Project: PRJEB53130
Metastasis is the leading cause of death in Breast Cancer (BC). However, despite tremendous research efforts, molecular determinants of the pro-metastatic phenotype are largely unknown. Here, we leveraged lentiviral barcoding coupled to single-cell RNA sequencing to trace clonal and transcriptional evolution of BC cells during metastatization in vivo. We showed that metastasis derive from rare pro-metastatic clones that are under-represented in primary tumors (PTs). Analysis of same clones in different PTs revealed that both low clonal-fitness and high metastatic-potential are independent of clonal origin. Differential expression analyses and classification approaches showed that pro-metastatic clones are distinguished by rare cells with concomitant over-expression of genes involved in extracellular matrix (ECM) remodeling, dsRNA and type-I interferon signaling, and stress response pathways (hypoxia, unfolded-protein and antioxidant responses). Gene-expression signatures from the identified pro-metastatic genes strongly predict disease progression in BC patients, independently of other known prognostic factors. Remarkably, most non-metastatic clones activated the same pathways, though at lower levels and not concomitantly within the same cell. Genetic silencing of single pro-metastatic genes from either of these pathways was sufficient to impair cell migration in vitro and prevent metastatization in vivo, with negligible effects on cell proliferation and PT growth in most cases. This study demonstrates that the pro-metastatic phenotype is acquired in rare PT cells through the concomitant hyper-activation of several stress response pathways, and provides novel molecular predictors of BC prognosis and therapeutic targets for metastasis prevention.
General