MULTI-OMIC LINEAGE TRACING PREDICTS THE MOLECULAR DRIVERS OF CANCER STEM CELLS IN TRIPLENEGATIVE BREAST CANCER EVOLUTION.

Triple-Negative Breast Cancer (TNBC) is the most aggressive breast cancer subtype, and patients still rely primarily on chemo-neoadjuvant therapy (NACT). Their poor prognosis is largely driven by extensive phenotypic heterogeneity, sustained by complex genetic and epigenetic interactions that are difficult to dissect without perturbing the native cellular context in which multiple phenotypically distinct subpopulations coexist. Recent advances integrating single-cell omics with different clonal barcoding have begun to uncover features linked to the aggressive behavior of cancer clones. However, the intricate gene regulatory mechanisms that define their identity and govern their fate remain insufficiently resolved. In this work, we developed GALILEO, a single-cell multiomic framework that combines lineage tracing with transcriptomic and chromatin accessibility profiling (via snRNA-seq and snATAC-seq) and integrates these with phenotypic assays. Applying GALILEO to the intrinsically heterogeneous and highly tumorigenic TNBC cell line SUM159PT, we identified two distinct subpopulations: drug-tolerant cells (DTCs) and tumor-initiating cells (TICs). DTCs are marked by a genomic alteration on chromosome 11 that confers reduced sensitivity to cytotoxic treatment. Conversely, TICs exhibit two baseline transcriptional states and share a set of co-accessible chromatin regions, indicative of a pre-existing epigenetic basis for tumor initiation. Their behavior is orchestrated by specific gene regulatory networks driven by a set of transcription factors that regulate the epithelial-to-mesenchymal transition (EMT) program. To validate these findings, we performed prospective isolation using TM4SF1 as a TIC-specific surface marker. TM4SF1high cells were confirmed to be enriched for tumor-initiating capacity through in vivo orthotopic transplantation. We further generated high-resolution transcriptional and epigenetic profiles of TM4SF1high cells via bulk RNA-sequencing and CUT&Tag, enabling precise annotation of the active regulatory elements underlying the networks reconstructed with GALILEO. Overall, this study elucidates the coexisting genetic, epigenetic, and transcriptional determinants shaping TNBC evolution and reveals the molecular architecture of rare, pre-encoded tumor phenotypes.