can provide valuable insight into how mammalian spinal cords may regenerate as well as potential treatment options. Performing gene knock-outs in specific subsets of NSCs within a restricted time period will allow study of the molecular mechanisms behind these regenerative processes, without being confounded by development perturbing effects. Described here is a method to perform gene knock-out in axolotl spinal cord NSCs using the CRISPR-Cas9 system. By injecting the CAS9-gRNA complex into the spinal cord central canal followed by electroporation, target genes are knocked out in NSCs within specific regions of the spinal cord at a desired timepoint, allowing for molecular studies of spinal cord NSCs during regeneration.
Introduction
The spinal cord of most vertebrates is unable to regenerate following injury, leading to permanent disability. Several salamanders, such as the axolotl, are notable exceptions. The axolotl can fully regenerate a structurally identical spinal cord and completely restore spinal cord function. Much of the regenerative capability of the axolotl spinal cord is due to ependymal cells. These cells line the central canal, and unlike those in mammals, axolotl ependymal cells remain as neural stem cells (NSCs) post-embryonic development. After spinal cord injury (e.g., from a tail amputation), these NSCs proliferate to regrow the ependymal tube and differentiate to replace lost neurons1,2,3. Uncovering how axolotl spinal cord NSCs remain pluripotent and become activated after injury can provide valuable information on developing new therapeutic strategies for human patients.
Due to advances in the CRISPR-Cas9 gene knock-out technique, performing knock-outs to decipher gene function has become easier and has been shown to have broad applicability in various species, including axolotls4,5,6,7,8. The recent release of the full axolotl genome and transcriptome now allows any genomic locus to be targeted and for better assess to off-target effects9,10,11,12,13,14. Optimized protocols have been developed for knock-out and knock-in in axolotls using the CRISPR-Cas9 system15. Delivery of the CRISPR-Cas9 machinery in the form of CAS9 protein-gRNA ribonucleoprotein (RNP) has been shown to be more efficient than using Cas9 and gRNA-encoding plasmids4. This is likely due to the RNP being smaller in size than plasmid vectors, its ability to create DNA breaks immediately, and protecting of the gRNA from RNA degradation. In addition, using RNPs bypasses transcription and translation; thus, it avoids issues such as promoter strength and optimal codon usage when plasmid elements are derived from a different species.
Loss-of-function studies are one of the general approaches to investigating potential functions of genes of interest. In order to study gene function during regeneration, a knock-out should ideally be performed just before an injury to avoid effects on development. In addition, the knock-out should be restricted to both the NSCs and region of regeneration. A knock-out of the target gene in all NSCs (including those in the brain, which is the case in Cre-LoxP systems), may produce effects not related to regeneration that can confound the interpretation of results. Fortunately, the structure of the axolotl spinal cord provides a unique opportunity for time- and space-restricted knock-out in NSCs. Most of the spinal cord NSCs are in contact with the central canal and constitute the vast majority of the cells in contact with the central canal16,17. Therefore, an injection of the CAS9-gRNA complex into the central canal, followed by electroporation, allows delivery to spinal cord NSCs in a desired region at a specific time4,18,19. This protocol demonstrates how this is performed, leading to highly penetrating knock-out in the targeted spinal cord NSCs. Subsequent analysis are then performed to study the effects on regeneration and NSC behavior.
Protocol
All animal experiments must be carried out in accordance with local and national regulations on animal experimentation and with approval of the relevant institutional review board.
1. Preparing the CAS9-gRNA RNP mix
- Design and synthesize gRNAs.
NOTE: Refer to other publications for designing and synthesizing gRNAs, including one exclusively concerning axolotls15,20,21,22. - Obtain CAS9-NLS protein by preparing in-house or purchasing it commercially.
- Prepare CAS9-gRNA mix by mixing 5 µg of CAS9-NLS protein, 4 µg of gRNA, 0.9 µL of 10x CAS9 buffer, and fill up the volume to 10 µL with nuclease-free H2O. Aliquot the mixture and store at -80 °C if not used immediately.
