Back

Distal-less homeobox 5 gives rise to myofibroblastic carcinoma-associated fibroblasts to promote collective breast cancer invasion

Mezawa, Y.; Kumegawa, K.; Morita, K.; Yang, L.; Hirakuri, K.; Yamashita, K.; Shirakihara, T.; Sasaki, R.; Onagi, H.; Kutomi, G.; Maruyama, R.; Orimo, A.

2026-08-31 cancer biology
10.64898/2026.08.30.745418 bioRxiv
Show abstract

Tumor-promoting myofibroblastic carcinoma-associated fibroblasts (myCAFs) are induced by activation of transforming growth factor-{beta} signaling. However, the molecular basis of myCAF-specific transcriptional programs regulated by TGF-{beta} signaling remains poorly understood. Using a meta-analysis of single-cell RNA-seq data from 132 human breast tumor and non-tumor tissues, we show that myCAFs activate gene regulatory programs relevant to skeletal and cardiovascular development that are associated with poorer outcomes in breast cancer patients. Of note, distal-less homeobox 5 (DLX5), a master transcription factor for skeletal development, is activated in human breast myCAFs at both epigenetic and transcriptional levels. DLX5 expression is also initiated by TGF-{beta}1 treatment in human mammary fibroblasts. Immunoprecipitation and CUT&RUN assays using DLX5-expressing fibroblasts demonstrate that DLX5 interacts with Smad2/3/4 proteins, enabling their cooperative occupancy at shared genomic binding sites of target genes, thereby promoting canonical TGF-{beta} signaling and the myCAF state. DLX5-primed myCAFs also enhance paracrine TGF-{beta} signaling and neuropilin-2 expression to promote collective tumor invasion. Our findings indicate that DLX5 induces myCAF formation and promotes breast tumor progression in collaboration with canonical TGF-{beta} signaling.

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.