Abstract
In zebrafish, hematopoietic stem cells (HSCs) are born in the developing aorta during embryogenesis. From the definitive wave of hematopoiesis onward, blood homeostasis relies on self-renewal and differentiation of progeny of existing HSCs, or clones, rather than de novo generation. Here, we describe an approach to quantify the number and size of HSC clones at various times throughout the lifespan of the animal using a fluorescent, multicolor labeling strategy. The system is based on combining the multicolor Zebrabow system with an inducible, early lateral plate mesoderm and hematopoietic lineage specific cre driver (draculin (drl)). The cre driver can be temporally controlled and activated in early hematopoiesis to introduce a color barcoding unique to each HSC and subsequently inherited by their daughter cells. Clonal diversity and dominance can be investigated in normal development and blood disease progression, such as blood cancers. This adoptable method allows researchers to obtain quantitative insight into clonality-defining events and their contribution to adult hematopoiesis.
Keywords: Hematopoietic stem cells (HSCs), Zebrabow system, Colorimetric barcoding, Cre-loxP recombination, Lineage analyses, In vivo multicolor imaging
1. Introduction
The hematopoietic system is populated from a limited number of hematopoietic stem cells (HSCs) that are seeded in the adult marrow early in development. These HSCs replenish multilineage progenitors, which give rise to all differentiated blood cells. Normal function and output of HSCs ensure a balanced production of all peripheral blood cell types throughout the lifespan. Because there is no additional de novo generation of HSCs after early development [1], tools can be developed that label individual HSCs, or clones, early in development and allow for assessment of clones in the adult hematopoietic system [2]. Production in the number and type of daughter cells from individual HSC clones fluctuates throughout the lifespan. Progressive clonal expansion of an HSC that harbors a mutation is the basis for diverse human cancers and has been linked to a more than ten-fold increase in the risk of developing a blood cancer [3, 4]. An overrepresentation of progeny from a limited number of HSCs also arises under a number of adverse conditions such as after irradiation [2] and with advanced aging [3-5]. Hence, understanding the nature of clonal dynamics in normal and malignant cell populations is essential for the rational development of therapeutics to prevent HSC clonal dominance-related diseases.
To quantify clonal dynamics and the number of HSCs present in the zebrafish over time, the Zon laboratory adapted the Zebrabow methodology [6], specifically to the hematopoietic system. In the Zebrabow system, three fluorescent proteins (dTomato, cyan fluorescent protein (CFP), and yellow fluorescent protein (YFP)) are driven by a ubiquitous promoter and flanked by lox sites. The Zebrabow transgene expresses dTomato in its default state and upon stochastic recombination by cre or creERT2, either CFP or YFP is exclusively expressed. To generate a more diverse repertoire of colors, multiple independent integrations of the Zebrabow construct in the genetic background were established, which leads to different expression levels of each fluorophore and therefore multiple, unique color barcodes. The resulting unique color barcodes are inherited by all descendants of the labeled cell [2]. This system allows for clonal tracking of HSCs without the need for additional perturbation such as transplantation. Hence, the clonal dynamics of each HSC can be analyzed and its trajectory can be reconstructed. The Zon laboratory combines the Zebrabow system [6] with the draculin (drl) promoter [7] to drive Cre recombinase expression specifically for tracking HSCs and their progeny.
Using this robust method has led to several discoveries about HSC biology. For example, Henninger et al. revealed that the entire adult hematopoietic system in zebrafish is populated from approximately 21 HSCs [2]. Additionally, this study demonstrated that stressors, including sublethal irradiation and transplantation, reduce clonal diversity and lead to clonal selection [2]. More recently, the Zon laboratory has paired this system with mosaic mutagenesis using the CRISPR/Cas9 system to examine the effect of genetic perturbations on clonal hematopoiesis in adult zebrafish [8]. Furthermore, this system can be combined with fluorescence-activated cell sorting to prospectively isolate different clonal populations for downstream analysis such as single-cell RNA sequencing and DNA sequencing [8].
The strength and versatility of multicolor genetic labeling in zebrafish provides a well-established tool that can be readily adapted. The system is a powerful method to evaluate clonal diversity in both native and malignant settings and to better understand the genetics underlying these conditions. Using different Cre combinations and mosaic gene knockouts, experiments can be designed covering a wide array of approaches to suit the biological question. Thus, the Zebrabow system provides a resource for systematic anatomical and lineage studies during zebrafish blood development.