2. Preparing agarose plates for electroporation
- Prepare 200 mL of 2% (wt/vol) agarose solution in 1x DPBS. Dissolve by heating the solution in a microwave and pour it onto 10 cm Petri dishes to a depth of about 10 mm (deep enough to cover the whole animal).
- Allow the plates to solidify at room temperature. Store plates at 4 °C if not used immediately.
- Using surgical scalpels, cut the agarose plate to make a slit for holding the tail straight, a well for fitting in the body of the animal, and two extra wells for placing the electrodes (Figure 1E).
- Place the plate on ice and fill it up to the brim with ice-cold 1x DPBS. Use the same DPBS solution for making the plates to ensure consistent electrical conductibility in the set-up.
3. Configuring the electroporator
- Configure the electroporator. Set up the poring pulse to consist of one bipolar pulse of 70 V, with a duration of 5 ms, interval of 50 ms, and no voltage decay. Set up the transfer pulse to have four bipolar pulses of 40 V, with a duration of 50 ms, interval of 999 ms, and 10% voltage decay.
- Connect the electrodes to the electroporator and adjust the tweezers to be 7 mm apart. Submerge the electrodes in a beaker of DPBS before and in between electroporation.
4. Preparing and loading microinjection glass capillaries
- Perform a ramp test on a flaming/brown micropipette puller as per the manufacturer’s instructions to determine the heat value.
- Pull injection capillaries with tapered ends using the following parameters: heat = ramp test value, pull = 100, velocity = 100, time = 100.
- Observe the needle under a stereomicroscope. Using fine forceps, break the capillary at an angle so that it has a slanted tip. Break at a position where it is thin enough to target the central canal, while being strong enough to pierce the skin.
NOTE: The point at which the diameter is at about 20 µm or about 50% of the diameter of the spinal cord is a good starting point. Multiple tries may be needed to obtain optimal breaking. - Add Fast Green FCF solution to the RNP mix at a ratio of 1:30 and load about 5 µL of injection mix into the capillary with a Microloader pipette tip.
- Mount the capillary on the pneumatic pump, with the slanted tip facing downwards.
5. Configuring the pneumatic pump
- Configure the pneumatic pump. Set the hold to 0.5 pounds per square inch (psi), eject to 2 psi hold, and duration to gated.
- Test capillary and pneumatic pump settings by dipping the capillary into a drop of water and pressing the foot pedal to inject. Ensure that the injection mix comes out in a slow but steady stream.
- Adjust the hold pressure if the injection mix start leaking out spontaneously or water is sucked up the capillary. Increase the ejection pressure or break more of the capillary if the injection mix is coming out too slow. Decrease the ejection pressure if the injection is coming out too fast.
6. Injecting RNP mix into the spinal cord
- Anesthetize axolotls in 0.03% benzocaine solution for at least 10 min. Check that the animals are not responsive by flipping them upside down in the water.
- Pick up an animal with ring forceps and lay it on a bed of silicon, with the left side of the animal facing up and the tail pointing towards the right. Position the injection site in the middle of the field of view of the stereomicroscope (Figure 1A).
- Adjust the micromanipulator so that the capillary is coming in at 60° from the right side of the field of view (Figure 1B).
- Identify the spinal cord and central canal (Figure 1C).
NOTE: The spinal cord is the tubular structure above the notochord. - Aiming at the central canal, move the capillary forward to gently pierce the skin and muscle into the central canal of the spinal cord. Limit the movement of the capillary to small steps, as the spinal cord is right beneath the tail muscles.
- Press and hold down the foot pedal to inject the mix into the central canal. Observe whether the RNP mix spreads along the central canal as a blue line in both directions, which shows that the capillary has been correctly positioned (Figure 1D).
- If this does not occur, keep the foot pedal pressed while moving the capillary slightly in and out until the RNP mix starts going in. If the capillary becomes clogged by tissue, clear it by ejecting into water at higher pressures.
- Adjust the ejection pressure so that it is just enough to cause the RNP mix to spread in the central canal, but not so much that it blows up the structure.
- Continue to hold down the foot pedal so the injection mix spreads further along the central canal, until it reaches the terminal vesicle at the end of the spinal cord and ventricles in the brain.
