A simple and rapid shaking-based assay to genotype live, early developmental stage zebrafish embryos

Abstract

The use of zebrafish as an animal model for biomedical and toxicological research has increased dramatically over the past decade, alongside a growing need to adopt the 3Rs principles to ensure ethically acceptable animal experimentation. Currently, one of the main challenges concerns ‘surplus’ animals that are unavoidably generated as part of an experimental procedure and are unsuitable for experimental analysis because they do not have the desired genotype, are too old or have the wrong sex. However, justifying the sacrifice of animals for these reasons is morally debatable and current ethics legislation in some countries insists they should nevertheless be maintained and left to die of natural causes. It is therefore imperative to develop strategies which can identify unwanted animals at a sufficiently early, non-sentient developmental stage so that they can then be sacrificed in an ethically more acceptable manner. In this manuscript we present a reliable medium-throughput method for non-invasive genotyping of zebrafish at developmental stages when sacrifice is considered ethically acceptable. This method is based on the use of low frequency shaking to induce the detachment of a limited number of cells from the embryos. These cells are then analysed by polymerase chain reaction-based genotyping approaches.

Background

The zebrafish (Danio rerio) was first identified as an ideal animal model for laboratory research owing to the small size and transparency of its embryos, its high fecundity and extra-uterine development. ¹ Specifically, this species is particularly attractive for the non-invasive study of embryogenesis. ² However, zebrafish also share many aspects of physiology with other vertebrates, including humans, making them an ideal model for the study of many pathophysiological processes with the goal of developing new therapeutic approaches.^3,4 Indeed, the zebrafish has become a powerful model for many other research fields ranging from environmental toxicology ⁵ to cancer and genetics.^6,7 Consequently, the number of zebrafish used for research has increased enormously during the past two decades. However, there is also a growing urgency to minimize, wherever possible, this number by employing the 3Rs rule (Reduce, Refine, Replace).

The development of new protocols to generate fish cell cultures from fin biopsies of adults or early developmental stages provides complementary ex vivo zebrafish models that can serve as an alternative to live animal experimentation for drug screening and to study important biochemical pathways.^8–12 However, although the use of such alternative methods can substantially reduce the number of live animals required for research projects, there are still far more animals raised than are theoretically required. In 2017 in the EU, 12.6 million surplus animals were generated in research projects compared with the 9.4 million actually used in experiments, thereby increasing animal facility running costs. Of this number, 7% were zebrafish.^13–15 This situation has sparked an important discussion regarding the ethical status of these surplus animals. One of the foundational principles of animal protection is not to inflict pain, suffering or harm on animals without ‘reasonable cause’. Therefore, does the sacrifice of surplus animals that are not considered useful for research constitute a reasonable cause? This discussion is already leading to changes in legislation and, for example, in Germany, laws on animal welfare have become stricter and since 2021 now require the declaration each year of the number of surplus animals that have been sacrificed.^16,17 This situation challenges the research community to develop new strategies for minimizing the number of surplus animals whose presence leads to increased costs in order to maintain an adequate welfare status.

In the case of zebrafish, as for rats and mice, in order to establish and maintain genetically modified lines, large numbers of surplus animals with undesired genotypes are inevitably generated in the F0 to F2 generations before the correct line is stably established. This is also the case for the maintenance of mutant lines which exhibit embryonic lethality, are unfertile or exhibit adverse phenotypes in the homozygous state. These lines can only be propagated by heterozygote crosses, which, as well as generating the desired homozygous mutant, also produce large numbers of ‘undesired’ wildtype (WT) and heterozygous animals. As a consequence, the large number of animals generated in this way can be reduced only by strict planning of crosses and by reducing to a minimum the number of breeding events. Here, we present a new genotyping protocol that has been specifically developed in zebrafish to further reduce the number of animals raised during the establishment and maintenance of genetically modified lines. It is based on the rationale that if the genotype of each fish can be determined accurately during early embryonic non-sentient developmental stages prior to the stage of self-feeding (the age according to EU law from which the zebrafish becomes sentient and so is protected according to ethical considerations), it is possible to limit the number of fish raised to only those possessing the desired genotype. The onset of self-feeding in zebrafish is defined as 120 h post fertilization (hpf) ¹⁸ and is considered the stage when nervous system function has fully matured and the animal is able to perceive pain and distress. Until now, protocols for genotyping living animals at a high or medium throughput scale have been restricted mainly to juveniles and adults by employing fin biopsy or non-invasive skin swab techniques.^19–21

