Using Zebrafish to Advance Disease Research
by Mark Behlke
Introduction and Background on Zebrafish
Since the introduction of CRISPR-Cas9 gene editing in the last few years, the technology has generated substantial attention and excitement for its ability to make precise, permanent changes to DNA in animals and plants. By altering DNA sequences, CRISPR has the potential to provide novel therapies for patients suffering from diseases caused by single gene mutations, including sickle cell disease (SCD), severe combined immunodeficiency (SCID), cystic fibrosis, and Huntington’s disease.1-3 To use it effectively, researchers and clinicians must understand the many functions of disease- associated genes before they can be safely altered using genome editing technologies. Genes may have more than one function and changing a gene sequence may have unintended consequences. CRISPR can help accomplish a better understanding of gene function through experimental knock-out or knock-in of new genetic changes in model systems and examination of the resulting phenotypic effects. Understanding the genetic contributions and roles in disease is critical for the development of precision gene therapies.
An increasingly popular model organism for studying gene function and the genetic causes of human diseases is the zebrafish (Danio rerio), a freshwater fish belonging to the minnow family, native to Southeast Asia. It is a small, robust species that is cheaper to maintain than mice. Its transparency during its embryonic and larval stages is particularly useful for the direct observation of developmental and other physiological processes. They grow and reproduce quickly and so are ideal for high-throughput experiments.4 Moreover, once the complete zebrafish genome was fully sequenced, it was revealed that they have a surprisingly similar genetic structure to humans, sharing 70 percent of genes.5 Of the human genes known to be associated with disease,84 percent have a zebrafish counterpart or ortholog.6
High quality genome sequencing and annotation means that scientists have thousands of zebrafish genes with which to study genetic function and disease. To do this, researchers create zebrafish with specific genes deleted or mutated to investigate the effects of the loss of that gene and thus extrapolate its function. This phenotype may then be ‘rescued’ by inserting a normal copy of the gene back in. Similarly, genetic modifications are made to create models of disease for use in drug discovery. Zebrafish can be genetically altered to generate avatars for individual patients, which can be used to screen for personalized therapies.7
The CRISPR system evolved as a form of bacterial immune system where guide RNAs, which provide sequence specificity, combine with a CRISPR nuclease to form an active ribonucleoprotein (RNP) complex that can cleave a target DNA. For the case of Cas9, in the natural setting two annealed RNA molecules are needed to form the complete, active RNP complex.9 These two RNAs are called the CRISPR RNA (crRNA, which provides sequence specificity) and the trans-activating crRNA (tracrRNA, which binds Cas9). As an artificial simplification, the separate crRNA and tracrRNA can be fused to form a single chimeric RNA, called a single guide RNA (sgRNA).1 Both versions of RNP are active and can be used in genome editing. The sgRNA format provides added convenience for experimental protocols by allowing for expression (e.g., transcription) of the guide as a single unit when placed in an appropriate plasmid or viral vector.
To realize the immense potential of CRISPR in the discovery of new drugs and the development of precision medicine, companies like Integrated DNA Technologies (IDT) have been working to develop better tools and methods for researchers in the life sciences. The use of CRISPR-Cas9 in zebrafish has been somewhat restricted by the inconsistent and highly variable activity achieved when older versions of the technology are employed. Using these approaches, the efficiency with which target DNA sites are edited rarely approaches 100 percent and the resulting zebrafish end up being a random mix of cells with more than one genotype.