2. Materials
2.1. Genotyping of Tg (ubi:Zebrabow-M)
gDNA from fin clip (as previously described https://zfin.atlassian.net/wiki/spaces/prot/pages/356155929/Fin+Amputations).
Qiagen blood and tissue kit for gDNA extraction (Qiagen, Cat. #69504).
QX200 AutoDG Droplet Digital PCR System (automated droplet generation) or QX200 Droplet Generator (manual droplet generation).
Manual droplet generation: DG8 Cartridges for QX100/QX200 Droplet Generator (Biorad, Cat. # 186-4008).
Manual droplet generation: Droplet Generator DG8 Gaskets (Biorad, Cat. # 186-3009).
Manual droplet generation: Droplet Generation Oil for Probes (Biorad, Cat. # 186-3005).
Automated droplet generation: Pipet Tips for Automated Droplet Generator (Biorad, Cat. # 186-4120).
Automated droplet generation: DG32 Cartridge for Automated Droplet Generator (Biorad, Cat. # 186-4108).
Automated droplet generation: Automated Droplet Generation Oil for Probes (Biorad, Cat. # 186-4110).
twin.tec PCR Plate 96-well LoBind semi-skirted (Eppendorf, Cat. # 30129504).
ddPCR Supermix for Probes (no dUTP) (Biorad, Cat # 186-3024).
Pierceable Foil Heat Seal (Biorad, Cat. # 181-4040).
ddPCR Reader Oil (Biorad, Cat # 186-3004).
96-well generator tip boxes.
Green droplet cassettes.
Chilled 96-well adaptor for destination plate.
Ice bucket + chilled 96-well heat block.
FAM Probe standard diluted to 20× using nuclease free water (gapdh).
HEX Probe standard diluted to 20× (tdTomato for ubi:zebrabow) (see Note 1).
2.2. Zebrafish
2.3. creERT2-Mediated Recombination
4-Hydroxytamoxifen (4-OHT) solution (3.86 mg/mL in 100% ethanol, protect from light and vortex at RT for 20 min to produce a 10 mM stock. Store aliquots for single use at −20 °C) (Sigma, H7904) (see Note 2).
Pronase (Roche, Cat. # 10165921001; 50 mg/mL).
6-well plastic cell culture dishes.
15 or 50 mL conical tubes.
Aluminum foil.
2.4. Cardiac Bleed and Kidney Marrow Collection
Iced water.
Fish net.
Heparin (heparin sodium salt from porcine intestinal mucosa, Sigma, Cat. # 2106).
1.5 mL microtubes.
Blood buffer (for 250 mL blood buffer: 225 mL 1% PBS, 5 mL FBS (fetal bovine serum), sterile filter, and store at 4 °C. Immediately add heparin prior to use to 1 USP units/mL).
Surgical tools (forceps, needles, and scissors).
P10 pipette and tips.
Dissecting microscope.
Styrofoam board with 30g needles.
2.5. Analysis of the Recombined Kidney Marrow
Blood buffer (0.9% PBS, 5% FBS, 1000× heparin, as described in Subheading 2.4).
Collected kidney marrow samples suspended in blood buffer.
FACS tubes (Falcon 5 mL Round Bottom Polystyrene Test Tube, with Cell Strainer Snap Cap; Falcon, Cat. # 352235).
DRAQ7 live cell stain (a far-red fluorescent dye that only stains the nuclei in dead and permeabilized cells; Abcam, Cat. # ab109202).
P1000 pipette and P10 pipette.
Flow cytometer with 440, 488, 561, and 640 laser lines (e.g., BD LSR Fortessa).
FlowJo software (Becton, Dickinson & Company).
Zebrabow Color Analysis via Zebrabow App or MATLAB code (see Github link: https://github.com/jehenninger/2020_Avagyan_Henninger_et_al).
3. Methods
3.1. Genotyping of Adult Tg(ubi: Zebrabow-M)
Fin clip the adult Tg(ubi:Zebrabow-M) zebrafish [6] (Fig. 1a; as previously described). As a control, include a zebrafish with a single insertion and one with no insert (wild type).
