Zebrafish swimming towards cures: a scalable NAM platform for drug discovery

Abstract

Drug discovery faces escalating costs, extended timelines, and ~90% failure rates, prompting National Institutes of Health (NIH) and US Food and Drug Adminstration (FDA) guidance to reduce animal testing. As a New Approach Methodology (NAM), zebrafish (Danio rerio) bridges high-throughput in vitro assays and mammalian models. Its genetic tractability, optical transparency, rapid development, high fecundity, and genomic similarity to humans enable cost-effective whole-organism screening across drug development. In this review, we survey zebrafish applications from target identification to regulatory evaluation, and how advances in automation, imaging, artificial intelligence (AI), and genetic engineering are extending its reach. We highlight how zebrafish offer a complementary platform accelerating therapeutic discovery and precision medicine.

Teaser:

Amidst calls to move away from standard animal models, zebrafish provide a cost-effective, ethically sound alternative, enabling rapid, whole-organism drug screening with significant implications for therapeutic development.

Introduction

The druggable genome, proteins targetable by small molecules, comprises an estimated 3000–4500 genes, yet only 5–10% are addressed by FDA-approved drugs, leaving vast unexplored target space.¹ Drug development costs ~US$2.6 billion and 10–15 years per approved drug. ² Despite advances in high-throughput screening, computational modeling, and o’mics, ~90% of candidates fail in late-stage trials as a result of inadequate efficacy or unforeseen toxicity.^3–5 Although scalable, in vitro studies are reductionist and often miss systemic interactions.^6–8 These challenges underscore the need for NAMs that combine physiological relevance, scalability, and cost-effectiveness.

The zebrafish, Danio rerio is a powerful vertebrate model in drug discovery, bridging the gap between in vitro assays and mammalian in vivo studies. Atlhough commonly used in developmental biology since the 1960s, zebrafish have since established a firm foothold in biomedical research and preclinical drug discovery pipelines because of their conserved organ systems, rapid development, and high fecundity. Importantly, their larvae enable high-throughput, whole-organism chemical screening at a scale and cost unattainable in mammalian systems.^9–11

Zebrafish show high genetic and pharmacological concordance with humans: ~70% of human genes have at least one zebrafish ortholog, and nearly 84% of human disease genes are conserved.¹² Their small size and permeability to small molecules enable non-invasive compound administration and real-time longitudinal assessment of developmental, physiological, and behavioral phenotypes.^13–15 These features suit zebrafish to high-content in vivo phenotypic screening, identifying compounds and toxicity concurrently. Beyond screening, zebrafish offer genetic tools for target validation and precision medicine, including CRISPR/Cas9 knockout and knock-in models, tissue-specific transgenics, and morpholino knockdown (Figure 1).^16–20 Real-world successes include clemizole and lorcaserin, identified in zebrafish Dravet syndrome screens, which have shown clinical relevance (Table 1).²¹ As automation, genetic engineering, and AI-driven analytics evolve, zebrafish are poised to occupy a central role in drug discovery pipelines.

Figure 1.

Technologies for target selection, validation, and disease model development in zebrafish. This figure illustrates key genetic manipulation techniques utilized in zebrafish for interrogating biological pathways relevant to drug discovery, including target selection, validation, and disease model development. (a) Morpholinos are synthetic oligonucleotides that bind to specific mRNA sequences, blocking protein synthesis and enabling transient gene knockdown for functional analysis. (b) CRISPR-Cas9 enables precise gene editing through either non-homologous end joining (NHEJ) or homology-directed repair (HDR), facilitating the creation of gene knockouts or specific mutations to investigate gene function. (c) RNA overexpression involves the injection of mRNA into embryos to induce overexpression of target proteins, aiding the analysis of gene function and interactions. (d) Transgenesis uses transposase-mediated integration of exogenous DNA, such as the Tol2 plasmid, to create stable transgenic lines for studying gene expression patterns and disease models. These technologies provide powerful tools for studying gene function and disease mechanisms in zebrafish, enabling advances in target validation and drug discovery.

Table 1.

Examples of zebrafish-derived drug candidates in clinical trials

Therapeutic area	Compound	Indication	Clinical trial status
Oncology	ProHema (PGE2 derivative)	Leukemia	Phase II (NCT01627314)
	All-trans retinoic acid	Adenoid cystic carcinoma	Phase II (NCT00890500, NCT03999684)
Neurology	Clemizol (EPX-100)	Dravet syndrome	Phase II (NCT04462770)
Cardiology	Vemurafenib, trametinib	Arteriovenous malformations	Compassionate use, planned trials

In this review, we explore how zebrafish are being used increasingly across the drug discovery pipeline: from finding promising targets to preclinical testing and even regulatory issues (Figure 2). We highlight where zebrafish shine and where their limitations still need to be addressed.

Figure 2.

Zebrafish in the drug discovery and development pipeline. This figure outlines the role of zebrafish in various stages of the drug discovery and development process, spanning from basic research to regulatory approval. Zebrafish models are used throughout the pipeline for understanding disease biology, selecting and validating drug targets, and developing disease-relevant models. In basic research (−1 to 0 years), zebrafish are used to map molecular pathways, identify biomarkers, and refine disease models. During drug discovery (0–3 years), zebrafish aid in target selection, validation, high-throughput screening (HTS), hit-to-lead optimization, and early bioavailability and blood–brain barrier (BBB) studies. In preclinical development (3–6 years), zebrafish are utilized for in vitro safety, early pharmacology, and proof-of-concept studies, as well as pharmacokinetics (PK) and toxicology studies. During drug development (6–10 years), zebrafish models contribute to advanced efficacy, dose-finding studies, formulation testing, and safety pharmacology panels. Overall, zebrafish are increasingly recognized for their utility in expediting drug development processes, enabling early-stage testing, and advancing regulatory approval. Abbreviations: ADME, absorption, distribution, metabolism, excretion; IND, investigational new drug.

