Back

The atypical cadherin CELSR2 regulates distinct epithelial-mesenchymal states and oxidative phosphorylation in triple-negative breast cancer

Koren, J.; Saidani, N.-I. Y.; El kaoutari, A.; Daulat, A.; Finetti, P.; Josselin, E.; Audebert, S.; Camoin, L.; Gauthier, R. A.; Castellano, R.; GERAUD CREMIEUX, A.; Lamballe, F.; Bertucci, F.; Maina, F.; Borg, J.-P.; Walton, A.

2026-09-17 cancer biology
10.64898/2026.09.15.751697 bioRxiv
Show abstract

Triple-negative breast cancer (TNBC) is characterized by marked intratumoral heterogeneity, which contributes to metastatic dissemination, therapeutic resistance, and poor clinical outcome. Epithelial-mesenchymal plasticity is a major contributor to this heterogeneity. Identifying the molecular regulators governing epithelial-mesenchymal plasticity is therefore essential to understanding how distinct tumor cell populations arise and fuel TNBC progression. Here, we identify the adhesion G protein-coupled receptor CELSR2 as an independent poor-prognosis marker and regulator of epithelial-mesenchymal plasticity in TNBC. Spatial transcriptomic analyses revealed segregation of CELSR2high tumor cells within the epithelial tumor core and CELSR2low tumor cells at the invasive front. In line with the latter observation, CELSR2 expression decreases during EMT and its loss promotes acquisition of a hybrid epithelial-mesenchymal state. CELSR2high tumor cells display a CREB-associated oxidative phosphorylation (OXPHOS) program and promote cell proliferation and primary tumor growth. Conversely, CELSR2low cells display features of a more migratory phenotype. Collectively, our findings identify CELSR2 as a previously unknown regulator of epithelial-mesenchymal plasticity and reveal that heterogeneous CELSR2 expression shapes spatially and functionally distinct tumor cell populations, thereby contributing to intratumoral heterogeneity in TNBC.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.