NOTE: This may take up to 2 min depending on the size of the animal. It may be necessary to increase the pressure after some time to cause the RNP mix to enter the brain ventricles. - Proceed as soon as possible to electroporation after successful injection.
7. Electroporation
- Transfer the animal with ring forceps onto the prepared agarose electroporation plate (Figure 1E) on ice. Place the tail inside the slit so it is sandwiched by agarose (Figure 1F).
- Place the electrodes into the wells on both sides of the tail. Ensure that the entire animal and electrodes are covered by ice-cold PBS.
- Press the foot switch to start the electroporation.
- After electroporation is complete, return the animal to water.
- Proceed to inject more animals if necessary.
- Check that the electroporated animals are healthy after anesthesia wears off and that there is no damage done to the tail.
8. Assessing knock-out efficiency
- Fix the tail of the electroporated animal. Make cross-sections of the tail followed by immunohistochemical staining to assess the efficiency of the knock-out at the protein level.
NOTE: If a tail amputation is to be performed after electroporation, the cut off tail could be used for this purpose. Animals electroporated with gRNAs against Gfp or Tyrosinase, for example, can be used as a negative control4. - Alternatively, the electroporated spinal cord could be extracted and lysed for DNA. Perform genotyping PCR followed by Sanger sequencing or next-generation sequencing to assess the percentage of edited genetic loci15.
NOTE: The resulting editing efficiency will underestimate the actual editing efficiency for NSCs, due to the inclusion of non-electroporated cells and neurons that lack apical contact to the central lumen and thus will not receive the RNP mix.
Representative Results
Injection and electroporation of CAS9-gRNA complex against Sox2 into the axolotl spinal cord central canal led to a massive loss of SOX2 immunoreactivity in a majority of spinal cord NSCs, with gRNA against Tyrosinase (Tyr) as a control (Figure 2A). B3-tubulin (stained with TUJ1) is a marker for neurons and was not expressed in NSCs, and SOX2- TUJ1- cells surrounding the central canal were considered to be cells harboring Sox2 deletions. Quantification showed Sox2 knock-out in a significant number of NSCs by CAS9-Sox2-gRNA electroporation (Figure 2B), indicating that this method led to efficient and highly penetrating gene knock-out in spinal cord NSCs.
After simultaneous injection and electroporation, the mixture of two CAS9-gRNA complexes against Sox2 and Gfp, respectively, into transgenic axolotls with ubiquitous GFP expression (CAGGS-GFP), a high percentage of double knock-out NSCs were observed (Figure 2C,D), indicating that the protocol can be used to flexibly knock-out multiple genes.
Figure 1: Axolotl spinal cord injection and electroporation. (A) Schematic showing injection of CAS9-gRNA RNP into axolotl spinal cord central canal. (B) Micromanipulator set-up for the injection. (C) View through a stereomicroscope of the axolotl tail, with the spinal cord indicated. (D) View through a stereomicroscope of the axolotl tail with an injected spinal cord. (E) Schematic of the electroporation plate. (F) Schematic of the animal and electrodes placed inside the electroporation plate, ready for electroporation. Scale bars = 1 mm. This figure is adapted from Albors and Tanaka18. Please click here to view a larger version of this figure.
Figure 2: Representative knock-out results in axolotl spinal cord NSCs. (A) Spinal cord injection and electroporation of CAS9-Sox2-gRNA (n = 21) led to loss of SOX2 in the majority of NSCs at 15 days post-electroporation (dpE) by immunohistochemical staining. CAS9-Tyr-gRNA (n = 21) served as a negative control. Scale bars = 100 μm. (B) Quantification showed significant reduction in SOX2+ NSCs (p < 0.001 by Student’s t-test), counting SOX2- TUJ1- cells surrounding the central canal as NSCs harboring a Sox2 deletion. Error bars = SD. (C) Simultaneous injection and electroporation of Sox2-gRNA and Gfp-gRNA coupled to CAS9 (n = 6) led to loss of SOX2 and GFP in the majority of NSCs by immunohistochemical staining. CAS9-Tyr-gRNA (n = 6) serves as negative control. Scale bars = 100 μm. (D) Quantification showed that among all targeted NSCs having any deletion, a majority (94%) harbored double deletion (GFP− SOX2−), with only a small proportion (6%) having single deletion (GFP+ SOX- or GFP- SOX2+). This figure is adapted from Fei et al.4. Please click here to view a larger version of this figure.