In this paper, we present a genotyping method based on the use of low frequency shaking to induce the detachment of a limited number of cells from embryos which are then analysed by a polymerase chain reaction (PCR)-based genotyping assay. Application of this experimental approach can be easily scaled up to the medium or high throughput level. Importantly, our method permits the sacrifice of embryos with undesirable genotypes before they reach the 120 hpf stage and thereby limits the number of surplus animals that are raised at each generation. Furthermore, this method represents a replacement for the fin clip genotyping method in adult and juvenile animals, which is considered a harmful and pain-inducing procedure.

Zebrafish embryos

• Embryos obtained by breeding AB strain fish (WT);

• Embryos from the rtel1 mutant line obtained by breeding heterozygous fish (manuscript in preparation);

• Embryos from the mlxip^ka405 (mondoa) mutant line obtained by breeding heterozygous fish. ²² The ZFIN ID of the mondoa/mlxip^ka405 mutant line is ZDB-ALT-180628-2.

Material

• Petri dishes (Greiner Bio-One, Cellstar ref. 632102);

• Ninety-six-well cell culture plates with flat bottoms (Greiner Bio-one, Cellstar 655 180);

• Plastic Pasteur pipettes (VWR, ref. 612-1681);

• Pipette tips 10 µl (SARSTED, Germany ref. 70.3050.200);

• Pipette tips 200 µl (SARSTED, Germany ref. 70.3030.100);

• Pipette tips 1000 µl (SARSTED, Germany ref. 70.3010.255);

• Ninety-six-well PCR plates (PeqLab, ref. 82-0600-A);

• Q-PCR Clear Seal Klebefilm (Steinbrenner Laborsysteme GmbH, Germany, ref SLAM0560);

• Injection needles TW100F-4 (World Precision Instruments).

Reagents

• 1× E3 medium: 5mM NaCl (Roth, ref. 3957.1); 0.17 mM KCl (Roth, ref. 6781.3); 0.33 mM CaCl₂ (Roth, ref. 5239.2); 0.33 mM MgSO₄ (Roth, ref. P027.1);1

• Methylene blue stock 1.0% (Sigma Aldrich, ref. 03978);

• Sodium hypochlorite 12% (Roth ref. 9062.3);

• D-PBS 1× (Dulbecco’s Phosphate Buffered Saline (CaCl₂, MgCl₂), Gibco, ref. 14190-094);

• Pronase from Streptomyces griseus (Sigma-Aldrich, ref. 10165921001);

• Proteinase K (Thermo Fisher Scientific, ref. EO0491);

• Tris-HCl pH 8 (Tris: Roth ref. 4855.2; HCl 37% Roth);

• KCl (Roth ref. 6781.3);

• 5× PK Lysis buffer for genomic DNA extraction:

• 50mM Tris-HCl pH 8;

• 250mM KCl;

• 1.5%Tween20 (stock concentration 100%, Roth ref. 9127.2);

• 0.025 mg/ml PK;

• PCR reactions:

• GoTaq Polymerase and GoTaq Green buffer (Promega, ref. M7845);

• dNTPs 10 mM Mix (Promega, ref. U1511);

• PCR primers (Sigma-Aldrich);

• DNA-ladder 100 bp (New England BioLabs, ref. N3231S);

• BslI restriction enzyme (New England BioLabs, ref. R0555S);

• Agarose, universal PeqGold, (VWR, Life Science, ref. 35-1020);

• Ethidium Bromide 10 mg/ml (Roth, ref. 2218.1);

• MEGAshortscript™ T7 Transcription Kit (Invitrogen);

• TrueCut™ Cas9 Protein v2 (Invitrogen).

