During the project period (August 2019 – July 2025) the team established a new laboratory at Institut Curie and built collaborations across Europe and the US.
For Aim 1, intravital multiphoton imaging and organotypic brain slice cultures were used to visualise GBM cells migrating along vessels. We quantified how chemoradiation increased the number of vessel‑co‑opting cells and demonstrated that GSCs preferentially localise to blood vessels after irradiation. This work provided three‑dimensional datasets of co‑option dynamics.
For Aim 2, transcriptomic and phosphoproteomic screens on irradiated GBM cells identified key pathways underlying vessel co‑option. A major discovery was a therapy‑induced cell state named VC‑Resist that lies between proneural and mesenchymal subtypes and is highly vessel co‑opting. VC‑Resist cells exhibit FGFR1–YAP1 activation, G2M arrest, senescence and stemness pathways, leading to resistance to DNA damage. This work culminated in a 2024 Nature Communications paper presenting VC‑Resist as a driver of chemoradiation resistance and tumour dissemination.
In parallel, we identified a proneural–mesenchymal hybrid (GB‑Hybrid) cell state through dual fluorescent reporter experiments, single‑cell RNA sequencing and nuclear proteomics. GB‑Hybrid cells co‑express proneural and mesenchymal markers, are highly proliferative and show elevated transcriptional activity. At the molecular level they depend on nuclear pore complexes and nucleocytoplasmic shuttling of transcription factors. This therapy‑induced state contributes to resistance to radiotherapy and chemotherapy by increasing tumour cell plasticity. The GB‑Hybrid findings, published in Neuro‑Oncology 2025, extend the project’s focus on resistance and plasticity beyond VC‑Resist.
For Aim 3, we evaluated α6‑integrin and Wnt signalling as co‑option drivers. A 2021 Cancers article showed that integrin α6 supports stemness in proneural GSCs and mediates radioresistance in mesenchymal GSCs; inhibiting α6 reduced stem‑like traits in both subtypes. In proof‑of‑concept studies we inhibited Wnt7a in orthotopic GBM models, finding that combination of anti‑co‑option treatment and temozolomide prolonged survival (Cancer Cell 2018). By 2023 our transcriptomic screens identified additional mediators, including FGFR1–YAP1 signalling, which were validated using patient‑derived models and animal studies.
As the project progressed, it became clear that Wnt7 was not the predominant driver of therapy resistance. We therefore shifted our focus from Wnt7 targeting to dissecting and exploiting the vulnerabilities of the VC‑Resist and GB‑Hybrid cell states. High‑throughput screens and functional assays produced a list of putative targets involved in FGFR1–YAP1 signalling, senescence pathways, nuclear pore complexes and transcription factor transport. These targets represent potential strategies to prevent transitions towards VC‑Resist or GB‑Hybrid and thereby sensitise GBM cells to therapy. The shift in emphasis demonstrates the project’s adaptability and commitment to addressing the most relevant resistance mechanisms.
Beyond experimental research, the project produced a comprehensive review on vessel co‑option in glioblastoma (Angiogenesis 2019), a protocol for in vivo compression and imaging of solid stress in mouse brain (Nature Protocols 2020), and an edited book Brain Tumors (Neuromethods 2020) that assembled methodologies for brain tumour research. Dissemination activities included numerous invited seminars and conference presentations across Europe, North America and Asia. A mid‑term report highlighted that results were presented at the International Nanotumor Intravital Imaging Webinar Series, the EACR Conference on Structural Microenvironment and the Serbian Association for Cancer Research meeting. During the final period, findings were showcased at the French Society of Angiogenesis Congress (June 2025) and the Brain Tumor Microenvironment Symposium (January 2025). Social media dissemination via institutional accounts and press releases amplified visibility, reaching over 20 000 accounts worldwide. The lab also produced graphical abstracts and videos to communicate discoveries to broader audiences.