Discussion
The described protocol allows time and space restricted gene knock-out in the NSCs in the axolotl spinal cord. The current protocol allows specific targeting of NSCs at a defined time and location with high penetrance. It avoids potential undesired effects originating from gene knock-out in NSCs in other regions such as the brain that occur when using the Cre-LoxP system. It also avoids developmental effects originating from a persistent knock-out, allowing the study of gene function focused on regeneration. In addition, this protocol can be performed in wild-type animals, avoiding the long waiting time needed to generate additional transgenic axolotls, which is required for the Cre-LoxP system. It also offers the versatility to knock-out multiple genes simultaneously and is highly penetrating, allowing experiments to be performed on F0 animals. Importantly, the procedure is shown to have no observable effect on tail and spinal cord regeneration4.
The use of CAS9 protein-gRNA RNP complexes also displays much higher efficiency than systems using Cas9 and gRNA-expressing plasmids4. This is likely due to smaller size of the RNP complex compared to plasmids, as well as differential codon usage affecting expression of the CAS9 protein from plasmids derived from mammalian systems. In addition, since the CAS9 protein-gRNA RNP complexes can induce breaks immediately, knock-out occurs quickly. 60%-70% of modified genomic loci have been observed by genotyping PCR within 24 h of electroporation (data not shown). There is also a possibility to perform knock-ins using this strategy by co-electroporating a repair template. A number of knock-in strategies are available, the details and considerations of which are out of scope for this publication, but they are described in detail in another publication15.
On the other hand, this protocol does have a few limitations. Staining or genotyping PCR is generally needed to assess the extent of knock-out in each experimental animal, which can be laborious. While gene knock-out from this method is largely restricted to the NSCs surrounding the central canal, it cannot be ruled out that other cell types could also be targeted, because 1) they are also in contact with the central canal, or 2) the RNP mix has leaked out of the central canal before electroporation. These factors should be considered when interpreting the results.
gRNA design
With the recently published axolotl genome and transcriptome, identifying axolotl genes and their sequences has become much easier9,10,11,12,13. Detailed guides in designing gRNAs has been described elsewhere, including one that deals exclusively with axolotls15. It is advised to design and test at least three gRNAs for editing efficiency beforehand and to choose the optimal design for the actual experiment. Efficiency of the gRNA can be tested by injecting the CAS9-gRNA mix into freshly laid axolotl eggs at the one-cell stage23. Editing efficiency can be then assessed by genotyping the hatched larvae.
Injection considerations
Since the injection site of the spinal cord suffers from physical damage caused by the needle and spreading of the RNP mix to tissues outside the central canal, it is important to perform the injection at positions away from the region of interest to be analyzed. For example, if a tail amputation is performed after electroporation to study the effects of a targeted gene in spinal cord regeneration, it is necessary to perform the injection where the site will be removed by amputation.
Entering the central canal with an injection capillary is the most critical step of this protocol, and new users are advised to practice before the actual experiment. Injection is best performed on smaller animals (less than 2.5 cm long) at the posterior end of the spinal cord, where it is easiest to position the capillary tip into the central canal. It is advised to observe the initial spread of the blue-colored RNP mix closely. There is the possibility that the RNP mix may be injected into the meningeal space around the spinal cord. A sharp and quick spread as a line along the middle of the spinal cord indicates successful injection into the central canal18. If nothing exits or the spread stops prematurely, it is recommended to keep the foot pedal pressed while moving the capillary slightly in and out. If this fails, it may be necessary to re-position the capillary or check for a clogged opening. For the RNP mix to spread into the brain ventricles, it is often necessary to increase the ejection pressure towards the end. Having the brain ventricles filled up is not necessary for electroporation of the spinal cord, but it is the best indication that the RNP mix has been injected into the central canal.