In genetics, this mixed genome phenomenon is referred to as mosaicism.8 Like humans, zebrafish have two copies of every gene present in their genome. To completely remove a gene’s function, both copies (alleles) must be edited, which requires very high activity of the CRISPR machinery. To improve the efficiency of CRISPR-Cas9 for gene editing in zebrafish, Professor David Grunwald and his team at the University of Utah (USA) employed the latest IDT-developed CRISPR tools to create a new, highly efficient CRISPR-Cas9 method for the generation of zebrafish with deletion mutations and lost gene functions with close to 100 percent efficiency and minimal mosaicism (see Figure 1).8
Development of a highly efficient CRISPR method
From a biochemical standpoint, CRISPR- Cas9 gene editing can be performed with RNP complexes using sgRNAs that are either chemically synthesized or made enzymatically using in vitro transcription (IVT); as an alternative, crRNA:tracrRNA complexes (annealed crRNA and tracrRNA) that are chemically synthesized can be used (this latter approach is referred to as the duplex guide RNP, or dgRNP, method [see Figure 2]).10 The improved CRISPR-Cas9 method developed by Professor Grunwald and IDT demonstrated that the dgRNPs, as well as chemically- synthesized sgRNA-containing RNPs, performed with greater potency and consistency than RNPs made with IVT-produced sgRNAs in zebrafish mutagenesis. The major difference between IVT-produced sgRNA and chemically synthesized sgRNA or crRNA:tracrRNA is that IVT-produced sgRNAs contain supernumerary guanines (one or two extra “G” residue[s] are positioned at the beginning of the sequence), which are needed to make the enzymatic IVT process more efficient. However, this results in a final sgRNA that has an extra base or two at one end which often does not match the target DNA sequence.
During the development of the new method, the effect of supernumerary guanine nucleotides at the 5’ end of sgRNAs was found to reduce the efficiency of the Cas9 RNPs in cuttingnthe zebrafish genomic DNA target. Even the presence of a single unpaired supernumerary guanine at the 5’ end of an sgRNA significantly diminished the RNP activity. Chemically synthesized sgRNAs or dgRNAs, on the other hand, are made to be a perfect match to the target DNA and do not have extra residues. By eliminating the supernumerary guanines and ensuring that the 5’ end of the guide RNA is an exact match to the genomic DNA target sequence, the gene editing efficiency of the RNP was greatly improved. Chemically synthesized sgRNA and crRNA:tracrRNA both worked remarkably well when used in RNPs.
The team chose to use dgRNPs for further experiments. With this extremely efficient dgRNP method, virtually all the copies of the targeted gene in the zebrafish embryos were mutated, consistently containing the desired bi-allelic insertion–deletion mutations (indels) introduced by the CRISPR-Cas9. The resulting disruption of gene function reached very high efficiency at a level of mosaicism in the F0 embryos (the generation emerging from eggs injected with CRISPR-Cas9 RNPs) sufficiently low that phenotypes could be investigated directly without the need to breed daughter generations, greatly speeding up the discovery process. The deletion mutations induced using this method often resulted in a complete loss of function, with the embryonic F0 generations closely resembling null mutants to a remarkable degree. Moreover, the consistent Cas9 activity and mutagenesis did not give rise to any confounding non-specific traits resulting from off-target editing effects.9
This highly efficient CRISPR-Cas9-based method for generating deletion mutations and F0 embryos that lack gene function in zebrafish enables researchers to investigate the functions of individual genes. Large numbers of gene functions can thus be quickly screened in F0 embryos in a single experiment. Combinations of genes can also be examined, as this CRISPR-Cas9 method can be used to target two genes simultaneously, through the application of two sets of duplex-RNA-guided RNPs. By inducing mutations in two distinct genes, connections between their functions can be observed. At the same time, any redundancy between these genes and their roles can also be counteracted, producing a more complete loss of function. The simultaneous targeting of two sites was also achieved with two coding regions on the same chromosome and is especially important in zebrafish because single point mutations are often phenotypically overcome by redundant genes. Therefore, the targeting of two genomic sites often gave much better results than targeting only one site.
