A brief guide to the in vivo chemogenetic cell ablation approach in zebrafish

Abstract

In vivo cell ablation technologies are essential tools for understanding biological processes within living animal models. The nitroreductase (NTR)/metronidazole system enables highly effective spatiotemporal cell ablation. Using transgenic zebrafish that combine NTR2.0 with the QF3/QUAS binary gene expression system, conditions to achieve efficient cell type–specific chemogenetic ablation were optimized. This approach provides a versatile in vivo platform for investigating developmental processes and regeneration, as well as for disease modeling and drug discovery.

Graphical abstract

INTRODUCTION

Cell ablation systems that enable the selective removal of specific cell populations represent powerful and essential technologies to provide critical insights into fundamental biological processes, including cell lineage and function, tissue homeostasis, regeneration, and behavior in multicellular organisms (Bruttger et al., 2015, Xiao et al., 2022). Such knowledge not only advances our understanding of developmental programs and disease pathogenesis but also facilitates the drug discovery of novel therapeutic targets (Keum et al., 2024, Liu et al., 2019).

Over the past several decades, diverse methods for targeted cell ablation have been developed, each with distinct features, advantages, and limitations (Keum et al., 2024, Xiao et al., 2022). The characteristics of the most commonly used systems, such as physical laser, toxins, enzymes, and photosensitizers, are summarized in Table 1 (Bruttger et al., 2015, Chung et al., 2013, Labbaf et al., 2022, Makhijani et al., 2017, Marshall et al., 2022, Sharrock et al., 2022). Among these, the bacterial nitroreductase (NTR)/ metronidazole (MTZ) system is a widely used chemogenetic approach for targeted cell ablation. NTR reduces nitro groups (−NO₂) to amino groups (−NH₂), thereby converting the otherwise nontoxic prodrug MTZ into cytotoxic metabolites that bind DNA, cause double-strand breaks, and trigger apoptosis (Fig. 1a). This system provides the spatiotemporal control of cell ablation through promoter-driven NTR expression and MTZ treatment, and has been widely used in zebrafish regeneration research (Chung et al., 2013, Xiao et al., 2022).

Table 1.

Diverse methods for targeted cell ablation system

Method	Specific tool	Mechanism	Advantage	Limitation	Reference
Physical	Laser ablation	Converts intense light into heat to damage cellular proteins, causing cell death within seconds	Allows for precise spatial and temporal control	Time-consuming, requires expensive equipment, and may cause thermal damage to surrounding tissues	Marshall et al. (2022)
Optogenetic (photosensitizer/reactive oxygen species [ROS])	miniSOG, miniSOG2	Genetically encoded photosensitizers produce ROS, upon light excitation, which damages cellular components and activates apoptosis	Combines genetic specificity with precise control using light	The depth of light penetration into tissue is limited	Makhijani et al. (2017)
Chemogenetic (enzyme/prodrug)	Nitroreductase (NTR1.0 or 2.0)/metronidazole (MTZ)	NTR converts the nontoxic prodrug MTZ into a cytotoxic product that causes DNA damage and induces apoptosis in NTR-expressing cells	High cellular specificity; can eliminate NTR-expressing cells	The reaction time is relatively slow; may be less effective on certain cell types; impossible to ablate only a subset of cells	Chung et al., 2013, Sharrock et al., 2022
Chemogenetic (toxin/antitoxin)	Kid/Kis toxin system	Bacterial toxin Kid acts as an endoribonuclease, interfering with protein synthesis, inducing apoptosis rapidly. The antitoxin Kis can inhibit Kid function	Extremely rapid kinetics compared to the NTR system; tunable system (severity and timing can be modulated by expressing Kis)	Efficacy needs further validation across various cell types; a specific gene expression system is required	Labbaf et al. (2022)
Genetic/toxin	Diphtheria toxin (DT/DTR)	First, engineer targets cells to express the diphtheria toxin receptor (DTR); then, administration of DT binds specifically to DTR-expressing cells, leading to internalization of the toxin and inhibition of protein synthesis	High cellular specificity; can eliminate DTR-expressing cells	Extreme toxicity; potential off-target effects; cannot be used in most nonrodent species (rabbits, pigs) due to endogenous DTR	Bruttger et al. (2015)

Fig. 1.