Target identification and validation in zebrafish models

Given that many human disease pathways are conserved in zebrafish, genetic and chemical screens can identify disease-causing genes and drug targets in a whole-organism context. ^12,17,22 Forward genetic screens and morpholino knockdowns have enabled discovery of genes that modulate disease phenotypes.^23,24 Nasevicius and Ekker first established morpholino knockdown as a rapid method for gene inactivation in early embryos, generating ‘phenocopies’ of known mutants.²⁴ Today, these approaches are combined with transgenesis and targeted gene editing (ZFNs, TALENs, and CRISPR/Cas9) to create precise disease models.²⁵ CRISPR/Cas9 in particular has revolutionized zebrafish genetics: Jao et al. achieved 75–99% mutagenesis at multiple loci with high biallelic disruption and efficient germline transmission, establishing zebrafish as a scalable knockout method.²⁵ F₀ ‘crispants’ and stable mutant lines are now routinely used to test disease gene function.^22,23,26

Applications across disease areas

Zebrafish models have been applied to a range of disorders. In oncology, researchers model human cancers in zebrafish and perform in vivo chemical and genetic screens to find novel targets and drugs. For example, Leonard Zon's group used zebrafish to identify modulators of hematopoietic stem cells; a small-molecule screen revealed prostaglandin E2 (PGE2) to enhance hematopoietic stem cell (HSC) formation with subsequent studies leading to clinical trials of PGE2 analogs to improve human bone marrow transplants.²⁷ Similarly, Ablain et al. conducted targeted sequencing of human mucosal melanomas and then, in zebrafish, expressed candidate oncogenes or knocked out tumor suppressors using transgenics and CRISPR in mosaic F₀ fish, confirming SPRED1 as a tumor suppressor in KIT-driven melanoma.²⁸ Drug testing in adult transgenic fish showed that the KIT inhibitor dasatinib killed KIT-driven melanomas, unless SPRED1 was inactivated; in that case, only the MEK inhibitor trametinib was effective.¹⁷ This demonstrated how CRISPR-based disruption of a human candidate gene can alter tumor behavior and guide therapy selection in vivo.

Zebrafish are also valuable models for cardiovascular and metabolic disease because their heart develops across 72 h, is easily imaged and shares many disease-related pathways and mutations with humans, including those affecting sarcomere function, ion channels, and angiogenesis. Zebrafish are used to study drug-induced cardiotoxicity because their conserved cardiac electrophysiology mirrors human long-QT and arrhythmia responses.²⁹ In the context of metabolic disorders, such as obesity, type 2 diabetes mellitus (T2DM), and hyperlipidemia, zebrafish models are generated through diverse approaches, including diet-induced obesity (DIO) via high-fat or high-cholesterol diets, genetic manipulations of target genes, such as pdx1, plxnd1, or lxra, and chemical inductions with agents including glucose or streptozotocin. These models recapitulate key human pathologies, including adiposity, hepatic steatosis, impaired glucose tolerance, insulin resistance, and elevated lipid profiles.³⁰ In the pdx1 mutant model of maturity-onset diabetes of the young (MODY), the dosing of metformin, which is widely prescribed for diabetes, restored β cell mass, normalized blood glucose levels, and improved survival rates. ³¹ These phenotypes mirror what is seen with metformin in patients with T2DM, validating the translational relevance of the model.

In neurological and behavioral disorders, zebrafish models leverage the transparency of the zebrafish larva to directly observe the neural circuitry and behavior in real time. Neurons, glia, and neurotransmitter systems are highly conserved in fish, and zebrafish larvae exhibit measurable behaviors. One notable example is epilepsy: a zebrafish mutant lacking the voltage-gated sodium channel gene scn1Lab models Dravet syndrome (a severe human pediatric epilepsy caused by mutations in human SCN1A). Baraban et al. showed that scn1Lab mutant larvae exhibit spontaneous electrographic seizures and convulsions.^21,32–34 A focused screen of ~320 drugs identified clemizole (an FDA-approved antihistamine) as a potent inhibitor of seizures in these fish.²¹ This result not only validated zebrafish for epilepsy drug discovery, but also suggested repurposing an existing drug for Dravet syndrome. Other neurodevelopmental and neurodegenerative models are under active development: for example, zebrafish expressing human α-synuclein model aspects of Parkinson’s disease, and knockdowns of disease-risk genes (e.g., smn1 in spinal muscular atrophy or shank3b in autism) are used in screens for suppressors.³⁵

Screening hit identification and optimization

Zеbrafish embryos readily absorb chemicals from water and can be arrayed in microplates for high-content screening^36–38 (Figure 3). Whole-organism screening, optical transparency, and high homology to humans enable rapid phenotypic assessment of compound libraries and early derisking of candidates. The in vivo milieu adds biologically relevant filters: it assesses target engagement in a whole organism; compound activity correlates with blood–brain barrier permeability; and the dependence of development on tightly orchestrated signaling means that nonspecific or broadly inhibitory compounds rapidly induce lethality, providing an intrinsic selectivity filter. Thus, zebrafish pipelines have been used to identify bioactive small molecules, elucidate structure–activity relationships (SARs), and support drug repurposing.

Figure 3.

High-content phenotypic screening in zebrafish for drug discovery. Key steps in a high-content phenotypic screening process using zebrafish to identify and validate potential drug candidates comprise: (1) spawning and arraying fish and compounds: zebrafish embryos are exposed to a variety of compounds in a multi-well plate format, enabling large-scale screening; (2) hit identification: after exposure to compounds, phenotypic outcomes are captured using high-content imaging and analyzed via software, identifying positive hits (compounds that induce a desired phenotype) versus negative hits (compounds with no effect); (3) dose response validation and secondary assays: positive hits are further validated by dose–response assays to determine their potency and efficacy across various concentrations. Secondary assays can also be used to confirm the observed phenotypes and validate the activity of the compounds; and (4) target deconvolution: the final step involves determining the molecular target of the identified hit compound by investigating its interaction with specific proteins, helping to elucidate its mechanism of action. This high-throughput zebrafish phenotypic screening approach enables efficient and rapid drug discovery, particularly for novel disease-modifying compounds.

Diverse phenotypic screening modalities

The versatility of zebrafish supports an array of screening modalities that leverage their genetic tractability, compatibility with high-throughput imaging, and automated analysis. Here, we categorize these modalities into morphological screens (including those assisted by transgenes), behavioral screens, and gene expression reporter screens, each offering unique modes of screening for hit identification and optimization.