Perform the gDNA extraction using the Qiagen blood and tissue kit (Qiagen, Cat. #69504). Follow the manual instructions using 180 μL ALT buffer plus 20 μL proteinase K solution per fin clip and elute in 50 μL nuclease-free water.
For optimal ddPCR results, dilute the gDNA solution to 10–15 ng/μL with nuclease-free water.
Thaw 2× Supermix on ice (to avoid freezing and thawing cycles aliquot to 1 mL and freeze).
Pipette the PCR reaction with the gDNA samples into one 96-well Eppendorf plate placed in a 96-well heat block on ice using the volumes described in Table 1.
Pipette 22.5 μL master mix into each well and add 2.5 μL gDNA into the respective well. Mix via pipetting up and down five times. Steps 7–12 describe the automated droplet generation and steps 13–19 the manual droplet generation.
-
Automated droplet generation:
Place the sample plate into the correct position (see manual) and make sure the plate is trapped within the clamps.
Place a clean destination plate into a chilled 96-well adaptor and place it into the correct position (green light under the block appears).
Place up to two generator tip boxes into the correct positions.
Place up to three sets of droplet generator cassettes into the correct positions. The green side of the cassette faces the left side (green light appears).
Close the automated droplet generator. On the front panel, select the correct number of samples and wait for the initialization to determine the time to finish the droplet generation.
Heat seal the destination plate using 180 °C for 5 s and place it into a thermocycler. Make sure that the 96-well plate will fit into the thermocycler.
-
Manual droplet generation:
Put cartridge into the holder with the flat end to the right and close it properly.
Transfer 20 μL reaction into the middle wells of the cartridge (touch the tip to the bottom and pull up as dispensing the solution to avoid bubbles). Be careful that no bubbles are in the center bottom of the well over the port since droplets will not be generated. Samples in all eight wells are needed to induce the generation.
Transfer 70 μL oil into the oil wells.
Put gasket over four posts.
Open the QX200 door, put cartridge in (unidirectional), and close the door (droplet generation begins automatically and takes 2 min per 8 samples).
Remove cartridge, discard gasket, and gently transfer the 40 μL of emulsion into the 96-well twin-tec plate. Pipette slowly and at an angle so as not to shear droplets (air bubbles are fine at this point). Make sure that the 96-well plate will fit into the thermocycler.
Heat seal the destination plate using 180 °C for 5 s and place it into the thermocycler.
Run the following PCR program in a thermocycler (Table 2). After the PCR, the samples are stable at 4 °C for up to 24 h.
Read the plate on QX200 plate reader and in the QuantaSoft software setup Ch1 as gapdh and Ch2 as dTomato with dye set FAM/HEX.
Choose CNV2 (indicates wild-type copy) and setup a template labeling all parameters.
Export a CNV file for a later analysis using Excel or analyze within QuantaSoft (see Note 3).
Analyze copy number: Divide the concentration of tdTomato with the gapdh concentration times 2 (tdTomato conc./gapdh conc. × 2 = copy number).
Keep zebrafish with more than 20 insertions/copies (Fig. 1b).
Fig. 1.
Genotyping strategy for Tg(ubi:Zebrabow-M) zebrafish. (a) Schematic of the Zebrabow color cassette depicting the location of the ddPCR primer location within the tdTomato sequence. (b) Number of insertions are calculated from the event reads for each sample. Insertions vary and zebrafish with more than 20 insertions should be kept. The single insertion has one insertion detected and the wild-type control has no insertions, which indicated a successful genotyping procedure
Table 1.
ddPCR reaction mix for Tg(ubi:Zebrabow-M) genotyping
| Volume per well/reaction (μL) | Mix for a 96-well plate (+10%) (μL) | |
|---|---|---|
| Super mix | 12.5 | 1320 |
| Probe 1 (20×; gapdh) | 1.25 | 132 |
| Probe 2 (20×; tdTomato) | 1.25 | 132 |
| DNA sample (15–25 ng/μL) | 2.5 | – |
| Water | 2.5 | 792 |
| Total | 25 | 2376 |
Table 2.