Electroporation considerations
It is advised to optimize the electroporation parameters beforehand to achieve sufficient delivery of the RNP to a majority of NSCs without causing extensive damage to other tissues or killing the animal. This protocol has been optimized for animals 2.0-2.5 cm long in size18. A reduction in transfer pulse voltage to 35 V is likely necessary when working with animals smaller than this. On the other hand, larger animals will require higher voltage and/or more pulses. In addition, the distance between the electrodes affects the efficiency. A reduction of the distance to 6 mm may be necessary for larger animals.
Divulgaciones
The authors have nothing to disclose.
Acknowledgements
We thank Prof. Elly M. Tanaka for her continuous and long-term support. This work was supported by a National Natural Science Foundation of China (NSFC) Grant (317716), Research Starting Grants from South China Normal University (S82111 and 8S0109), and a China Postdoctoral Science Foundation Grant (2018M633067).
Materials
| Agarose | Sigma-Aldrich | A9539 | |
| Benzocaine | Sigma-Aldrich | E1501-100G | |
| Benzocaine 0.03 % (wt/vol) | Mix 500 ml of 10× TBS, 500 ml of 400% (wt/vol) Holtfreter’s solution and 30 ml of 10% (wt/vol) benzocaine stock solution. Fill up the volume to 10 L with dH2O. The solution can be stored at room temperature for up to 6 months. | ||
| Benzocaine 10 % (wt/vol) | Mix 50 g of benzocaine in 500 ml of 100% (vol/vol) ethanol. The solution can be stored at room temperature for up to 12 months. | ||
| Borosilicate glass capillaries 1.2 mm O.D., 0.94 mm I.D. | Stutter Instrument | BF120-94-8 | |
| CaCl2·2H2O | Merck | 102382 | |
| CAS9 buffer, 10x | Mix 200 mM HEPES and 1.5 M KCl in RNase-free water. Adjust pH to 7.5. Filter sterilize, aliquot and store at −20 °C for up to 24 months | ||
| CAS9-NLS protein | PNA Bio | CP03 | |
| Cell culture dishes, 10cm | Falcon | 351029 | |
| Dumont #5 – Fine Forceps | Fine Scientific Instruments | 11254-20 | |
| Electroporator | Nepa Gene | NEPA21 | |
| BEX | Pulse Generator CUY21EDIT II | ||
| Fast Green FCF | Sigma-Aldrich | F7252-5G | |
| Fast Green FCF Solution, 5x | Dissolve 12.5 mg of Fast Green FCF powder in 10 mL of 1× PBS. | ||
| Flaming/Brown Micropipette Puller | Stutter Instrument | P-97 | |
| Holtfreter’s solution 400% (wt/vol) | Dissolve 11.125 g of MgSO4·7H2O, 5.36 g of CaCl2·2H2O, 158.4 g of NaCl and 2.875 g of KCl in 10 L of dH2O. The solution can be stored at room temperature for up to 6 months. | ||
| KCl | Merck | 104936 | |
| MgSO4·7H2O | Merck | 105886 | |
| Microloader pipette tips | Eppendorf | 5242956003 | |
| Micromanipulator | Narishige | MN-153 | |
| NaCl | Merck | 106404 | |
| Pneumatic PicoPump | World Precision Instruments | SYS-PV830 | |
| Ring Forceps | Fine Scientific Instruments | 11103-09 | |
| Stereomicroscope | Olympus | SZX10 | |
| Tris base | Sigma-Aldrich | T6066 | |
| Tris-buffered saline, 10x | Dissolve 24.2 g of Tris base and 90 g of NaCl in 990 ml of dH2O. Adjust pH to 8.0 by adding 10 ml of 37% (vol/vol) HCl. The solution can be stored at room temperature for up to 6 months. | ||
| Tweezers w/Variable Gap 2 Round Platinum Plate Electrode, 10mm diameter | Nepa Gene | CUY650P10 |
Referencias
- O’Hara, C. M., Egar, M. W., Chernoff, E. A. G. Reorganization of the ependyma during axolotl spinal cord regeneration: Changes in intermediate filament and fibronectin expression. Developmental Dynamics. 193 (2), 103-115 (1992).
