Extracellular vesicles carry information. We decode it.

We quantitatively characterize the dynamics of extracellular vesicle (sEV) trafficking in the tumor microenvironment by measuring both vesicle release and uptake across multiple cell types, including different tumor, and stromal cells. Using dual-color in-cell EV labeling combined with flow cytometry, we can simultaneously track sEV production and internalization without isolation, thus preserving physiologically relevant concentrations. This allows us to calculate the EV-dynamic profile, a robust metric expressing the balance between vesicle release and uptake for each cell type. By applying this method under baseline and stress conditions, we identify cell type–specific communication patterns and, importantly, can monitor shifts in vesicle exchange during treatments — capturing not only changes in release but the full dynamic balance of vesicle trafficking.

Extracellular vesicles carry a diverse molecular cargo that plays a pivotal role in immunomodulation, emerging as key regulators of immune responses during tumor progression. Tumor-derived EVs often harbor immunosuppressive factors that impair the antitumor activity of immune cells. In this study, we aim to investigate the immunological effects of blood-derived small EVs on B cells from different brain tumor patients. Since brain tumor-derived EVs primarily interact with microglia, we further examined their immunomodulatory properties in a microglia–B cell co-culture system, focusing on vesicle-driven alterations in cytokine and chemokine secretion, modulation of immune checkpoint proteins (PD-1, CTLA-4) and surface marker expression, as well as their influence on B cell proliferation and apoptosis.

To capture the dynamic interplay of extracellular vesicles (sEVs) within the tumor microenvironment, we employ complex 3D cell culture models that recapitulate key structural and functional features of in vivo tumor tissues. These models are operated under two distinct conditions: a static culture with steady medium, allowing controlled baseline observations, and a dynamic flow system (MIVO®) that better mimics physiological fluid circulation in the human body. The inclusion of dynamic flow enables more accurate modeling of vesicle trafficking and exchange between tumor, fibroblast, endothelial, and immune cells. By integrating advanced imaging with quantitative vesicle tracking, we map the multidirectional exchange of vesicles across different cell populations, revealing how sEV communication networks orchestrate tumor progression, immune modulation, and fibroblast remodeling. This dual-condition approach provides a versatile platform for dissecting cell–cell communication at high resolution and for evaluating therapeutic strategies aimed at disrupting pathogenic vesicle signaling pathways.

The vesicle research community, including our research group, posits that it may be more appropriate to consider the vesicle as the fundamental unit of information, rather than focusing on the molecular contents of vesicles through conventional omics approaches. In this context, Raman spectroscopy — a type of vibrational spectroscopy — offers a compelling alternative, as it enables the characterization of the complete molecular profile of a biological sample. Our group employs this molecular fingerprint to facilitate the diagnosis of various tumor types, including central nervous system tumors, through the analysis of blood-derived extracellular vesicles.

We are building a high-coverage EV-proteome library by deep DIA-LC-MS profiling of plasma EVs from healthy volunteers. Quantitative intensity tables are processed with Weighted Gene Co-expression Network Analysis (WGCNA) to translate thousands of peptides into a biologically interpretable surface protein atlas. Applying the same network to tumor-derived datasets highlights the concerted protein-expression changes that distinguish CNS-tumor EVs from healthy controls and nominates minimal-invasive biomarker panels. Combined with machine-learning classifiers, this atlas-driven workflow offers a reproducible, expandable pipeline for diagnostic discovery and for monitoring therapeutic response in future multicenter cohorts.

Our group aims to unravel how the blood-derived protein corona dictates the immunological fate of circulating extracellular vesicles. Using nascent endothelial sEVs as a model, we map the balance between “eat-me” (IgG / C1q) and “don’t-eat-me” (CD47 / factor H) corona blocks and link these signatures to functional read-outs—complement activation, macrophage uptake and PBMC cytokine release. This growing atlas will serve as a reference for forthcoming comparisons with glioblastoma plasma and already highlights surface modules that can be exploited as biomarkers or therapeutic targets to overcome tumour-EV immune evasion.