Equipment

• Zeiss Stemi 2000 stereomicroscope;

• Olympus SZK7 stereomicroscope with Olympus EP50 camera;

• Incubator, Binder KBW, GmbH, ref. 9020-0090;

• Unithermix 2 pro, LLG Labware equipped with Plate96 block;

• Gene Amp PCR System 9700 thermocycler (Applied Biosystems);

• SimpliAmp thermal cycler A24811 (Thermo Fisher Scientific);

• Gel electrophoresis power supply (Biorad, Power/PAC3000);

• Gel electrophoresis tanks (VWR model 700-0777, USA);

• Gilson, P20, P200 and P1000 pipettes;

• Electronic multichannel pipettes INTEGRA;

• E-BOX gel-documentation imaging system (Vilbert);

• Injector MINJ-D (Tritech Research).

Methods

Husbandry, breeding and collection of zebrafish eggs

Fish were maintained under 14:10 h light:dark cycle conditions in a water recirculation system according to standard methods. ¹ More specifically, each water system is maintained with biofilters and ultraviolet units. The water quality is set at 26°C, pH 6.8–7.2, ammonia and nitrite < 2 ppm, nitrate < 50 ppm and at circa 300 μS/cm. Fish are housed in 9-l tanks at a maximum density of five adult fish per litre. Feeding was performed three times per day (with flake food or live Artemia larvae). All husbandry was performed according to European Legislation for the Protection of Animals used for Scientific Purposes (Directive 2010/63/EU) (General licence for fish maintenance and breeding: Az.: 35-9185.64/BH KIT IBCS-BIP Karlsruhe Institute of Technology (KIT)). All experiments were performed before the onset of the free-feeding stage and so did not require ethical board approval according to the EU Animal Protection Directive 2010/63/EU. The maintenance of the rtel1 zebrafish line was approved by the Tel-Aviv University Animal Care Committee (04-18-035 and 04-18- 051) and conducted in accordance with the National Council for Animal Experimentation, Ministry of Health, Israel. Eggs derived from 6–9 months old fish were collected within 3 h following fertilization and washed three times with 1x E3 fish medium (5 mM NaCl; 0.17 mM KCl; 0.33 mM CaCl₂; 0.33 mM MgSO₄), or directly after fertilization for injections. After removing unfertilized eggs, live eggs were then incubated at 28°C in a petri dish in the presence of 1x E3 medium containing 0.1% methylene blue (Sigma Aldrich) to prevent fungal contamination. Between 10 and 24 hpf, eggs were bleached in 0.0045% sodium hypochlorite (Roth) using a standard procedure ¹ and then incubated in E3 plus Pronase (20 µg/ml) to facilitate hatching. At this stage, embryos were cleaned and washed extensively with E3 buffer in order to minimize any potential contamination by maternally derived cell material. Embryos were then transferred into single wells of a flat-bottomed 96-well plate (Embryo-plate) in 0.2 ml of E3 embryo medium and left to develop at 28°C until the start of the genotyping protocol that was set at one, two or three days post fertilization (dpf) depending on the trial. The transfer of each embryo was performed using plastic Pasteur pipettes (Roth) with wide tips.