- Mchedlishvili, L., Mazurov, V., Tanaka, E. M. Reconstitution of the Central Nervous System During Salamander Tail Regeneration from the Implanted Neurospheres. Plant, Soil and Environment. 916 (8), 197-202 (2012).
- Nordlander, R. H., Singer, M. The role of ependyma in regeneration of the spinal cord in the urodele amphibian tail. Journal of Comparative Neurology. 180 (2), 349-373 (1978).
- Fei, J. F., et al. Tissue- and time-directed electroporation of CAS9 protein–gRNA complexes in vivo yields efficient multigene knock-out for studying gene function in regeneration. npj Regenerative Medicine. 1 (1), 16002 (2016).
- Fei, J. F., et al. CRISPR-mediated genomic deletion of Sox2 in the axolotl shows a requirement in spinal cord neural stem cell amplification during tail regeneration. Stem Cell Reports. 3 (3), 444-459 (2014).
- Flowers, G. P., Timberlake, A. T., McLean, K. C., Monaghan, J. R., Crews, C. M. Highly efficient targeted mutagenesis in axolotl using Cas9 RNA-guided nuclease. Development. 141 (10), 2165-2171 (2014).
- Flowers, G. P., Sanor, L. D., Crews, C. M. Lineage tracing of genome-edited alleles reveals high fidelity axolotl limb regeneration. eLife. 6, (2017).
- Fei, J. -. F., et al. Efficient gene knockin in axolotl and its use to test the role of satellite cells in limb regeneration. Proceedings of the National Academy of Sciences. 201706855. , (2017).
- Nowoshilow, S., et al. The axolotl genome and the evolution of key tissue formation regulators. Nature. 554 (7690), 50-55 (2018).
- Bryant, D. M., et al. A Tissue-Mapped Axolotl De Novo Transcriptome Enables Identification of Limb Regeneration Factors. Cell Reports. 18 (3), 762-776 (2017).
- Smith, J. J., et al. A chromosome-scale assembly of the axolotl genome. Genome Research. 373548, (2019).
- Smith, J. J., et al. Sal-Site: Integrating new and existing ambystomatid salamander research and informational resources. BMC Genomics. 6, 1-6 (2005).
- Campbell, L. J., et al. et al Gene expression profile of the regeneration epithelium during axolotl limb regeneration. Developmental Dynamics. 240 (7), 1826-1840 (2011).
- Stewart, R., et al. Comparative RNA-seq Analysis in the Unsequenced Axolotl: The Oncogene Burst Highlights Early Gene Expression in. the Blastema. PLoS Computational Biology. 9 (3), (2013).
- Fei, J. -. F., et al. Application and optimization of CRISPR–Cas9-mediated genome engineering in axolotl (Ambystoma mexicanum). Nature Protocols. 13 (12), 2908-2943 (2018).
- Holder, N., et al. Continuous growth of the motor system in the axolotl. Journal of Comparative Neurology. 303 (4), 534-550 (1991).
- Tazaki, A., Tanaka, E. M., Fei, J. F. Salamander spinal cord regeneration: The ultimate positive control in vertebrate spinal cord regeneration. Biología del desarrollo. 432 (1), 63-71 (2017).
- Albors, A. R., Tanaka, E. M. High-Efficiency Electroporation of the Spinal Cord in Larval Axolotl. Salamanders in Regeneration Research: Methods and Protocols. 1290, 115-125 (2015).
- Albors, A. R., et al. et al Planar cell polarity-mediated induction of neural stem cell expansion during axolotl spinal cord regeneration. eLife. 4, 1-29 (2015).
- Cong, L., et al. Multiplex Genome Engineering Using CRISPR/Cas Systems. Science. 339 (6121), 819-823 (2013).
- Doench, J. G., et al. Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9. Nature Biotechnology. 34 (2), 184-191 (2016).
- Graham, D. B., Root, D. E. Resources for the design of CRISPR gene editing experiments. Genome Biology. 16 (1), 260 (2015).
- Khattak, S., et al. Optimized axolotl (Ambystoma mexicanum) husbandry, breeding, metamorphosis, transgenesis and tamoxifen-mediated recombination. Nature Protocols. 9 (3), 529-540 (2014).