Moreover, the new method also demonstrated that genomic DNA more than 50 kilobases in length could be consistently deleted. Being able to make such large deletions can be particularly useful for the preliminary screening of gene functions, as the effects of large deletions are likely to be more noticeable and impactful than those resulting from small deletions in zebrafish. In addition, the indels can be induced in the F0 embryos that are heritable; the ability to make such mutants is especially useful for the characterization of null phenotype and to facilitate experiments where the generated mutations can be recovered in subsequent zebrafish generations.9 Areas of potential application The optimization of the consistency and efficiency of CRISPR-Cas9 for genome editing in zebrafish has numerous potential applications.9 Indeed, zebrafish are being employed for a whole host of uses for which better CRISPR-Cas9 protocols would be of great value.8
When it comes to drug discovery, especially for first-in-class compounds, phenotypic screens in cells are the tool currently favored by most pharmaceutical and biotechnology companies. However, these cell-based models are unable to capture the effects of small molecules and biotherapeutics on many complex biological processes. To examine whole-organism physiology, animal models such as mice are preferred. However, they take much longer to grow and mature compared with zebrafish. Consequently, zebrafish are gaining more attention as a model species for high-throughput and large-scale screening in drug discovery research.4,7
In the time since zebrafish were first used intensively approximately 20 years ago, chemical phenotypic screens have identified lead compounds for clinical development and several such investigational medicines have progressed, or soon will progress, to clinical trial stage.8 These include the identification of new indications for existing medicines, such as the rheumatism treatment, leflunomide, which was isolated using a zebrafish-based screen showing that the drug had potential for treating melanoma (skin cancer).11 Another therapy that started as a screen in zebrafish was the discovery that treatment with the small lipid prostaglandin dmPGE2 can be used to expand hematopoietic stem cells, which is now used to improve transplantation from umbilical cord blood donors.12
Zebrafish are being used in the selection of personalized medicines. Personalized medicine avatars of patients are being created using zebrafish patient-derived xenografts (zPDXs). These zebrafish avatars are used to identify the best treatment plans for particular patients. For example, two patients with a lymphatic anomaly possessed a mutation, which was then engineered into zebrafish, as well as human cells in vitro. Screening of candidate treatments using these two model systems indicated that inhibitors of mitogen-activated protein kinase (MEK) enzymes may be effective in treating the lymphatic anomaly. Indeed, in one patient, treatment with the MEK inhibitor, trametinib, led to dramatic clinical improvement.7,19
Zebrafish avatars can also be useful for determining treatment dose and duration, validating potential disease-causing mutations, and for searching for therapies that might reverse disease once the phenotype is characterized. Researchers at the University of Edinburgh (UK) and University College London (UCL) Great Ormond Street Hospital (UK) have modelled the effects of human sporadic vascular malformations using zebrafish avatars. Responses to targeted therapies identified using these avatars are used to guide personalized treatment decisions, including the compassionate use of certain treatments, such as repurposed cancer therapies, in individual cases.7,20,21
The experimental and clinical validity and utility of zebrafish as a model organism has been clearly demonstrated and is rapidly being adopted among academic researchers working in the life sciences and medical clinics. Although broad application in the biopharmaceutical industry has yet to occur, a growing number of lead compounds are being identified and characterized using zebrafish models and screens, which are advancing into clinical drug development. As such, the timely addition of new, improved CRISPR- Cas9 methods optimized for highly efficient and consistent genome editing in zebrafish is very welcome. Methods such as these will be powerful tools in biomedical research and the development of personalized and stratified medicines to combat human disease. Such tools could also then accelerate the translation of precision medicines to the clinic for numerous diseases, especially those with a clear genetic etiological component.
Dr Mark Behlke is Chief Scientific Officer at Integrated DNA Technologies (IDT) and has directed many activities since joining the company in 1995, with a focus on novel molecular biology applications of oligonucleotide-based technologies. Before joining IDT, Dr Behlke was a Physician Postdoctoral Fellow of the Howard Hughes Medical Institute at the Whitehead Institute, MIT. He was a Resident Physician in Internal Medicine at Brigham and Women’s Hospital, Boston. Dr Behlke is an inventor on over 50 issued US patents, has numerous pending patent applications and is an author of over 130 scientific publications and book chapters.
1. Dever DP, Porteus MH. Th e changing landscape of gene editing in hematopoietic stem cells: a step towards Cas9
clinical translation. Curr Opin Hematol. 2017; 24: 481–488.
