When to Integrate High-Throughput In Vivo Screening in Early Drug Discovery Pipelines

The early stages of drug discovery are defined by the need to evaluate as many compounds as possible, as quickly and cost-effectively as possible, while generating data that will actually predict what happens later in the clinic. In vitro assays offer throughput and cost-efficiency but lack the physiological context needed to detect systemic effects. Mammalian in vivo models offer biological relevance but are too resource-intensive to deploy at scale in the early stages of a pipeline. 

Zebrafish have emerged as a practical solution to fill this gap. As a whole-organism vertebrate model that is compatible with automated, high-throughput formats, zebrafish occupy a unique position in the drug discovery workflow. They provide in vivo data at a throughput that approaches in vitro assays, while delivering a level of physiological complexity that cell-based models cannot replicate. 

Strategic timing and integration of high-throughput in vivo screening in early drug discovery workflows

In a conventional drug discovery pipeline, in vitro assays are used to generate initial data on potency, selectivity, and ADME properties. Promising compounds enter then a much more resource-intensive round of in vivo testing in mammalian models, typically rodents. The problem with this binary structure is that in vitro assays, by definition, cannot capture systemic toxicity, organ-level effects, or pharmacokinetic behavior in a living organism. As a result, a significant number of compounds that pass in vitro screens fail at the in vivo stage, generating data that is expensive to produce and offers limited mechanistic insight into why the compound failed. Not to speak about the number of mammals that could have been saved. 

Zebrafish integrate naturally between these two stages. Their small size, high fecundity, and optical transparency make them amenable to the same miniaturized, automated formats used in in vitroscreening: 96-well plates, liquid handling systems, and automated imaging. At the same time, they share 70% of protein-coding genes with humans and 84% of human disease-related genes. The result is a model that combines the scale of in vitroscreening with the biological relevance of in vivo data.

Zebrafish also support the 3Rs principles of Replacement, Reduction, and Refinement, since they are not considered protected animals until they start feeding independently around 5 days post-fertilization (dpf), according to EU Directive 2010/63. They are considered a New Approach Methodology (NAM), thus providing systemic in vivo data without the ethical and regulatory constraints of conventional vertebrate models.  

The optimal point to introduce zebrafish screening is after a primary in vitro screen has generated an initial hit list, and before compounds are advanced into rodent studies. At this stage, the goal is not to replace mammalian models but to use zebrafish data to de-risk the transition. Compounds that show clear toxicity signals or lack of efficacy in zebrafish can be deprioritized before incurring the costs of rodent experiments, while compounds that perform well carry a higher probability of success in mammalian models.

The specific endpoints that zebrafish can address at this stage include general and organ-specific toxicity (cardiotoxicity, hepatotoxicity, neurotoxicity, developmental toxicity), behavioral phenotypes, and disease-model efficacy in transgenic or chemically induced disease lines. A single zebrafish model can generate safety and efficacy data, something that in vitro assays or mammalian models at this stage cannot do at comparable throughput.

Impact of introducing in vivo HTS between in vitro and mammalian models in early decision-making

The practical impact of inserting a zebrafish in early drug discovery is clearly illustrated by one of ZeClinic's recent case studies:

A small biotech company working on cardiac disease, specifically Dilated Cardiomyopathy (DCM) and Hypertrophic Cardiomyopathy (HCM), had already performed a first round of screening using human cardiomyocyte models, narrowing a library of 300 commercial compounds to 20 hits. 

Testing all 20 compounds directly in mammalian models would cost over $1.5 million and take more than 24 months, partly due to the extended breeding and growth periods required to reach the necessary sample sizes. ZeClinics' solution was to develop tailored zebrafish genetic models targeting the key genes involved in both cardiomyopathies using CRISPR/Cas9 technology, and to run a pre-screening campaign in zebrafish larvae under 5 days post-fertilization across all 20 compounds. We identified 5 candidates for progression to rodent evaluation. 

The consequences were substantial: 

• They reduced the timeline by 40%. The preclinical phase was compressed from 24 months to 14 months
• They reduced costs by 60%. Total preclinical screening costs fell from $1.5 million to $600,000.
• They saved hundreds of mice that would otherwise have been required to test 20 compounds instead of 5. 

This has implications beyond budget management. Faster decision-making at the compound selection stage means that promising compounds reach mammalian validation and ultimately clinical development faster. In a field where the average time from discovery to approval exceeds a decade, compressing the early stages of the pipeline by even a few months represents meaningful value.

Operational considerations: throughput, cost, and data quality in HTS deployment

Integrating zebrafish high-throughput screening for drug discovery involves operational decisions around assay formats, compound delivery, and data interpretation. In practice, however, these decisions do not need to be made internally. Working with a specialized CRO means that the infrastructure, expertise, and validated protocols are already in place: you define the scientific question, and we deliver interpretable, decision-ready data. We manage the operational considerations. 

At ZeClinics, we have the expertise and the capabilities to offer high-throughput screening services based on zebrafish, including two VAST BioImager for high-content morphological imaging and DanioVision for quantitative behavioral analysis across large numbers of individuals simultaneously. Deep-learning-assisted phenotyping captures sub-lethal, organ-specific, and functional effects that go beyond mortality and gross morphology.

The question of data quality, however, goes beyond throughput and reproducibility. The data generated at the zebrafish stage only justify the investment if it predicts what will happen later in mammalian models and, ultimately, in the clinic. This is where the biological relevance of zebrafish becomes the central argument. Zebrafish share approximately 84% of human disease-related genes and recapitulate a wide range of human pathologies at the whole-organism level.

A concrete example of this translational value comes from a collaboration with IOCB Tech on pharmacoresistant epilepsy. Our zebrafish epilepsy model reproduces epileptic convulsions in a stereotyped, concentration-dependent way when exposed to the convulsant agent pentylenetetrazole (PTZ), making them a practical model for anticonvulsant drug screening. Using a standardized light/dark protocol combined with DanioVision video tracking, ZeClinics screened a series of neuroactive steroids synthesized by IOCB Tech. The screenidentified a leading compound with clear anticonvulsant efficacy in zebrafish. The zebrafish data predicted the mammalian outcome, confirmed by subsequent validation in a rodent model. 

This is what translational value looks like in practice: move fast, quantify early, and only advance what works. For organizations looking to reduce attrition and compress timelines without building internal zebrafish infrastructure, a CRO partnership is the most direct path to achieving that.

ZeClinics can help you

By Javier Terriente

Javier is the co-founder of ZeClinics and ZeCardio Therapeutics, two biotech firms specializing in zebrafish-based preclinical drug discovery for cardiovascular, neural, and toxicology applications. He combines scientific leadership with business acumen, having successfully driven fundraising efforts and strategic partnerships.

Currently leading scientific efforts at ZeCardioTx (and formerly CSO at ZeClinics), Javier also serves on the Board of Directors of AseBio, where he advocates for industry collaboration. His academic background includes a PhD in Molecular Biology and a Marie Curie Fellowship. Recognized as an expert in zebrafish models, he has published extensively and has supervised five industrial PhD theses.