Zebrafish guide RNA injection for mosaic rheb knockout embryos

CRISPR single guide RNAs (sgRNAs) were designed using the ChopChop web interface (https://chopchop.cbu.uib.no/) and synthesized using cloning-free PCR with a 60 bp gene-specific oligo (TAATACGACTCACTATA GGTACATGTGGATGTTGATGGTTT TAGAGCTAGAAATAGCAAG) containing a T7 promoter, a 20 base spacer region specific to the rheb target site and one overlap region that anneals with the constant oligonucleotide (AAAAGCACCGACTGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAACTTGCTATTTCTAGCTCTAAAAC). The specific and constant oligonucleotides were annealed, filled-in with T4 DNA polymerase and transcribed into a sgRNA using the MEGAshortscript™ T7 Transcription Kit (Invitrogen) followed by ammonium acetate precipitation according to the manufacturer’s instructions. The sgRNA (300 ng/µl) was injected together with Cas9 protein (GeneArt Platinum Cas9 Nuclease, Invitrogen, 300 ng/µl) into one cell stage WT embryos.

Embryo shaking step

The genotyping procedure started with the substitution of all the E3 medium in each well with 50 μl of 30 mM Tris-HCl pH8.0 using a multichannel pipette, to prevent cross-contamination. Embryos were shaken for 30 min at 250 or 500 rev/min at 28°C using an Unithermix 2 pro, LLG Labware equipped with a Plate96 block.

After the shaking step, 40 µl of the solution was carefully removed from each well using a multichannel pipette and was transferred to the equivalent position on a separate 96-well plate (Genomic-plate). Immediately afterwards, 200 µl of E3 fish buffer was added to each embryo in the original Embryo-plate and each embryo was observed through an Olympus SZK7 stereomicroscope equipped with an Olympus EP50 camera to verify their health. Embryos were then incubated at 28°C in a Binder incubator to continue their normal development.

Genomic DNA extraction and genotyping of zebrafish embryos

Ten microlitres of 5x PK lysis buffer were added to each 40 µl sample in each well of the Genomic-plate containing the released cells. The Genomic-plate was sealed using an optical adhesive cover and then incubated at 55°C for 3 h to extract genomic DNA (gDNA) from the cells released from each embryo during the shaking step. Proteinase K was then inactivated by incubation of the plate for 10 min at 95°C. The Genomic-plate was subsequently stored at 4°C until further processing. Genotyping of the embryos was performed by PCR analysis using target gene-specific primers (Table 1).

Table 1.

List of primers used to amplify genomic sequences via polymerase chain reaction (PCR). For each pair of primers, the sequences (Fw, forward primer; Rv, reverse primer), the specific annealing temperatures, the size of the PCR fragments in bp and in which Figure the corresponding PCR reactions are shown.

Gene	Primer pairs used	PCR annealing temperature	Notes
fblx3a	Fw: 5′-TGCTTGCACCTGGTTTGTTT-3′Rv: 5′-CAAGAAGGGGCAGGTACTGAA-3′	62°C	Figure 1(a) and (b) and Figure 2(b) 239 bp
p53	Fw: 5′-GATAGCCTAGTGCGAGCACACTCTT-3′Rv: 5′-AGCTGCATGGGGGGGAT-3′	62°C	Figure 2(a) 220 bp
rtel1	Fw: 5′-CCAGAAAGTCAATGGTGTCC-3′Rv: 5′-GTGGTCTCTCCATGCCAG-3′	56°C	Figure 3(a) 72 bp mut 88 bp WT
mondoa mlxipka405	Fw: 5′-CGCATACTCCTTTATCTTGC-3′Rv: 5′- CATGCTTTTTCCATTAAAACC -3′	55°C	Figure 3(b), digestion with Bsl1 restriction enzyme 397 bp mut 297 + 104 bp WT
rheb	Fw: 5′-GAGTCTGAGACTTTCCTCGTAATTTATCAC-3′Rv: 5′-CGAGTTTCACAAAGCATCGTGCTC-3′	59°C	Figure 4320 bp fragment

mut: mutation; WT: wildtype.

