Back

Single-cell transcriptomics uncover conserved molecular mechanisms and functional diversification in multilayered epithelia

Merle, C.; Huyghe, M.; Kimber, E.; Faraldo, M. M.; PANTALACCI, S.; Semon, M.; Fre, S.; Journot, R. P.

2025-11-14 cell biology
10.1101/2025.11.14.688286 bioRxiv
Show abstract

Multilayered epithelia, including skin, cervix, thymus, and prostate, arise from all three germ layers and perform diverse physiological functions, yet share a conserved layered architecture in which a supportive basal layer underlies differentiated suprabasal cells. To determine whether this common organization reflects a shared genetic program or tissue-specific regulation, we assembled a comprehensive single-cell atlas of 14 murine Multilayered epithelia. We identified a conserved p63-Notch axis that universally governs basal cell identity maintenance and suprabasal commitment. The p63-centered transcriptional program is shared, while Notch signaling recruits context-specific transcriptional modules to generate functional diversity. The layered architecture of epithelia spatially and temporally decouples conserved basal from context-specific suprabasal networks, compartmentalizing functional innovation across multilayered epithelia. Comparative genomics traces the p63-Notch axis to the root of vertebrates over 500 million years ago, showing that it was repeatedly co-opted through the incorporation of newly evolved genes in the suprabasal domain to generate novel epithelial functions. Our work establishes that evolution of multilayered epithelia operates through modular, compartmentalized diversification of a conserved scaffold, providing a unifying principle for how complex tissues achieve functional innovation while preserving structural integrity.

Matching journals

The top 4 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.