Morphological screens (including transgene-assisted)

Morphological screens in zebrafish focus on visible structural changes in embryos or larvae, providing a direct readout of developmental perturbations induced by compounds.³⁹ These assays are particularly effective for identifying modulators of developmental signalling, organogenesis, tissue remodeling, and toxicity, because alterations in shape, size, or patterning can be rapidly scored using brightfield microscopy. Hong and colleagues pioneered the use of zebrafish for chemical screening with the identification of dorsomorphin, a BMP signaling inhibitor, using a simple dorsalization of the primary axis as a phenotype readout in embryos.⁴⁰ Transgene-assisted morphological screens elevate this modality by incorporating fluorescent reporters to highlight specific cellular or tissue processes, enabling high-content analysis with greater sensitivity. Transgenic lines, such as fli1:EGFP, which labels endothelial cells, have been instrumental in quantifying vascular remodeling in response to anti-angiogenic compounds, contributing to the preclinical validation of VEGF inhibitors for cancer therapeutics. ^41,42 Similarly, transgenic models for photoreceptor morphology in retinal degeneration, where compounds restoring cone or rod structure are prioritized, and haploid transgenic screens for vascular mutants that uncover essential regulators of blood vessel integrity. Phenotype-driven screens using cldnb:EGFP, which label migratory posterior lateral line primordium (PLLp) cells have also identified collective cell migration inhibitors, demonstrating the ability of the platform to uncover compounds for metastasis suppression in cancer models.⁴³ These approaches provide easy readouts for structural changes affected by altered developmental signaling for hit identification.

Behavioral screens

Behavioral screens exploit the rich repertoire of quantifiable movements and responses in larval zebrafish, offering insights into neurological and psychiatric drug effects. These assays utilize high-throughput locomotor tracking systems with infrared cameras and software for trajectory analysis to profile compound-induced changes in activity patterns, making them ideal for neuroactive library screening.⁴⁴ Simple behavioral paradigms include the photomotor response (PMR), where embryonic light-induced movements serve as a proxy for neurotoxicity or excitability, and visual motor response (VMR) assays, which have been used to identify beta-blockers, such as carvedilol for ameliorating retinal degeneration phenotypes in retinitis pigmentosa models.⁴⁵ Thigmotaxis, a wall-hugging behavior correlated with anxiety, has been used to screen anxiolytics, while startle response assays probe sensory-motor integration.^46,47 A landmark study screened over 5000 compounds using rest–wake cycling as a readout, identifying novel hypnotics and modulators of arousal states via conserved pathways like melatonin signaling.⁴⁴

Recent breakthroughs in deep learning and machine vision have enabled sophisticated behavioral fingerprinting: deep autoencoder-based pattern recognition was shown to identify neuroactive and developmental neurotoxicants with higher sensitivity compared with traditional statistical methods, classifying subtle abnormalities in larvae exposed to environmental chemicals and pharmaceuticals.⁴⁸ This approach allows for both anomaly detection in large toxicology screens and prediction of chemical actions on neural circuits. Social behavior assays, talhough less commonly high throughput, have assessed social approach and maintenance in zebrafish, with alterations used to model neurodevelopmental disorders such as autism spectrum conditions.⁴⁹ Pharmacological modulation of these behaviors is shedding light on conserved social circuits.⁴⁹

Gene expression reporter screens

Gene expression screens use transgenic lines with fluorescent or luminescent reporters driven by pathway-specific response elements, enabling real-time, non-invasive monitoring of molecular responses to compounds. This modality is especially suited for identifying modulators of intracellular signaling cascades, where overall expression levels (e.g., upregulation or downregulation) serve as the primary readout, facilitating high-throughput quantification via plate readers or imaging.⁵⁰

An example of a reporter assay is the use of the FGF signaling reporter dusp6, which was used to generate the Tg(dusp6:d2EGFP)) line, which responds to fibroblast growth factor (FGF) pathway activation through destabilized GFP expression. In a high-content chemical screen, this transgenic system identified novel FGF activators and inhibitors, including (E)-2-benzylidene-3-(cyclohexylamino)-3,4-dihydro-2H-indazol-5-ol (BCI), which attenuates Dusp6 to expand cardiac cell lineages and enhance heart regeneration.⁵¹ Similarly, signaling pathway reporter (SPR) lines for Wnt, Notch, and Hedgehog have been applied to screen for pathway-specific modulators, integrating into broader systems biology approaches.⁵⁰

Another key illustration is the Glucocorticoid Responsive In vivo Zebrafish Luciferase activitY (GRIZLY) assay, a transgenic luciferase reporter system for glucocorticoid signaling.⁵² This screen detects in vivo activity with high sensitivity and specificity, capturing effects from metabolically activated compounds or those altering endogenous hormone production. In chemical library screens, it has identified novel agonists and antagonists, aiding the search for anti-inflammatory drugs with reduced side effects. The assay provides a simple, time-resolved, quantitative readout via luciferase activity in zebrafish larvae, operable as a stand-alone platform or integrated into HTS workflows, with detailed protocols emphasizing data normalization, quality control, and visualization.⁵²

Therapeutic screens (also called rescue or suppressor screens)

Therapeutic screening in zebrafish centers on identifying compounds that rescue disease phenotypes in established genetic or chemically induced models.^9–11,53 These assays capitalize on the conserved physiology and organ architecture of zebrafish to enable simultaneous evaluation of efficacy, toxicity, and mechanism of action within a single, whole-organism context. Conceptually, these screens parallel traditional phenotype-based assays, but their focus is inverted, seeking agents that restore disrupted biological function rather than perturb normal physiology. A defining example is the gridlock suppressor screen by Peterson et al., in which mutation of hey2 (gridlock) ablates dorsal aortic circulation; screening of ~5000 compounds revealed PI3K inhibitors capable of rescuing blood flow in mutant embryos.^54,55 In genetic disease models, such as those modeling muscular dystrophy, zebrafish offer sensitive morphological readouts that can be leveraged for quantitative screening. One widely used approach exploits the birefringent property of organized muscle fibers, which appear bright under polarized light against a dark background. This visual assay enables rapid identification of compounds that restore muscle integrity in dystrophic embryos, with standardized parameters, ensuring reproducibility and comparability across studies.^56,57 Through therapeutic screening approaches, zebrafish studies have yielded lead compounds for diverse indications, including cardiovascular, neurological, and metabolic disorders, which continue to inform preclinical efficacy and translational research.