Genotype assignment depends on the mutation and on genotyping strategy. If the genomic mutation significantly modifies the size of the PCR product as compared with the amplicon obtained with the WT gDNA, then the genotype can be assessed immediately after the PCR, based on the size of the amplified bands (e.g. rtle1 mutant). Another classically used strategy is the introduction or deletion of a restriction site in the mutated genomic locus. In such a case, the PCR amplification is followed by a restriction enzyme digestion step (e.g. mondoa mutant). However, when the mutation consists of the addition, deletion or exchange of very few base pairs which do not affect a restriction enzyme recognition site, then sequencing of the PCR product is required to assign the genotype (e.g. rheb mutant).

For PCR reactions, 10 µl of gDNA-containing sample was transferred to the equivalent position on a 96-well plate (PCR-plate) and processed as follows. The PCR reactions were performed in a final volume of 50 µl (final concentrations: 1x GoTaq Polymerase Green buffer; 0.4 µM of the forward and reverse primers; 0.2 mM dNTPs mix and 5 units of GoTaq Polymerase) and involved 5 min denaturation at 95°C followed by 40 cycles of 30 s denaturation at 95°C, 30 s annealing at a primer pair-specific temperature (Table 1) and 30 s elongation at 72°C. Finally, the samples were incubated at 72°C for 5 min and then at a storage temperature of 4°C. The PCR products were visualized by running 5–10 µl of each reaction on an agarose gel and then examined and documented with the E-BOX gel documentation imaging system (Vilber) in order to assign a genotype to each embryo. In the case of weak PCR amplification, the product can be reamplified using the same pair of primers or a set of nested primers.

Elimination of embryos with non-desired genotypes

Embryos with non-desired genotypes were collected and sacrificed before 120 hpf via rapid chilling and by the addition of bleach (5.75% sodium hypochlorite). ²³ Embryos were maintained in this solution for at least 20 min and then the solution was filtered using a coffee filter paper to which the embryos stick and were then disposed of immediately at –20°C in a cadaver bag until incineration.

Results

Our goal was to develop a simple method whereby we could detach and collect a limited number of cells from the body of each developing embryo that would be sufficient for use in a sensitive PCR-based assay. We reasoned that the movement generated by low frequency shaking of an embryo during a short time window might release sufficient cells from the body to yield amplifiable DNA material. We first established the method via the detection of two different genomic loci in WT embryos. Then, as a proof of principle, we used the method to genotype three different types of mutations.

Establishment of the method

We first tested the effects of low frequency agitation on WT embryos from the A/B strain. Eggs were collected and washed extensively in order to minimize any potential contamination by any maternal cells. Then eggs were aliquoted into a 96-well plate (Embryo-plate) supplemented with 0.2 ml fish medium (E3) per well and incubated at 28°C. At 3 dpf, the E3 medium in the Embryo-plate was substituted with 50 μl of 30 mM Tris-HCl pH8.0 and subjected to 30 min of shaking at 28°C at two different frequencies, 500 rev/min or 250 rev/min. The 30 mM Tris-HCl pH8.0 solution was chosen as a buffer to enhance optimal embryo survival as well as to permit optimal PCR amplification from the recovered material. ²⁴ After shaking, 40 μl was sampled from each well to perform the gDNA extraction using cell lysis buffer containing PK for 3 h at 55°C, while each well on the plate containing an embryo was immediately refilled with 0.2 ml E3 medium and returned to 28°C to allow the embryos to recover and to continue normal development. Our results show that even with slower shaking at 250 rev/min, it is possible to successfully amplify embryo-derived genomic DNA using a primer pair specific for the fblx3a gene (Figure 1 and Table 1).

Figure 1.

Low frequency shaking of embryos releases sufficient cells for polymerase chain reaction (PCR) detection of genomic DNA. Three-day post fertilization embryos were subjected to different shaking conditions ((a) 500 rev/min and (b): 250 rev/min) and a fragment of the fblx3a gene was PCR amplified from the extracted genomic DNA. A 100 bp DNA marker (M) and negative PCR controls (–) are indicated. The specific amplicons (indicated by an arrow) are clearly distinguished from a faint non-specific band of higher molecular weight.

