# Enabling mechanistic studies of EVs in vivo: a protocol for isolation and cell-specific labelling in larval zebrafish

## TL;DR

This paper introduces a new method to isolate and label extracellular vesicles in zebrafish larvae, enabling detailed study of their role in living organisms.

## Contribution

The study presents a novel protocol for cell-specific EV labeling and isolation in zebrafish larvae, enhancing in vivo mechanistic studies.

## Key findings

- Bacillus licheniformis protease outperforms collagenase in tissue dissociation without damaging cells.
- GW4869 inhibits EV biogenesis in a dose-dependent manner, validating the protocol's sensitivity.
- The UAS: CD63-GFP construct allows cell-specific EV labeling and tracking in intact zebrafish tissues.

## Abstract

Extracellular vesicles (EVs) are critical mediators of intercellular communication in development, physiology, and disease. In vivo models such as Drosophila melanogaster, Caenorhabditis elegans, and Danio rerio (zebrafish) now provide powerful platforms to visualize EV dynamics in real time. However, the full potential of these models remains underutilized due to the lack of reliable, cell-specific EV labelling tools and robust EV isolation protocols. Here, we present an optimized workflow for the isolation of EVs from zebrafish larvae and the in vivo labelling of EVs in a cell-type-specific manner.

To isolate EVs from larval zebrafish, we used size exclusion chromatography (SEC). By comparing different tissue digestion methods and performing step-by-step optimisation of sample preparation prior to SEC, we established a novel protocol that enables EV isolation without compromising cell viability. EV size and concentration were assessed by nanoparticle tracking analysis (NTA), with subsequent characterization by transmission electron microscopy (TEM) and Western blotting. To evaluate the sensitivity of our protocol, we treated zebrafish larvae with GW4869, a known inhibitor of EV biogenesis, and assessed the dose-dependent effects on EV release. To specifically label EVs from distinct cell types, we have generated a UAS: CD63-GFP construct which can be expressed under control of the Gal4 transcriptional activator.

Through a systematic comparison of tissue dissociation techniques, we identify Bacillus licheniformis protease as a superior alternative to conventional collagenase treatment, which compromises cell integrity. Treatment with GW4869 confirmed that EV biogenesis and release can be inhibited in a dose-dependent manner and demonstrated that our protocol is sensitive enough to detect and quantify changes in EV levels. To enable cell-specific EV tracking in vivo, we combined the UAS: CD63-GFP construct with a radial glia-specific Gal4 driver line, providing a proof-of-concept for targeted EV imaging in intact tissues.

These advances provide a versatile toolkit for mechanistic studies of EV function in vivo. The broad availability of cell-type-specific Gal4 driver lines in zebrafish and Drosophila will now allow researchers to trace EV dynamics from virtually any cell type, while our isolation protocol enables rigorous, quantitative EV analyses across developmental and pathological contexts.

The online version contains supplementary material available at 10.1186/s12964-025-02433-3.