SAR studies

A more unique use for zebrafish in drug discovery is its incorporation as an early in vivo model driving in vivo SAR studies, enabling the refinement of drug candidates by assessing potency, selectivity, and whole-organism bioactivity. Although dorsomorphin was discovered in a phenotypic screen in zebrafish, embryos had two distinct phenotypes: the first was axis shortening caused by dorsalization via inhibition of ALK2; the second was the loss of intersomitic vessels resulting from inhibition of another receptor tyrosine kinase (RTK), VEGFR2. Iterative synthesis of analogs and testing in vivo identified key structural requirements for receptor specificity and phenotype segregation, resulting in more selective ALK2 inhibitors, such as DMH1, and more selective VEGF inhibitors, including DMH4. Zebrafish models have also been pivotal in the identification of prostaglandin E2 analogs, such as dmPGE2, which enhances hematopoietic stem cell formation and has progressed to clinical evaluation.^27,58

Crucially, in vivo SAR avoids artefacts inherent to typical in vitro assays. Cell-based and cell-free systems operate in nonphysiological conditions, heterologous overexpression, artificial ionic composition and temperature, detergent-solubilized receptors, or simplified membranes, and lack accessory proteins and metabolic competence.^59,60 They often require supraphysiological agonist levels to achieve adequate signal-to-noise. ^61,62 These deviations alter receptor conformations, shift signaling amplification, bias pathway selection, and mask or exaggerate competitive versus allosteric mechanisms, producing potency, efficacy, and selectivity shifts that poorly predict in vivo pharmacology.^64,66 Measuring effects in an intact organism with native receptor expression, tissue context, metabolism, and compensatory signaling, zebrafish in vivo SAR prioritizes molecules that are more likely to translate to mammalian systems.^59,68

Drug repurposing

Zebrafish screens are particularly well suited to drug repurposing efforts because of their compatibility with human-approved libraries. Phenotype-driven identification of unexpected activities in known drugs enables rapid repositioning for new indications. A notable example is the discovery of cemizole, which stemmed from a screen using a transgenic zebrafish line containing a Nav1.1 (scn1Lab) mutation, which accurately models the genetic basis of Dravet syndrome in humans.^21,64,65 From a library of 320 compounds, clemizole was specifically identified for its ability to inhibit both convulsive behavior and electrographic seizures in this mutant zebrafish model. This promising finding led to clemizole advancing to Phase II clinical trials for Dravet syndrome (ClinicalTrials.gov identifier: NCT04462770).

Furthermore, leflunomide, an FDA-approved DHODH inhibitor for the treatment of rheumatoid arthritis, was repositioned for oncology based on zebrafish screening. Specifically, NSC210627 was identified through chemical genetic screens designed to find compounds impacting neural crest cell development, self-renewal, and lymphatic activity in zebrafish embryos. These developmental processes are often relevant to cancer pathogenesis.⁶⁶ The screen identified structurally distinct DHODH inhibitors implicating leflunomide in pathways crucial for tumorigenesis. Leflunomide is now in multiple Phase I/II and cancer prevention trials across previously treated metastatic triple-negative breast cancer, PTEN-null advanced solid malignancies, relapsed/refractory multiple myeloma, high-risk smoldering myeloma, and MEN1 syndromes.

Preclinical efficacy studies

Although mammalian models remain the preclinical gold standard, zebrafish provide a versatile vertebrate platform for whole-animal drug evaluation.^11,12,17 Their conserved drug targets, small size, and rapid development enable efficient pharmacological testing across diverse disease models. More than two decades of studies show that compounds identified in zebrafish often retain activity in mouse and human systems.^17,39,40,66 Thus, zebrafish bridge in vitro assays and mammalian studies by enabling early in vivo assessment of small-molecule efficacy, mechanism, and toxicity.

Oncology and personalized medicine

Zebrafish cancer models include both genetic tumors and transplanted human cells, and are increasingly used to evaluate anti-cancer drug efficacy.⁶⁷ Patient-derived xenografts (PDX) in zebrafish, sometimes called ‘zAvatars’, are showing great promise for personalized oncology. In this approach, tumor cells or biopsies from a patient are implanted into zebrafish larvae, which are then treated with candidate therapies. The transparent bodies allow direct imaging of tumor regression and dissemination in response to drugs. Zebrafish PDX models require only ~100–200 patient cells per fish and yield results in under a week, orders of magnitude faster and less tissue-intensive compared with mouse PDX (Figure 4). Mouse PDX typically requires 10⁵–10⁷ viable cells implanted in immunodeficient hosts and 6–24 weeks of engraftment before tumors are evaluable, with expansion and treatment studies extending total timelines to 4–9 months.^68–70 This speed and minimal sample requirement make zebrafish particularly attractive for chemosensitivity profiling. A recent clinical study of metastatic colorectal cancer compared 55 patient outcomes to responses of their matched zAvatars. Strikingly, the zAvatar predictions of chemotherapy sensitivity agreed with the patients’ clinical responses in 91% of cases, and patients whose zebrafish tests were sensitive had significantly longer progression-free survival.⁷¹ Likewise, zAvatars have been used to forecast responses to radiotherapy and immunotherapy in other cancers. In a personalized bladder cancer study, zebrafish were engrafted with a patient’s tumor and treated with BCG immunotherapy; the zebrafish model correctly identified which patients would benefit, because regression in the fish tumors corresponded to clinical remission in the donor.⁷² These successes highlight how zebrafish platforms can translate genomic and pathological tumor data into functional drug sensitivity assays, guiding precision medicine.^71,73

Figure 4.

Generation of zebrafish patient-derived xenografts (zAvatars). Tumor cells obtained from a patient biopsy are implanted into zebrafish embryos to generate zAvatars. These models reproduce aspects of the patient’s tumor biology and provide a rapid, transparent in vivo system for assessing drug efficacy and resistance in a personalized context.

Neurological disorders

Zebrafish also serve as powerful models of neurological and neurodegenerative disease, supporting screens for neuroactive small molecules. The zebrafish nervous system contains homologs of major neurotransmitter circuits (dopaminergic, serotonergic, etc.), with many receptor sequences nearly identical to those in humans.^74,75 High-throughput behavioral assays, using specialized equipment such as the Daniovision, can quantify larval locomotion, response to stimuli, or seizure activity, providing phenotypic readouts relevant to disease states. In epilepsy research, for example, a zebrafish model of Dravet syndrome (with loss of the scn1lab sodium channel) exhibits spontaneous high velocity ‘swim seizure’ behaviors. Large-scale screens in scn1lab mutant larvae successfully reidentified standard DS therapies (valproate, stiripentol, and fenfluramine), validating the predictive power of the model.⁷⁶ Recently, more than 3000 compounds were screened in this model, leading to a shortlist of several novel antiseizure candidates for further preclinical testing.⁷⁶ Similarly, zebrafish models of Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis (ALS) are under active study. Advanced imaging and reporter lines allow visualization of protein aggregates or neuronal death, and behavioral paradigms (e.g., habituation or arousal) can indicate subtle deficits. As a consequence, high-throughput screens in zebrafish have uncovered new neuroprotective leads: larvae treated with libraries of FDA-approved drugs yielded clusters of compounds that phenocopy known calcineurin inhibitor effects pointing to candidates for dementia prevention.⁷⁷ Given that behavioral or cellular phenotypes in zebrafish often mirror human pathology, therapeutic hits in fish could inform clinical development in neurology.