Schematic overview of the generation and applications of a cell type–specific ablation zebrafish model. (a) Upon MTZ incubation, only NTR-expressing cells convert MTZ into cytotoxic metabolites, leading to DNA damage and apoptosis. (b) Cell type-specfic promoter drives QF3 transcriptional activator expression, which induces NTR2.0 expression under the 13xQUAS sequence. (c) The combined QF3/QUAS-NTR2.0 system enables the generation of cell type–specific ablation zebrafish models, which can be applied to developmental and regeneration studies, disease modeling, and high-throughput drug screening. All schematic illustrations were created using BioRender.com.

In this MiniResource, we describe methods and applications for achieving versatile and precise tissue–specific cell ablation in zebrafish, using the recently developed NTR2.0 in combination with QF3/QUAS binary gene expression system.

MAIN BODY

System Design

NTR/MTZ Cell Ablation System

First-generation NTR systems (eg, Escherichia coli NfsB, or NTR1.0) often required relatively high, and potentially toxic, MTZ concentrations (eg, 10 mM) to achieve effective ablation (Chung et al., 2013). In contrast, NTR2.0, a rationally engineered variant from Vibrio vulnificus, markedly enhances MTZ–mediated ablation efficacy by approximately 100-fold, allowing robust ablation at much lower concentrations (eg, 0.1-2 mM) (Sharrock et al., 2022). This advancement supports persistent cell-loss models and permits ablation of cell types that were formerly resistant (eg, neurons and macrophages), with retained cell type-specificity and no observable bystander effects.

QF3/QUAS Binary Gene Expression System

To facilitate spatial and temporal cell ablation, a binary gene expression system is indispensable. Although the GAL4/UAS system is the most widely used genetic tool in animal models, its use in zebrafish is limited by transgene silencing, hampering its broader applications (Hong et al., 2021). To address this issue, a novel genetic tool based on the QF/QUAS system, originating from the quinic acid–sensing system of the fungus Neurospora crassa, was adapted in zebrafish (Potter et al., 2010). This system was engineered to overcome driver-associated cytotoxicity and transgene silencing by incorporating the optimized transactivator QFDBD-2xAD*-VP16* (hereafter QF3, also coined “EQ-On”) together with a 13xQUAS effector that dictates robust transgene expression (Fig. 1b; Hong et al., 2021).

Taken together, combining the NTR2.0 and QF3/QUAS systems allows a highly versatile in vivo cell ablation system in a spatiotemporal manner. Upon treatment of MTZ at defined time points, targeted tissues can be selectively ablated within living organisms, with varying degrees of ablation depending on the drug treatment conditions. Furthermore, in animals with regeneration abilities, including zebrafish, subsequent withdrawal of MTZ allows regeneration of the ablated cells. Thus, this model enables developmental and regenerative studies and can also be used for disease modeling and small molecule screening (Fig. 1c). In this report, we highlight its applications and utilities by presenting experimental implementation of neuron– and intestine–specific cell ablation.

Validation and Representative Applications

Transgenic Lines

In this study, the AB line was used as the wild-type zebrafish background. To achieve cell type–specific expression of NTR2.0, the Tg(th:QF3) and Tg(cldn15la:QF3) driver lines that harbor relevant regulatory elements were generated to target dopaminergic (DA) neurons and intestinal cells, respectively. The Tg(th:QF3) line was established by modifying a previously described Tg(th:QF2) strategy (Altbürger et al., 2023), and the Tg(cldn15la:QF3) line was constructed using the cldn15la promoter subcloned from the Addgene plasmid (#125026) originally deposited by Dr John Rawls’ lab. The effector line Tg(13xQUAS:YFP-2A-NTR2.0) was designed to place yellow fluorescent protein (YFP) and NTR2.0 downstream of the 13xQUAS sequence, linked by a self-cleaving 2A peptide. This construction enables simultaneous expression of YFP and NTR2.0, allowing visualization of NTR2.0 expression via the YFP signal. YFP-2A-NTR2.0 cassette was obtained from the Addgene plasmid (#158651) originally deposited by Dr Jeff Mumm’s lab. All plasmids were purchased from Addgene website (https://www.addgene.org/). All transgenic lines were generated using the Tol2 transposon system (Kwan et al., 2007). The expression construct plasmids were mixed with Tol2 transposase mRNA with the final concentration of 50 and 30 ng/μl, respectively, and 1 to 2 nl of the mixture was injected into the 1-cell stage of embryos to generate transgenic line. After 3 months, those fish were crossed with AB wild-type fish to identify stable lines by screening expression of the target gene.