PCR program for Tg(ubi:Zebrabow-M) genotyping
| 95 °C | 10 min | |
|---|---|---|
| 94 °C | 30 s | 40 cycles |
| 60 °C | 60 s | |
| 98 °C | 10 min | |
| 4 °C | Forever |
3.2. Genotyping of Tg (drl:creERT2) [7]
Cross a potential Tg(−6.3drl:creERT2,alpha-crystallin:ϒFP) adult zebrafish with a wild-type zebrafish.
Let the embryos develop until alpha-crystallin:ϒFP is strongly expressed in the eye from approximately 24 h postfertilization (hpf) (green eyes indicate a positive embryo).
Select all embryos with green eyes and grow them to adulthood. Adult Tg(−6.3drl:creERT2,alpha-crystallin:ϒFP) zebrafish can be identified using a fluorescent binocular with a GFP filter.
3.3. Zebrabow Fish Breeding and Color Induction
To generate the experimental zebrafish, cross Tg(ubi:Zebrabow-M) [6] zebrafish with Tg(drl:creERT2) [7] zebrafish (Fig. 2).
Collect the fertilized embryos in a 10 cm dish containing E3 medium (if desired, a genetic intervention may be introduced at the one-cell stage following standard protocols such as CRISPR-induced mutations [9] or morpholino knockdown [10]).
At 24 hpf, embryos can be either dechorionated manually using forceps or using a Pronase treatment (Use 1 mL of the Pronase stock solution for a 10 mL working solution of 1 mg/mL. Pour the working solution into a small petri dish and place the embryos still in their chorion into the solution. Swirl the embryos until the chorion becomes soft and embryos start to dechorionate, typically after 3–5 min at 24 hpf. Immediately transfer the embryos into a petri dish with fresh E3 medium. Wash three times in fresh E3 medium.).
The dechorionated embryos should be placed at a density of 25–35 embryos per well into a 6-well dish.
Based on the number of embryos to be treated, create a 15 μM mixture of 4-OHT (rehydrated in ethanol) in E3 medium in a tin foil-covered conical tube (protected from light).
Using a thin-tipped pipette, remove all E3 media from the 6-well dish without disturbing embryos and immediately replace with 6 mL per well of the 15 μM 4-OHT working solution.
Wrap the 6-well plate in aluminum foil and incubate at 28.5 °C for 4 h.
After 4 h, remove the 6-well plate from the incubator and transfer the 4-OHT treated embryos to a 10 cm dish in fresh E3 media.
Sort the embryos using a fluorescent binocular for green expression of cryaa:GFP in the eyes (drl:creERT2 genotyping) and a red ubiquitous expression of tdTomato (ubi:Zebrabow-M genotyping).
Keep the embryos at 28.5 °C in an incubator until analysis or transfer the zebrafish to your facility’s nursery.
Fig. 2.
Experimental workflow for a Zebrabow analysis. Double transgenic Tg(ubi:Zebrabow-M);Tg(drl:creERT2) single-cell embryos are collected and a genetic manipulation could be performed. At 24 hpf, 4-OHT treatment is performed for 4 h to induce the color recombination in drl-CreERT2 positive cells. After growing the zebrafish to adulthood, the color barcodes can be analyzed via flow cytometry and clones of interest could be sorted for further analysis using RNA- or DNAseq techniques
3.4. Determine Hematopoietic Clone Number and Clone Size
Isolate the zebrafish kidney marrow by first, euthanizing the zebrafish via ice-cold water submersion; second, making a lateral cut along the length of the underside of the zebrafish using a micro-scissor. Third, make two parallel incisions to the initial cut so that the skin of the zebrafish can be fixed to a Styrofoam board with 30g needles exposing the animal’s innards.
Remove internal organs leaving only the kidney marrow exposed using fine tip forceps. Collect the kidney marrow using fine tip forceps and place into 300 μL blood buffer on ice until all samples are ready for the analysis via flow cytometry. A total of 300 μL volume is recommended for a 3-month postfertilization (mpf) zebrafish, and it should be modified depending on the size of the kidney marrow.
Immediately prior to analysis, pipette the isolated marrow in blood buffer with a 1000 μL pipette 30 s up and down to mechanically dissociate the kidney marrow into a single cell suspension and to release the hematopoietic cells. Slowly pipette the suspended kidney marrow through the cell strainer attached to a FACS tube.