For this method to be as widely useful as possible, it would be advantageous to be able to apply it with even earlier developmental stages, which certain countries have defined as meriting protection according to differing ethical considerations. Therefore, after the success of these pilot experiments, we next tested the efficiency of this methodology applied at different embryonic stages ranging from 1 dpf to 3 dpf (24, 48 and 72 hpf). Embryos at each of the three developmental stages were vibrated for 30 min at 250 rev/min and 28°C and then PCR was performed using primers specific for two different zebrafish genes (p53 and fblx3a). We obtained specific PCR products with both pairs of primers using the shaking-derived material that had been obtained at each developmental stage (Figure 2).

Figure 2.

Shaking embryos as early as one day post fertilization (dpf) releases enough cells for polymerase chain reaction (PCR) detection of genomic DNA. 1 dpf, 2 dpf and 3 dpf zebrafish embryos were shaken at 250 rev/min and fragments of the p53 (a) and fblx3a (b) genes were PCR amplified from extracted genomic DNA. 100 bp DNA marker (M) and negative PCR controls (–) are indicated. The specific amplicons (indicated by an arrow) are clearly distinguished from a faint non-specific band of higher molecular weight for fblx3a and a primer-derived band for p53.

With 1 dpf-old embryos we observed variability in the intensity of the amplified bands from sample to sample with respect to the 2 and 3 dpf embryos, presumably because of the low number of cells released by the shaking procedure at this early stage. However, the efficiency of amplification can be easily improved by a second amplification step using identical or nested primers.

All the embryos subjected to the vibration procedure at 1, 2 or 3 dpf were subsequently assessed until 5 dpf for normal development in the context of the 96-well plate by observation with an Olympus SZK7 stereomicroscope after the genotype assessment was completed. Normal development was observed for all three groups of embryos in terms of their overall length, degree of pigmentation, presence of a normal heart beat, size of the yolk sack, relative dimensions and form of the eyes and the head region as well as the presence of spontaneous movement.

Proof of principle applications

We next tested our genotyping methodology using 2 dpf siblings derived from heterozygote incrosses of two different mutant lines (rtel1 and mondoa (mlxip^ka405, ZFIN ID: ZDB-ALT-180628-2). The resulting clutches consist of a mixture of WT, homozygous mutant, and heterozygous embryos. For the rtel1 incrossed embryos, analysis of the PCR products on a 4% agarose gel showed two bands corresponding to either the WT (upper band, 88 bp) or the mutant (lower band, 72 bp) allele that contains a deletion of 16 bp. Thus, based on the presence of only the upper band, only the lower band or both bands, among the 32 embryos analysed, six were identified as WT, seven as homozygous mutants and 19 as heterozygotes (Figure 3(a)). For the mondoa incross embryos, the PCR products were first digested with the BslI restriction enzyme, as the mutation deletes the BslI restriction site that is present in the amplicon from the WT allele. ²² Among the 12 embryos analysed (Figure 3(b)), three were identified as WT (two bands corresponding to the digested amplicon), three as homozygous mutants (only one undigested upper band) and six as heterozygotes (three bands).

Figure 3.

Genotyping of rtel1 and mondoa mutant embryos. Two-day post fertilization (dpf) embryos derived from a heterozygous incross of the zebrafish mutant rtel1 (a) and mondoa mutant lines (b) were shaken at 250 rev/min. Polymerase chain reaction (PCR)-based genotyping was performed on the extracted genomic DNA after restriction digestion with Bsl1 for mondoa. 100 bp DNA markers (M) and negative PCR controls (–) are indicated. Black arrows indicate the PCR products derived from the wildtype and mutant alleles. The specific amplicons are clearly distinguished from the faint smaller primer-derived band.