Mating and Embryo Collection

Driver and effector transgenic lines were put together in a mating chamber (typically 3 males and 2 females), separated by sex with a divider placed between them. The next morning, the divider was removed and the fish were allowed to mate for 1 hour. Embryos were collected and maintained in E3 medium at 28.5℃. YFP-positive larvae carrying both the driver and effector transgenes were selected at 2 days post-fertilization (dpf). Fifteen selected larvae were placed in each well of a 6-well plate containing 6 ml of MTZ solution (Fig. 2a). All experiments were conducted in accordance with the guidelines of the Korea Research Institute of Bioscience and Biotechnology (KRIBB) and were approved by KRIBB-IACUC (approval number: KRIBB-AEC-25110).

Fig. 2.

Validation of DA neuron and intestinal cell ablation in zebrafish. (a) The experimental timeline of (c). Driver and effector fish were mated, and the collected eggs were maintained at 28.5℃. YFP-positive larvae were selected at 2 dpf and incubated with MTZ (50, 100, 500 μM, and 2 mM) for 24 hours at 3 dpf. Fifteen larvae were placed in 6 ml of MTZ solution per well of a 6-well plate. Fluorescence imaging was conducted at 4 to 5 dpf to examine ablation and regeneration of DA neurons. (b) A dorsal-view schematic of the ventral diencephalic dopaminergic neuron cluster (DC); PTar (DC2), Anterior PT (Posterior Tuberculum) rostral group; Hdm (DC3), Dorsal medial hypothalamus; PTac (DC4), Anterior PT caudal group; Ptp (DC5), Posterior PT group; PTN (DC6), Posterior tuberal nucleus; Hc (DC7), Caudal hypothalamus. (c) Representative images showing ablation of DA neurons at different MTZ concentrations following 24 hours of treatment and subsequent regeneration after MTZ washout. The fluorescence image in the leftmost column shows DA neuron clusters (DC2–DC7) at 4 dpf, annotated according to the schematic in (b). A dorsal view; rostral to left. (d) Quantification of DA neuron volume shown in (c). Data are presented as mean ± SEM (n = 8 per group). ns, not significant. ***P < .001, ****P < .0001 (RM one-way ANOVA with Tukey's post-hoc test). (e) Representative images showing DA neurons in Tg(th:QF3, 13xQUAS:EGFP) larvae. Larvae were treated with DMSO or 10 mM MTZ for 24 hours starting at 3 dpf and imaged at 4 dpf. Scale bar, 25 µm. (f) Quantification of DA neuron volume shown in (e). Data are presented as mean ± SEM (n = 8 per group). ns, not significant. P = .2925 (Welch's t-test). (g) Representative images of anti–cleaved caspase-3 immunostaining in Tg(th:QF3, 13xQUAS:YFP-2A-NTR2.0) larvae. Larvae were treated with 10 mM MTZ for 3 hours at 4 dpf. YFP+ DA neurons (green), cleaved caspase-3+ apoptotic cells (red), and merged images are shown. Arrows indicate YFP+ cells undergoing apoptosis. Scale bar, 3 µm. (h) Quantification of cleaved caspase-3 and YFP colocalization. Graph shows the percentage of YFP+ cells that were cleaved caspase-3-positive in DMSO control and 10 mM MTZ-treated groups. Data are presented as mean ± SEM (n = 10 per group). ****P < .0001 (Welch’s t-test). (i) Representative images showing ablation of DA neurons immediately after 2 hours of treatment with 10 mM MTZ. A dorsal view; rostral to left. (j) Representative images showing ablation of intestinal cells immediately after 2 hours of treatment with 10 mM MTZ. A lateral view; rostral to left. Rectangular dotted lines in schematics in (i) and (j) represent the regions of imaging. Scale bars: 100 µm. All schematic illustrations were created using BioRender.com.

MTZ Treatment for In Vivo Cell Ablation

The ablation efficiency of DA neurons was assessed by treating Tg(th:QF3, 13xQUAS:YFP-2A-NTR2.0) larvae at 3 dpf with MTZ at different concentrations (50, 100, 500 μM, and 2 mM) for 24 hours (Fig. 2a). DA neurons in the ventral diencephalic cluster (Fig. 2b) were visualized by YFP fluorescence (Fig. 2c). Quantitative volume analysis using 3D rendering IMARIS software (Oxford instruments) with surface tool demonstrated the dose–dependent ablation efficiency, showing approximately 30% of DA neurons remaining after treatment with 50 and 100 μM MTZ, while near-complete ablation was achieved with 500 μM and 2 mM MTZ (<10% remaining) (Fig. 2c, d). Furthermore, following an 18-hour washout after 24 hours of treatment with 2 mM MTZ, DA neurons began to be rapidly regenerated, reaching approximately 21% of DMSO-treated controls (Fig. 2c, d, the last column). To confirm that the observed ablation is specifically mediated by NTR rather than nonspecific MTZ toxicity, EGFP-expressing Tg(th:QF3, 13xQUAS:EGFP) larvae were treated with 10 mM MTZ for 24 hours and no significant loss of DA neurons was observed (Fig. 2e, f).