Add 1 μL DRAQ7 into the mixture prior to analysis via flow cytometry. DRAQ7 is used as a viability stain and detected in far-red channel, thereby not interfering with the Zebrabow channels/colors.
Samples should be run through a flow cytometer equipped with an 85 μm filter and lasers detecting dTomato, YFP, and CFP protein. An example of the gating strategies is presented in Fig. 3 (see Note 4).
Record a minimum of 20,000 myelomonocyte events for a robust analysis of the color barcoding. Import the .fcs files from the flow cytometer into the FlowJo software.
Add a compensation matrix containing each color to be analyzed to the sample.
Process the data using FlowJo as depicted in Fig. 3. Using a “OR GATE” or Boolean gate function, create a population that includes cells expressing “blue and green” or “red” and export the compensated colors exclusively (see Note 5).
Download the Zebrabow Color Analysis Application here: https://github.com/jehenninger/2020_Avagyan_Henninger_et_al. Blind the files for an unbiased analysis of color clusters.
Load the exported .fcs file into the Zebrabow Color Analysis Application and observe the clustering solutions provided. A full description of the software can be found here: https://github.com/jehenninger/2020_Avagyan_Henninger_et_al. Representative plots obtained by the program can be observed in Fig. 4.
Evaluate clustering solutions produced by the Zebrabow Color Analysis Application and fuse or split clusters that were falsely labeled using the embedded functions, respectively.
Create a destination folder and save the files. Saving the files will generate folders with all analyzed data including images (Fig. 4) and clone size information.
Using the backgating images provided by the Zebrabow Color Analysis Application, identified clusters can be isolated for downstream applications such as transplantation, DNA sequencing, RNA sequencing, or metabolic assays.
Fig. 3.
Representative flow cytometry gating strategy to prepare data for the Zebrabow Color Analysis Application. An example of flow cytometry data obtained from a single, dissociated kidney marrow of a Tg(ubi:Zebrabow-M);Tg(drl:creERT2) zebrafish treated with 4-OHT at 24 hpf. The gating strategy is depicted from left to right as indicated by the arrows. Single cells are identified via FSC-A and FSC-H and further by SSC-W and SSC-H, live cells are identified by the exclusion of DRAQ7 staining in the APC-Cy7 channel, and myelomonocytes are identified by their characteristic FSC-A and SSC-A properties. From the myelomonocyte gate, tdTomato, CFP, and YFP are used to identify all cells for analysis in the Zebrabow Color Analysis Application. Two gates are separated, either cells expressing CFP detected in the DAPI channel and/or YFP detected in the FITC and cells expressing tdTomato detected in the dsRED channel. From these gates a Boolean gate is generated, indicating all cells that will be analyzed by the Zebrabow Color Analysis Application. FSC-A Forward scatter area, FSC-H forward scatter height, SSC-A side scatter area, SSC-H side scatter height, SSC-W side scatter width, Comp compensation
Fig. 4.
Example graphs obtained using the Zebrabow Color Analysis Application. (a) A three-dimensional rendering of myelomonocytes from the kidney marrow of a 3 mpf Tg(ubi:Zebrabow-M);Tg(drl:creERT2) zebrafish, 4-OHT treated at 24 hpf. (b) An individual cluster is highlighted in two-dimensional CFP/YFP and RFP/YFP plots. These plots can be used to inform gating strategies to sort specific clones using a flow cytometer. (c) The expression level of tdTomato, CFP, and YFP of each myelomonocyte is plotted in the ternary plot, respectively. (d/e) Two different methods of graphical representation are depicted, representing the percentage of analyzed cells belonging to each annotated color cluster
4. Notes
Probe information sheet, see Table 3.
4-OHT is generally unstable and light sensitive (work quickly and keep the solution protected from light).
- Data will only be collected for wells marked as containing samples. It is normal to see “no call” initially. Positive signals for FAM/HEX ≈12,000 and negative ≈2000.
- If the droplet reader shows so-called “rain” between the positive and negative droplets on the ϒ axis, the annealing temperature should be increased.
In some instances, a 405 or 488 nm laser may be used for CFP detection. However, for optimal detection of CFP signal and subsequent cluster analysis, optimal detection using a 440 or 458 nm laser is recommended.