Finally, we used this method to genotype mosaically modified embryos. Specifically, the aim was to verify the efficiency of this method to detect mutations present in cells mixed with non-modified cells. For this purpose, we performed our established genotyping protocol with 2 dpf embryos that had been injected with an sgRNA targeting the rheb locus and CAS9. PCR amplification of the genomic sequence was performed with primers binding before and after the CAS9 protospacer-adjacent-motif (PAM) targeted area (Figure 4(a)). PCR products obtained from the cells from seven embryos were sequenced and analysed to distinguish genetically modified from non-genetically modified embryos (Supplementary material Figure 4 online). The mosaic mutant cells exhibit either disrupted sequences in the case of the accumulation of multiple mutations within the same embryo (Figure 4(b), sequence examples 6101340 and 6101345), or deletions or insertions before the PAM (Figure 4(b), sequence examples 6101342, 6101344, 6101346) compared with sequences read from PCR products derived from non-mosaic WT embryos (Figure 4(b), examples sequence 6101341).

Figure 4.

Detection of CRISPR/Cas induced mutations in the rheb locus. One cell stage embryos were injected with a single guide RNA targeting the rheb gene synthesized with a Maxiscript Kit (Invitrogen) followed by ammonium acetate precipitation. Embryos (one day post fertilization) were shaken at 250 rev/min. The targeted locus was polymerase chain reaction (PCR) amplified from the extracted genomic DNA using primers spanning the targeted protospacer-adjacent-motif (PAM; red box). (a) Amplicons resolved on an agarose gel. A 100 bp DNA marker (M) and negative PCR controls (–) are indicated. (b) Sequences of seven randomly selected PCR products showing either no mutation (no mut. (WT)), deletion, insertion, or multiple mutations (multiple mut.) are presented. Original sequences are also presented in the Supplementary material with the original sequencing number.

Discussion

One of the complications of working with the zebrafish, as well as other animal models, is the unavoidable production of surplus animals for various reasons, ranging from them having an unwanted genotype, to sex or age. In particular, the establishment and maintenance of transgenic and mutant lines generates a large number of surplus animals. Given that these animals cannot be used in experiments, they accumulate in animal facilities, impacting on space as well as on cost and maintenance time. An ethical debate has now started in Europe concerning whether the extra costs and space required to maintain these surplus animals is sufficient a justification to sacrifice them. Based on consequent changes to the law on animal experimentation in the EU and Germany which now requires more detailed documentation of the fate of surplus animals, future, more widespread legislative changes in Europe seem likely to greatly limit the permitted sacrifice of surplus animals. It is therefore increasingly urgent to develop strategies to minimize the number of surplus animals being raised in animal facilities. Here, we have tackled this complex issue by establishing a method that enables genotyping of zebrafish to be performed at a much earlier developmental stage. Thus, this method allows the identification and sacrifice of surplus animals at developmental stages when it is ethically acceptable and legally permitted.

We have tackled this complex problem of reducing the number of surplus animals by developing a new protocol to genotype live zebrafish embryos before they reach the 120 h post-fertilization stage (see scheme of the protocol in Figure 5). At this stage, larvae are able to feed autonomously and are considered by EU legislation as sentient animals which are subjected to animal protection legislation.

Figure 5.

Simplified scheme of the genotype protocol. hpf: hours post fertilization; PCR: polymerase chain reaction

