A team at Université Paris Cité has spent six years chasing an unusual observation: Tiny particles released by certain breast cancer cells seemed to make aggressive tumors behave less aggressively. Their new study, published in Molecular Therapy, identifies what's inside those particles that does the work, and offers a possible blueprint for building RNA-based cancer drugs around combinations of molecules rather than single targets.
The story started in 2020, when the group, led by molecular biologist Sébastien Jauliac, reported that extracellular vesicles (EVs) released by breast cancer cells expressing a transcription factor called NFAT3 could reduce invasion and metastasis in aggressive cancer models. That raised an obvious follow-up question: what was actually inside those vesicles doing the work?
Fifteen microRNAs, one signal
In the new paper, the researchers compared the RNA cargo of vesicles from NFAT3-expressing cells against vesicles from cells with NFAT3 knocked down. That comparison narrowed a long list of candidate microRNAs down to a specific set of 15, which the team named miR-Comb 15. The team noted that none of the miRNAs could replicate the comprehensive therapeutic impact of the full signature when administered in isolation. Together, they inhibited both invasion and proliferation across triple-negative breast cancer and pancreatic cancer models, cancers with notoriously few effective targeted therapies.
Jauliac told DDN the shift from single-target to combination thinking was central to the project. "Rather than targeting a single molecular pathway, combinatorial RNA strategies may enable simultaneous modulation of multiple cancer-driving networks," he said.
Solving the delivery problem
Vesicles pulled from tumor cells aren't a viable therapeutic starting point, since they risk carrying oncogenic material along with the useful cargo. To get around that, the team turned to vesicles derived from HEK293T cells, a standard, non-tumor cell line already widely used in lab research, and loaded them with miR-Comb 15 using a pH-gradient technique that pushes RNA into vesicles without damaging them.
The team tested that engineered EV platform against two alternatives: vesicles derived from human adipose-derived stem cells and synthetic liposomes. All three carried the miRNA combination, but the HEK293T-derived vesicles delivered the most consistent activity in cell models and were the only platform that reliably slowed tumor growth in mice.
Jauliac was careful to frame the work as foundational rather than clinic-ready. "I see this work less as a finished therapeutic product than as a design framework for multi-target RNA medicines inspired by natural EV biology," he said. Before any translation toward a therapy, the team flagged several open questions: how the vesicles distribute through the body, which cells actually take them up, how to manufacture them at scale with consistent quality, and long-term safety.
Why it matters for RNA drug design
For drug developers, the appeal isn't any single molecule but the underlying logic. Aggressive cancers are typically driven by several interacting genetic programs rather than one dominant mutation, which is part of why single-target drugs so often lose effectiveness.
Jauliac's team argues that identifying the smallest combination of RNAs capable of disrupting that broader network, rather than picking one target and optimizing around it, could be a more durable design principle for next-generation RNA therapeutics. The extracellular vesicle platform they built to deliver that combination is also engineered to be scalable, using a cell line and loading method already established in other RNA-delivery work, providing a practical experimental framework for further development toward a translatable delivery system.