2. Cai L, Fisher AL, Huang H, Xie Z. CRISPR-mediated genome editing and human diseases. Genes Dis. 2016; 3:
3. Bashir H. Emerging therapies in Huntington’s disease. Expert Rev Neurother. 2019; 19: 983–995.
4. Burke E. Why use zebrafi sh to study human diseases? NIH Intramural Research Program. I am Intramural Blog.
Tuesday, August 9, 2016. https://irp.nih.gov/blog/author/elizabeth-burke (accessed January 2020).
5. Howe K, Clark MD, Torroja CF, et al. Th e zebrafi sh reference genome sequence and its relationship to the human
genome. Nature 2013; 496: 498–503.
6. Wellcome. Zebrafi sh genome yields signifi cant similarity to human genome. News. April 19, 2013. (accessed
7. Cully M. Zebrafi sh earn their drug discovery stripes. Nat Rev Drug Discov. 2019; 18: 811–813.
8. Hoshijima K, Jurynec MJ, Klatt Shaw D, et al. Highly effi cient CRISPR-Cas9-based methods for generating deletion
mutations and F0 embryos that lack gene function in zebrafi sh. Dev Cell. 2019; 51: 645–657.e4.
9. Jinek M, Chylinski K, Fonfara I, et al. A programmable dual RNA-guided DNA endonuclease in adaptive bacterial
immunity. Science. 2012; 337: 816–821.
10. Jacobi AM, Rettig GR, Turk R, et al. Simplifi ed CRISPR tools for effi cient genome editing and streamlined
protocols for their delivery into mammalian cells and mouse zygotes. Methods. 2017; 121–122: 16–28.
11. White RM, Cech J, Ratanasirintrawoot S, et al.
12. Hagedorn EJ, Durand EM et al. Getting more for your marrow: boosting hematopoietic stem cell numbers with
PGE2. Exp Cell Res. 2014 Dec 10;329(2):220-6.
13. NIH. ClinicalTrials.gov. Th e use of trifl uoperazine in transfusion dependent DBA (DBA). NCT03966053. https://clinicaltrials.gov/ct2/show/record/NCT03966053 (accessed January 2020).
14. NIH. Genetics Home Reference. Sickle cell disease. https://ghr.nlm.nih.gov/condition/sickle-cell-disease#diagnosis
(accessed January 2020).
15. Sickle Cell Disease Foundation. Who is aff ected? http://www.scdfc.org/who-is-affected.html (accessed January
16. Khemani K, Katoch D, Krishnamurti L. Curative therapies for sickle cell disease. Ochsner J. 2019; 19: 131–137.
17. NIH. ClinicalTrials.gov. Stem cell gene therapy for sickle cell disease. NCT02247843. https://clinicaltrials.gov/ct2/show/NCT02247843?recrs=abdfm&rslt=Without&type=Intr&cond=SICKLE+CELL&age=1&phase=014&draw=2
(accessed January 2020).
18. Collins F. A CRISPR approach to treating sickle cell. NIH Director’s Blog. April 2, 2019. https://directorsblog.nih.gov/2019/04/02/a-crispr-approach-to-treating-sickle-cell/ (accessed January 2020).
19. Vakulskas C. How might gene editing be used to cure disease? Journal of Precision Medicine. 2019; 5. https://www.thejournalofprecisionmedicine.com/wp-content/uploads/2019/12/jpm419-Vakulskas.pdf (accessed January 2020).
20. Li D, March ME, Gutierrez-Uzquiza A, et al. ARAF recurrent mutation causes central conducting lymphatic
anomaly treatable with a MEK inhibitor. Nat Med. 2019; 25: 1116–1122.
21. Al-Olabi L, Polubothu S, Dowsett K, et al. Mosaic RAS/MAPK variants cause sporadic vascular malformations
which respond to targeted therapy. J Clin Invest. 2018; 128: 5185.