Cardiovascular and metabolic disease

Cardiovascular disease modeling is another area of active zebrafish research. The zebrafish heart develops rapidly and is anatomically similar to early human cardiac structures; notably, zebrafish adults can regenerate cardiac tissue after injury (unlike mammals). These features have been leveraged in small-molecule screens for heart therapies. For example, in addition to the heart failure reporter screen mentioned above, other efforts have focused on genetic heart disease. A notable case is the use of a kcnh2 (hERG) mutant fish with prolonged cardiac repolarization: screening for suppressors of the long-QT phenotype yielded compounds that normalize heart rhythm.⁷⁸ Beyond the heart, zebrafish larvae have been used to model vascular development and metabolic disorders (e.g., lipid metabolism and glucose homeostasis), enabling in vivo testing of anti-atherosclerosis or antidiabetic drugs. For example, dietary and genetic models of hepatic steatosis, such as larvae fed a high-fat/high-cholesterol diet, develop rapid, quantifiable liver lipid accumulation that has been used for medium-to-high throughput screens to identify anti-steatosis compounds and modulators of lipid handling.^30,79 Likewise, transgenic β cell-ablation paradigms have enabled large chemical screens that uncovered adenosine-pathway agonists (e.g., NECA) as enhancers of β cell regeneration, with hits subsequently validated in mammalian models.^80,81 The whole-animal context is crucial: zebrafish screens inherently assess systemic effects on blood flow, circulation, and organ function, providing a translational link to potential human therapies that in vitro assays lack.

Infectious disease and host–pathogen models

Infectious disease is a growing focus of zebrafish research becaue of their conserved innate immune system and the ease of real-time infection imaging. Zebrafish embryos (and larvae) can be infected by immersion or microinjection of human pathogens (bacteria, fungi, or viruses) to create in vivo models of infection.⁸² These models are now being used for antimicrobial drug screening. For example, larvae infected with Mycobacterium marinum (a close relative of Mycobacterium tuberculosis) recapitulate many features of human TB, including granuloma formation. In a recent screen for antitubercular compounds, zebrafish infection assays identified a novel aspartyl-tRNA synthetase inhibitor with in vivo efficacy (not previously recognized by standard in vitro tests).⁸³ Importantly, whole-animal infection models can reveal drug activities missed by cell culture: the frontline TB drug pyrazinamide is inactive against M. tuberculosis in vitro at low pH but is curative in patients; this paradox was resolved when zebrafish models demonstrated that pyrazinamide requires host metabolic activation to kill bacteria.⁸⁴ More broadly, zebrafish embryos offer a rapid and ethical platform to evaluate antibiotics (and antivirals) for efficacy and toxicity simultaneously. In one case, a zebrafish model of latent mycobacterial infection was engineered to have drug-tolerant ‘persisters’; this model is now used to screen for drugs that shorten TB treatment by killing dormant bacteria.⁸⁵ Likewise, zebrafish have been used to study Staphylococcus, Pseudomonas, fungal pathogens and even viral infections (e.g., influenza and herpes), allowing compound screening in a living host context. Overall, zebrafish infection models provide a cost-effective bridge between in vitro hits and mammalian testing, improving the translational yield of antimicrobial discovery.

Together these models enable whole-organism assessment of small molecules across diverse pathophysiological contexts. Lower cost moves preclinical results earlier in the pipeline and improves success rates in mammalian models, while enabling in vivo testing of drug combinations and dosing regimens that are cost-prohibitive in mammals, yielding new targeted-plus-immunomodulator combinations for cancer before human studies. This translational impact has driven growing FDA and NIH recognition of zebrafish as a NAM.

Quantitative metrics in zebrafish drug screening

Although commonly touted as a bridge between cellular assays and rodent models, the superior throughput of zebrafish screening is infrequently supported by specific quantitative comparisons.^10,86 An adult mating pair produces hundreds of synchronized embryos each week, which can be arrayed one to three per well in standard 96-well plates. This scale enables truly large screens: for example, libraries of 10 000–100 000 compounds have been assayed in zebrafish embryos.⁸⁶ By contrast, equivalent vertebrate studies in rodents are limited by small litter sizes and slow breeding. In practice, a single zebrafish facility can test on the order of 10³–10⁴ compounds per week (depending on automation and readouts), whereas mouse- or rat-based in vivo screens seldom exceed tens of compounds in the same time frame.

The cost efficiency of zebrafish assays is also dramatically better than for mammals. Husbandry and infrastructure are simple and inexpensive: a small zebrafish facility (dozens of 2–3-L tanks) can be built for just a few thousand dollars.⁸⁷ Maintenance labor is minimal, and feeding requirements are trivial for larval stages. These savings compound when running screens, with estimates that a zebrafish toxicology screen can cost nearly 500-fold less than a comparable rat-based screen.⁸⁷ In addition, assays can be run in parallel without the space and regulatory burdens of rodents. Beyond the cost of animals zebrafish represent, time efficiency as a major advantage. Zebrafish embryos develop rapidly: by 4–5 days post-fertilization they already have a beating heart, functioning liver, and formed brain. Thus, treatments and phenotypic readouts can be carried out in days.^10,88 For instance, behavioral or cardiac assays often begin by 3–5 days post-fertilization and conclude within a week. By contrast, mammalian models require weeks to months: mouse pups are born only after ~3 weeks of gestation plus postnatal maturation, and tumors in xenograft models can take several weeks to establish.^89,90 Given their rapid embryonic development, a zebrafish-based screens can yield preliminary toxicity and developmental toxicity or efficacy data in days, accelerating the pipeline.

