Single-cell ATAC profiling of mammary gland tumours in MMTV-PyMT mouse model of breast cancer

Disseminated cancer cells rarely succeed in forming macrometastases, yet they continue to be the most life-threatening cause of patient mortality. The determinants driving metastatic competence, however, are still poorly understood. Using data from human breast cancer patients and genetic mouse models, we uncover a non-linear relationship between the expression of the EMT transcription factor Prrx1 in the primary tumor and metastatic outcome. Multi-omics analyses in mouse tumors reveal that Prrx1 functions as a dual regulator, promoting invasive behavior while simultaneously inducing dormancy. Cells with intermediate Prrx1 levels exhibit the highest metastatic fitness, whereas extreme levels (negative or too high) lead to less metastasis, typical of a hormetic behavior, and respectively due to insufficient invasiveness or to dormancy. These findings provide mechanistic insight into the regulation of dormancy and its link to metastatic potential. Additionally, we propose that combined signatures of invasion and proliferation allow robust prognosis prediction in breast cancer patients. Our study further highlights Prrx1-driven heterogeneity, established in the primary tumor, as a key determinant of metastatic competence.