Identification of a shared persistence program in triple-negative breast cancer across treatments and patients

Acquisition of resistance to anti-cancer therapies is a multistep process, which initiates with the survival of drug persister cells. Understanding the mechanisms driving the emergence of persister cells remains challenging, primarily because of their limited accessibility in patients. Here, using mouse models to isolate persister cells from patient tumors, we determine the identity features of persister cells from eight patients with triple-negative breast cancer (TNBC). Combining over 80 transcriptome studies, we reveal hallmarks of the persister state across patient models and treatment modalities: high expression of basal keratins together with activation of a stress response and inflammation pathways. Patient-derived persister cells are transcriptionally plastic and return to a common treatment-naïve like state upon relapse, regardless of the treatment they have been exposed to. Leveraging gene regulatory networks, we identify AP-1, NFKB and IRF/STAT as the key drivers of this hallmark persister state. As a proof of concept, we show that FOSL1 - an AP-1 member - is sufficient to drive cells to the persister state by binding enhancers and reprogramming the transcriptome of cancer cells. On the contrary, cancer cells without FOSL1 have a decreased ability to reach the persister state. By defining hallmarks of drug persistence to multiple therapies of the standard of care, our study provides a resource to design novel combination therapeutic strategies to limit resistance.