Current genotyping methods are mainly restricted to analysing juvenile and adult stages in zebrafish. In embryos, a micro-scale fin clip method that needs to be performed under a microscope involves PCR analysis of the DNA derived from a fin biopsy of anaesthetized embryos.^20,25 Although it is possible to perform this method without ethical permission at embryonic stages preceding 120 hpf, it is a laborious method not applicable to analysing large numbers of embryos in parallel. Another zebrafish embryo genotyping method described recently is easier to perform. ²⁴ However, this protocol involves Proteinase treatment performed directly on sedated embryos and with agitation at an elevated temperature of 37°C. Although it is possible to perform this protocol without ethical permission at embryonic stages before 120 hpf, and it does serve to reduce the scale of maintained surplus fish, it also involves non-physiological, aggressive treatments that risk affecting the future growth, phenotype or behaviour of the animals when analysed in future experiments. Interestingly, another genotyping method has been described that is also based on the use of PCR to analyse cellular material mechanically released from developing embryos. ²⁶ However, that method relies on the use of a specialized microfluidic device with a roughened glass surface that combined with harmonic oscillations induces the release of sufficient cells from sets of 24 embryos. In contrast, our shaking-based method uses standard laboratory material and equipment and can be directly scaled-up. While each genotyping method has its strengths and weaknesses, one critical parameter is the efficiency and sensitivity of the PCR amplification step, which clearly will differ depending on the gene under investigation. Amplifying a large, complex genomic DNA fragment clearly requires a larger amount of starting cellular material, so here, methods, for example, involving fin clipping may be the better choice. Alternatively, if a shorter fragment is targeted (100–400 bp) with a less complex sequence, then material from very small numbers of cells can be reliably amplified, thus broadening the scope of appropriate methods. Thus, the original design of the strategy for generating a transgenic or mutant line is a key step affecting the choice of genotyping method.

Our shaking-based method is applied in 96-well plates and so can be performed in parallel with a large number of embryos under physiological conditions. The method involves gentle shaking for 30 min at 250 rev/min, a velocity that most likely resembles water movement in the natural habitat of the zebrafish. ²⁷ It is possible to perform the assay at the 1 dpf embryo stage; however, in this case some embryos may be lost during handling and processing of the samples because of their transparency and poor visibility. Clearly, the genotyping results must be obtained before the embryos reach the 120 hpf stage. This is easily achievable by optimizing the PCR conditions. In cases where re-amplification of the PCR product is required, this will extend the time required to obtain the final genotype by a maximum of 2–3 h and so needs to be considered in the complete timetable for the analysis. In addition, it is valuable to design the strategy for identifying a transgene or a mutation where the genotype can be determined simply by analysing changes in the length of a PCR fragment, or by digestion with restriction enzymes as in the case of mutations creating or deleting a restriction site. An optimized strategy can greatly reduce the time required to obtain the genotype result.

There are cases where it is unavoidable that the PCR product needs to be sequenced in order to identify the genotype. This is typically required in procedures for establishing CRISPR/Cas mutant lines. In such cases, many injected individuals have to be analysed, classically at the juvenile stage and therefore this generates large numbers of surplus animals with undesired genotypes. The clarity of the sequence data we obtained following this experimental procedure demonstrates that it is possible to use this protocol to reliably identify single base mutations via sequence analysis at the embryonic stage. In addition, our method is particularly amenable to the analysis of large numbers of individuals and so represents an ideal replacement strategy in the establishment of CRISPR/Cas mutant lines. Furthermore, our data show that this optimized method can successfully predict the efficiency of the injection procedure and also of the chosen guide RNA, identifying the percentage of mosaic modified embryos within the F0 founder generation as soon as 2 dpf. Thus, it helps to drastically reduce the number of animals that need to be raised in order to obtain the desired mutant. Clearly, when using a commercial sequencing service, the final sequencing results must be obtained in a maximum of 24–36 h. Nevertheless, this method will undoubtedly make an important contribution to reducing the number of surplus zebrafish that need to be maintained following these commonly used experimental procedures.

Data availability statement

The data described in this manuscript are available from the corresponding author, upon reasonable request (daniela.vallone@kit.edu)

Ethics statements

All experiments were performed before the embryos’ free-feeding stage and so this study did not require ethical board approval according to the EU Animal Protection Directive 2010/63/EU. (General license for fish maintenance and breeding: Az.: 35-9185.64/BH KIT IBCS-BIP Karlsruhe Institute of Technology (KIT)). The establishment and maintenance of the zebrafish rtel1 mutant line used in this study was approved by the Tel-Aviv University Animal Care Committee (04-18-035 and 04-18- 051) and conducted in accordance with the National Council for Animal Experimentation, Ministry of Health, Israel.