For the analysis, the compensation step is mandatory for Zebrabow Color Analysis Application to accept the data. Exclude samples that have ≥25% of cells only expressing tdTomato as this is an indication that the cre recombination was inefficient.
Table 3.
Probe information for Tg(ubi:Zebrabow-M) genotyping
| Gene | Probe seq | 5′ Reporter | Int | 3′ Quencher | Primer 1 | Primer 2 |
|---|---|---|---|---|---|---|
| dTomato | CTACGTGGA [ZEN]CACC AAGCTGG ACAT | HEX | ZEN | Iowa Black FQ | GATGGTGTAG TCCTCGTTG TG | CACTACCTGG TGGAG TTCAAG |
| zf-gapdh | AAGGGTGAG [ZEN]GTTAA GGCAGAAGGC | 6-FAM | ZEN | Iowa Black FQ | GCATGACCA TCAA TGACCAG TTTG | AGTGCTTG TTTC TTCACAGG TTTAC |
Acknowledgments
We would like to thank Serine Avagyan and Jonathan Henninger for technical support and fruitful discussions. Figures were created with BioRender.com under an academic license. D.B. was supported by the German National Academy of Sciences Leopoldina project LPDS 2021-01. B.G. was supported by a Canadian Institutes for Health Research Postdoctoral Fellowship. Research reported in this publication was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health under Award Number U24DK126127. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
References
- 1.Bertrand JY, Chi NC, Santoso B, Teng S, Stainier DYR, Traver D (2010) Haematopoietic stem cells derive directly from aortic endothelium during development. Nature 464(7285):108–111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Henninger J, Santoso B, Hans S, Durand E, Moore J, Mosimann C, Brand M, Traver D, Zon L (2017) Clonal fate mapping quantifies the number of haematopoietic stem cells that arise during development. Nat Cell Biol 19:17–27 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sperling AS, Gibson CJ, Ebert BL (2016) The genetics of myelodysplastic syndrome: from clonal haematopoiesis to secondary leukaemia. Nat Rev Cancer 171(17):5–19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bowman RL, Busque L, Levine RL (2018) Clonal hematopoiesis and evolution to hematopoietic malignancies. Cell Stem Cell 22:157–170 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jaiswal S, Fontanillas P, Flannick J, Manning A, Grauman PV, Mar BG, Lindsley RC, Mermel CH, Burtt N, Chavez A, Higgins JM, Moltchanov V, Kuo FC, Kluk MJ, Henderson B, Kinnunen L, Koistinen HA, Ladenvall C, Getz G, Correa A, Banahan BF, Gabriel S, Kathiresan S, Stringham HM, McCarthy MI, Boehnke M, Tuomilehto J, Haiman C, Groop L, Atzmon G, Wilson JG, Neuberg D, Altshuler D, Ebert BL (2014) Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med 371:2488–2498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Albert Pan Y, Freundlich T, Weissman TA, Schoppik D, Cindy Wang X, Zimmerman S, Ciruna B, Sanes JR, Lichtman JW, Schier AF (2013) Zebrabow: multispectral cell labeling for cell tracing and lineage analysis in zebrafish. Development 140:2835–2846 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mosimann C, Panáková D, Werdich AA, Musso G, Burger A, Lawson KL, Carr LA, Nevis KR, Sabeh MK, Zhou Y, Davidson AJ, Dibiase A, Burns CE, Burns CG, Macrae CA, Zon LI (2015) Chamber identity programs drive early functional partitioning of the heart. Nat Commun 6:8146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Avagyan S, Henninger JE, Mannherz WP, Mistry M, Yoon J, Yang S, Weber MC, Moore JL, Zon LI(2021) Resistance to inflammation underlies enhanced fitness in clonal hematopoiesis. Colorful Clones in the Blood Science 374(6568):768–772. 10.1126/science.aba9304 [DOI] [PubMed] [Google Scholar]
- 9.Ablain J, Durand EM, Yang S, Zhou Y, Zon LI (2015) A CRISPR/Cas9 vector system for tissue-specific gene disruption in zebrafish. Dev Cell 32:756–764 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Brent LJ, Drapeau P (2002) Targeted ‘knock-down’ of channel expression in vivo with an antisense morpholino oligonucleotide. Neuroscience 114:275–278 [DOI] [PubMed] [Google Scholar]