Another key metric is compound use per treatment. The tiny size and aquatic dosing of zebrafish larvae mean that each test requires only microvolumes of drug solution. Typically, 1–3 larvae in a single well of a 96-well plate are housed in ~100–300 μl of embryo medium, into which compounds are added. In mouse, assuming a single dosing study, a reasonable concentration of 10 mg/kg injected into a 20-g mouse requires 0.2 mg of compound per mouse. In zebrafish, one embryo treated in 300 μl of media, at the common screening concentration of 10 μM (for a 300 MW molecule) is 0.9 μg of compound, a more than 200-times reduction in compound usage. The low compound requirements of zebrafish not only save on chemical costs, but also enable screening precious or limited molecules, such as those made in a medicinal chemistry campaign when milligram-level synthesis is the norm.

In summary, zebrafish screens provide a favorable balance of throughput, speed, and cost relative to mammalian testing. The high fecundity and small size of zebrafish allow thousands of embryos to be assayed in parallel. Each screen can be completed in days with minimal labor and requires only tiny drug doses per animal efficiencies to yield orders-of-magnitude gains: per-assay cost has been found to be several hundred times lower in zebrafish than in rodents and throughput is similarly expanded.

Zebrafish models in regulatory preclinical evaluation

Regulatory agencies now recognize zebrafish as a valuable new alternative method in drug development, provided its use is scientifically justified. The FDA’s current framework explicitly ‘permits and encourages’ validated non-animal assays (including model organisms) in place of, or to complement, traditional toxicology tests.^91–93 In its scientific roadmaps and guidance, FDA lists zebrafish as a promising ‘alternative organism’ for safety and efficacy screening.⁹³ For example, FDA’s New Alternative Methods program highlights zebrafish in research projects (e.g., developmental neurotoxicity studies) aimed at informing regulatory decision-making.⁹² In 2025, the FDA announced a paradigm shift, phasing out mandatory animal testing for new biologics and other drugs, and encouraged inclusion of NAM data in investigational new drug (IND) submissions to accelerate development and reduce animal use.⁹⁴ These initiatives imply growing openness to zebrafish data, especially when submitted through early-stage (Q-submission or pre-IND) consultations that clarify assay validity and context of use. Internationally, the ICH S5(R3) guideline on reproductive toxicology now explicitly acknowledges that ’qualified alternative assays, such as non-mammalian in vivo assays’ (e.g., zebrafish) can replace or reduce mammalian testing for detecting developmental hazards under defined circumstances.⁹⁵

Zebrafish data in IND submissions

In practice, zebrafish data are incorporated into IND applications chiefly as supportive safety information rather than as mandatory evidence. IND applications must include nonclinical pharmacology and toxicology data to justify first-in-human trials. Zebrafish studies can enrich these sections by flagging potential hazards or confirming on-target effects. Typical data types include in vivo embryotoxicity/teratogenicity screens, organ-specific toxicity assays (e.g., cardiac, hepatic, or neurobehavioral phenotypes), and high-throughput pharmacology studies. Sponsors often use such screens to triage compounds before committing to costly rodent trials.

Some INDs have explicitly reported zebrafish safety data, and have been ‘accepted’ by both the FDA and EMA in support of IND approval.⁹⁶ More recently, zebrafish developmental toxicity results have been submitted in investigational drug dossiers in the USA and abroad.⁹⁵ In their validation study (aligning to draft ICH S5(R3) requirements), 76% of known teratogens and non-teratogens were correctly classified by the zebrafish assay.⁹⁵ Such data were interpreted alongside mammalian studies to build confidence. Importantly, zebrafish data have never formally replaced the required rodent/nonrodent tox package, but regulators permit their inclusion as mechanistic or predictive evidence. Current FDA guidances encourage sponsors to justify any alternative model in the IND cover letter or pre-IND meeting request, ensuring reviewers understand how zebrafish results relate to human risk.

Regulatory guidance and case examples

Regulators have begun to provide formal contexts for zebrafish use. The revised ICH S5(R3) (finalized in 2020) specifically mentions qualified alternative assays: it states that data from a non-mammalian in vivo assay can be used for hazard identification under certain conditions.⁹⁵ This means, for example, that a properly validated zebrafish embryo assay might be acceptable for assessing embryo-fetal lethality or malformations in lieu of an additional mammalian test, once its concordance is demonstrated. No FDA guidance yet mandates zebrafish, but multiple guidance documents (ICH S5, S6, etc.) allow flexibility: sponsors ‘are encouraged to consult’ if using innovative models, as long as the data are scientifically valid.^91–93 The FDA’s recent roadmap for monoclonal antibodies (April 2025) even envisions a future in which NAM data (including potential zebrafish results) might partially replace traditional toxicity studies.⁹⁴

Concrete case examples are emerging. At the NCTR, zebrafish have been used to characterize the developmental neurotoxicity of environmental and drug exposures.⁹² Peer-reviewed studies also document predictive success: in a blinded validation of the classical zebrafish teratogenicity protocol, ~90% of reference teratogens were correctly identified.⁹⁵ As a case in point, aspirin (a known teratogen) induced characteristic malformations in zebrafish embryos at doses correlating to mammalian teratogenic levels. By contrast, agents without known developmental risk (e.g., certain analgesics) showed no effect in fish. Such concordance has helped convince regulators of the value of the assay.

Despite these advances, zebrafish remain a complementary regulatory model. The FDA still expects sponsors to justify any nonstandard data. Nevertheless, as the science matures and frameworks, such as the ICH S5(R3) guideline, provide qualification paths, zebrafish assays are gaining acceptance as fit-for-purpose tools. In summary, the FDA views zebrafish data as useful emerging evidence in IND packages, particularly for developmental and organ-specific safety questions, but requires that these data be rigorously validated and interpreted alongside conventional studies.^91–93

Challenges and limitations

Despite its advantages, the zebrafish model presents several pharmacological and technical challenges that complicate translation to human medicine. Compounds in zebrafish are typically delivered via immersion or microinjection rather than oral or intravenous routes, thus, effective uptake is highly sensitive to physicochemical factors. Small-molecule absorption through larval skin and gills depends strongly on water pH, buffering, and compound solubility.⁹⁷ Although zebrafish have many conserved drug targets (>80% of human disease genes have orthologs¹²), absorption, distribution, metabolism, excretion (ADME) processes remain only partially understood. For example, zebrafish larvae express orthologs of cytochrome P450s and transporters, but their enzyme activities and tissue distribution might differ from those in mammals. Extensive efforts are required to calibrate dosing: tracer studies show that compounds often accumulate in the yolk or exhibit limited brain penetration in early larvae. Mass-spectrometry and radiolabel methods can quantify internal exposures,⁹⁷ but these are low-throughput and technically demanding. In practice, therefore, correlating effective zebrafish doses with human pharmacology is nontrivial. Notwithstanding these issues, evidence from decades of screening indicates that zebrafish phenotypic outcomes generally correlate with mammalian efficacy.¹⁷ For instance, many drugs that modulate conserved targets (e.g., MEK inhibitors, serotonin ligands, and common analgesics) show similar effects in fish and humans.¹⁷ However, important caveats remain: for example, very large molecules or highly lipophilic drugs might fail to penetrate the embryo, and certain biotransformation steps differ; therefore, negative results in fish cannot categorically rule out human activity. Thus, careful in vitro characterization of zebrafish ADME and use of multiple assays are essential to interpret pharmacological findings.