To determine whether MTZ induces apoptosis in NTR-positive DA neurons, larvae were fixed and cryo-sectioned at 3 hours post MTZ treatment and processed for immunostaining with an anti–cleaved caspase-3 antibody (1:500; BD Biosciences, Cat. # 559565). More than 95% of YFP-positive DA neurons following MTZ treatment exhibited cleaved caspase-3 positivity, indicating that MTZ–mediated apoptotic ablation of NTR-positive DA neurons (Fig. 2g, h).

In addition, diverse cell types/tissues at different stages can be targeted for ablation using the combination of NTR/MTZ and QF3/QUAS systems. As an example, DA neuron–specific ablation using Tg(th:QF3, 13xQUAS:YFP-2A-NTR2.0) larvae was treated with 10 mM MTZ for 2 hours, the incomplete ablation of DA neurons was observed (Fig. 2i). In contrast, intestine-specific ablation using Tg(cldn15la:QF3, 13xQUAS:YFP-2A-NTR2.0) resulted in complete elimination of intestinal cells under identical condition (Fig. 2j). The differences in MTZ sensitivity between intestinal and neuronal cells may be due to the different route of MTZ exposure (eg, the intestine vs the blood–brain barrier), variations in QF driver promoter activity, cell-specific properties (such as drug metabolism and stress responses), and/or differential susceptibility to cell death.

Practical Considerations and Tips

The treatment of MTZ can be performed at a high concentration for a short duration or at a low concentration for a long duration at either early or late larval stages. Of note, it is clear that the required duration and concentrations of MTZ treatment may vary depending on target cell and tissue types for ablation. Therefore, parameters such as MTZ treatment conditions, developmental stages, and specific characteristics of cell and tissue types should be carefully considered to achieve successful ablation.

Tips

• 1.MTZ powder (Sigma-Aldrich, Cat. # M1547) was stored at 4°C protected from light. For each experiment, MTZ was freshly prepared by dissolving in DMSO and diluting with E3 medium to the desired final concentration, with the final DMSO concentration kept below 1% to avoid toxicity.
• 2.A previous report indicated that the long-term exposure to 10 mM MTZ is toxic to zebrafish (Sharrock et al., 2022), suggesting that concentrations of 1 mM or lower are more desirable for prolonged treatments.
• 3.To ensure reproducibility for experiments, the ratio of larvae to volume should always be kept constant (eg, 15 larvae/6 ml) during MTZ treatment.
• 4.Treatment 10 mM MTZ for 24 hours reliably induces efficient ablation across various cell types and experimental conditions, making it a recommended starting protocol for NTR/MTZ–based ablation experiments.

CONCLUDING REMARKS

This guide provides a novel and practical approach that combines the highly effective NTR2.0 enzyme with the robust, silencing-free QF3/QUAS binary gene expression system to achieve versatile and precise tissue–specific cell ablation in zebrafish. Optimization of MTZ concentrations and treatment duration confirmed the effectiveness of this system, revealing the dose-dependent efficiency and rapid regeneration after MTZ washout. We anticipate that this resource will be useful in a wide range of biomedical research fields for in vivo cell ablation studies.

Funding and Support

This work was supported by NRF Grant [RS-2022-NR067360, RS-2023-NR00225255] and the ABC-based Regenerative BioTherapeutics (ABC project) grant [RS-2024-00426031] funded by the Korea government (the Ministry of Health & Welfare and Ministry of Science and ICT) and KRIBB Research Initiative Program [KGM1102511, KGM9992521].

Author Contributions

Eun-Ji Lee: Writing – review & editing, Writing – original draft, Visualization, Investigation, Formal analysis, Data curation. Jeong-Soo Lee: Writing – review & editing, Writing – original draft, Supervision, Project administration, Funding acquisition, Data curation, Conceptualization. Jae-Geun Lee: Writing – review & editing, Writing – original draft, Visualization, Investigation, Data curation.

Declaration of Generative AI and AI-Assisted Technologies in the Writing Process

During the preparation of this manuscript, the authors utilized ChatGPT-4o (OpenAI) and Perplexity to enhance the clarity and readability of the text. All contents were subsequently reviewed and revised by the authors, who take full responsibility for the final version of the manuscript.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.