Technical issues also limit zebrafish drug studies. Microinjection, a cornerstone of disease modeling and xenografting, is labor-intensive and user dependent. Even trained researchers achieve only modest throughput (manual injections often take tens of seconds per larva) and results can vary by operator. For example, Ding et al. reported that manual embryonic injections yield ~60% success and ~70% survival, with high user-to-user variability.⁹⁸ Imaging and data collection similarly face constraints: small animal size demands high-resolution microscopes and careful mounting. Larvae must be oriented reproducibly, often in multi-well plates or agarose chambers, to enable automated microscopy. In practice, most labs rely on 96- or 384-well formats, but even then, tracking and quantification of complex phenotypes (morphology, cell migration, behavior) can be subjective. Behavioral assays and live imaging introduce additional noise. Furthermore, zebrafish husbandry lacks some of the rigor seen in rodent facilities, leading to batch effects. Factors such as strain background, feeding, light cycles, and water chemistry can subtly influence endpoints. The community has begun to address this: a recent ‘Zebrafishology’ guideline highlights common pitfalls and recommends standardized recording of husbandry conditions and experimental variables.⁹⁹ Nonetheless, inconsistent practices remain an ongoing issue.

The target-agnostic nature of phenotypic screening is both a strength and a limitation. It identifies compounds with desired biological effects without prior target knowledge, uncovering novel mechanisms or polypharmacology that target-based strategies miss, particularly valuable in zebrafish, where whole-organism phenotypes are rapidly assessed in a physiologically relevant context. However, target deconvolution, pinpointing the molecular mechanism behind a hit, remains challenging. Follow-up methods, such as affinity proteomics, CRISPR/Cas9 knockdowns, and computational modeling, can be confounded by off-target interactions, compound promiscuity, and vertebrate complexity. Transgenic lines and behavioral assays aid validation, but deconvolution often requires mammalian-model confirmation, slowing progression to clinical development.

To mitigate these challenges, researchers are developing multiple strategies. On the pharmacology side, combining zebrafish data with in vitro ADME profiling can guide dose selection. Fluorescent or radiolabeled probes can be used to measure uptake kinetics in vivo.⁹⁷ Technically, automation is a powerful corrective. Robotic injection systems (see below) dramatically reduce user dependence, as do automated larval sorters and liquid handlers. Microfluidic devices and agarose molds immobilize embryos for uniform delivery. Advanced reporter lines, such as transgenics that emit signals upon metabolic activation or toxic stress, help standardize readouts. Computational tools and machine learning also have a role in mitigation: AI image analysis can remove human bias (see below). Finally, coordinated efforts to share protocols, raw data, and metadata are improving transparency. As the field continues to mature, the combination of community guidelines and technical innovation will help address current limitations.

One final underappreciated barrier to broader zebrafish adoption is not technical feasibility, but model perception. In some clinical and regulatory settings, zebrafish studies are still discounted because the organism is evolutionarily distant from mammals, regardless of published evidence supporting conserved drug responses, disease mechanisms, and predictive phenotypes. This skepticism can create an adoption bias in which mammalian systems are presumed to be more relevant, while zebrafish models are required to provide additional layers of justification. Addressing this barrier will require rigorous benchmarking, standardized experimental design, transparent reporting of assay performance, and clearer communication that zebrafish are best used as fit-for-purpose vertebrate systems that bridge in vitro assays and mammalian studies. The convergence of high-content phenotyping, AI-driven analysis, and new regulatory interest in alternative models could make this an especially timely moment to overcome these perception barriers.

Emerging technologies and future perspectives

Rapid technological advances in automation and robotics are explanding the zebrafish toolbox. Microinjection is a foundational technique in delivery of cargoes into the zebrafish, ranging from RNA/DNA to cancer cells; however, it is also a technique that is labor intensive and with significant variability in speed and accuracy between users. State-of-the-art robotic microinjectors now perform high-throughput injections of embryos and larvae. Ding et al. engineered an automated system that locates each larva and injects precisely into the vasculature or tissue, operating at approximately twice the speed of trained humans. This platform achieved ~60% success and ~70% survival rates, comparable to expert operators, but halved the per-larva injection time.⁹⁸ Similarly, Guo et al. described a vision-guided batch injector with an 88-sample cycle; it attained over 92% injection success and ~94% survival, with each injection taking ~14 s.¹⁰⁰ Such robots vastly increase throughput: an automated system can inject hundreds of larvae per day, whereas manual methods are an order of magnitude slower. These systems also minimize human error (see below). In addition to injection, robotic platforms for embryo sorting, multiwell plate handling, and drug dispensing are under development. With these tools, high-volume screening and complex multi-step protocols can be automated, reducing variance between experiments.

Advances in microscopy to leverage the optical transparency of zebrafish are being made with new microscopes and analysis pipelines. Lightsheet and confocal systems now enable whole-embryo volumetric imaging over time, capturing cellular dynamics in vivo. Coupled with automated analysis, these systems can quantify complex phenotypes at scale. Sturtzel et al. developed a high-content imaging workflow for zebrafish xenografts: larvae are embedded in agarose within 96-well plates, and a software pipeline performs automated tumor detection and size quantification day-to-day.¹⁰¹ Similarly, Hoade et al. implemented the WiScan Hermes high-content microscope with an AI-based ‘Athena’ app: it automatically recognizes fish and anatomical landmarks in each well and enumerates fluorescently labeled cells with >98% accuracy compared with manual counting.¹⁰² These tools eliminate manual imaging bottlenecks and enable truly high-throughput phenotyping. Looking forward, integration of whole-brain imaging (via virtual brain atlases) and single-cell readouts (e.g., in situ transcriptomics in larvae) could link behavior, morphology, and molecular profiles. In all cases, these imaging advances harness quantitative data: screens can report effect sizes (fold-change in tumor area or behavioral metric units) rather than subjective scores, enhancing the rigor of readouts.

AI is now being deployed throughout zebrafish drug discovery, enabling new capabilities in data mining and experimental design. For example, ‘EmbryoNet’ is a deep convolutional network trained on over 2 million images of zebrafish embryos with defined pathway perturbations.¹⁰³ It can automatically recognize subtle morphogenetic defects linked to specific signaling pathways, even before they are obvious to human observers, and was used to flag unexpected developmental toxicity of FDA-approved drugs.¹⁰³ Similarly, ‘Marigold’ is a web-based pose-tracking app that uses a streamlined neural network to track ten body points on larval zebrafish in multiwell plates.¹⁰⁴ This body segmentation allows for greater granularity of detail in behavioral analyses. Crucially, Marigold runs in a browser without specialized hardware democratizing this complex behavioral analysis.¹⁰⁴ As these tools mature, AI will accelerate every stage of the workflow, from identifying hit compounds to characterizing phenotypes, complementing the inherent scalability of zebrafish.

Zebrafish are also increasingly used for personalized medicine. In oncology, PDX in fish, ‘zebrafish avatars’, are being tested as predictive bioassays. Advantages of zPDX include the need for only ~100–200 patient cells, a 5–7-day assay time, and the ability to directly visualize tumor behavior in the living host.⁷³ In practice, tumor cells from a biopsy can be implanted into dozens of larvae, which are then treated with candidate drugs; readouts such as tumor regression, angiogenesis, or migration can generate a chemosensitivity profile within a week. Early clinical studies are encouraging. Costa et al. reported that zebrafish avatars of colorectal cancer, treated ex vivo with the same chemotherapy regimens as the patient, predicted the patient’s progression with ~91% accuracy.⁷¹ Patients whose zAvatar showed drug sensitivity indeed had significantly longer progression-free survival.

Beyond oncology, precision genetic engineering is transforming zebrafish modeling. Precise editing in zebrafish can introduce patient-specific alleles and recapitulate disease etiology. New Cas variants and base-editors expand the editable sequence space. For instance, SpRY-based cytosine and adenine base editors achieved up to ~96% editing efficiency at nearly PAM-less sites in zebrafish, enabling insertion of mutations previously inaccessible.¹⁰⁵ These innovations mean that almost any human single-nucleotide variant can be modeled in vivo, in the correct genomic context. Moreover, emerging techniques for conditional and tissue-specific editing, including Cre/lox or CRISPR–Cas12 tools, allow perturbation of genes in desired cell types or developmental windows, mimicking the mosaic nature of many diseases. Taken together, these genetic tools promise ‘precision zebrafish models’ that closely mirror patient pathobiology, allowing in vivo investigations into ‘precision medicine’ at a fraction of the cost of the same in mouse, facilitating target validation and personalized drug testing.

Looking ahead, several trends point to a bright future. Integrative approaches will emerge: for instance, combining zebrafish with organ-on-chip or organoid models to capture multi-organ interactions. Whole-animal metabolomics or single-cell RNA-seq in drug-treated larvae could map the full scope of drug action. Microbiome-engineered fish could reveal gut–drug interactions. Advances in in vivo sensors (e.g., fluorescent reporters of neural activity, ion concentrations, or redox state) will allow dynamic readouts of drug effects in real time. Scalability will keep improving: robotic systems could eventually handle 384- or 1536-well plates, and in-line AI analysis could yield real-time decision-making during screens. In the clinic, consortium efforts might validate zebrafish assays against patient outcomes in large, systematic trials. Finally, there is potential for ‘digital twin’ models: computer simulations trained on zebrafish and human data to predict responses in silico.

Concluding remarks

The zebrafish model occupies a unique niche in drug discovery, offering a whole-animal system with high genetic and physiological relevance, yet with throughput orders of magnitude greater than rodents. In embryos and larvae, researchers can observe developmental and disease processes at single-cell resolution in vivo. This has already enabled discoveries, such as novel pathway modulators and drug repurposing leads that likely would have been missed in cell cultures. However, translating findings from fish to human requires care. Zebrafish and humans differ in size, anatomy, and some aspects of metabolism, which can introduce false negatives or positives. Thus, it is important to view zebrafish as a complementary bridge between cell-based screens and mammalian models.

Looking forward, the balance of evidence suggests that zebrafish will only become more powerful. New technologies are directly addressing current limitations: automated robotics reduce human variability; advanced imaging and AI extract richer quantitative data; and precise genome editing generates humanized disease models. At the same time, the accumulated 20-plus years of zebrafish pharmacology, together with rigorous standardization efforts, means that the field is maturing into a reliable preclinical platform. In sum, although not a panacea for the ills of drug discovery, the zebrafish has proven utility in phenotypic discovery and early translational research.

Highlight statements.

• Zebrafish enable high-throughput whole-organism drug screening.

• Genetic tools like CRISPR model human diseases in zebrafish.

• Zebrafish avatars predict patient drug responses with high accuracy.

• Automation and AI enhance zebrafish screening efficiency.

• Zebrafish data gain traction in regulatory preclinical evaluation.

Biographies

Charles H. Williams III

Charles H. Williams III is an assistant professor of Medicine at Michigan State University and chief scientific officer of Proton Bio Inc. His research focuses on the role of extracellular pH as an intercellular signaling system, in particular proton-sensing G-protein-coupled receptors, and their role in cancer, cardiovascular, and inflammatory diseases. Combining zebrafish and mammalian models, his laboratory develops and evaluates small-molecule inhibitors that modulate pH-dependent signaling and ferroptosis. His work integrates chemical genetics, translational pharmacology, and phenotypic screening to advance therapeutic discovery and precision medicine.

Veona Cutinho

Veona Cutinho is a research assistant in the Williams Laboratory in the Department of Medicine, Michigan State University. Her research investigates the role of proton-sensing receptors as potential therapeutic targets in tissue regeneration and traumatic brain injury. She holds a Bachelor’s degree in genomics and molecular genetics. Previously, she worked in Jose Cibelli’s Cellular Reprogramming Laboratory, studying epigenetic modulators of chromatin accessibility to enhance somatic cell nuclear transfer. Veona has over three years of experience in in vivo zebrafish research.