Single-Cell, Spatial and Functional Dissection of Cancer Cell States, Co-Evolving Ecosystems, and Vulnerabilities During Tumor Progression and Metastasis

Aggressive cancers often lack pharmacologically actionable mutations and do not respond to immune checkpoint blockade, thus deriving only modest clinical benefit from targeted and immune therapy. The heterogeneity of both transformed and healthy cells in the Tumor Microenvironment (TME) represents a critical obstacle to achieving more durable response in cancer patients. Recent insights, using multi-omics approaches, have shown that cancer cells can exist in a variety of transcriptionally distinct, yet co-existing states, some of which are already primed for metastatic progression or drug resistance. The plasticity of these states—i.e., the ability of cancer cells to reprogram across multiple states, either spontaneously or because of drug perturbations—and their homeostatic coexistence with other TME subpopulation, via paracrine molecular interactions, creates a constant challenge to therapeutic approaches by fostering the emergence of drug-resistance, tumor progression, and the creation of a pro-malignant, immunosuppressive milieu. Malignant states and transitions are only partially explained by sequential acquisition of somatic mutations, suggesting that they result from integration of a variety of cell-intrinsic and -extrinsic molecular cues that determine their lineage attribution, establishment, and interconversion. To date, several technical, clinical, and analytical challenges have hampered a comprehensive understanding of the natural biology of these processes in patients. Project 2 is dedicated to resolving the variability and plasticity of malignant cells and of the healthy cells that define the TME by developing and applying a battery of technical and analytical tools for the dissection of cancer heterogeneity at the single-cell level, and for the nomination, validation and testing of novel drivers of tumor-progression and therapy response and resistance. We will delineate these concepts in a defined biological context, that is the progression from a primary tumor towards brain-metastatic disease. To this end, we will leverage a series of innovations from CaST investigators, including (a) multi-modal single-cell profiling from archival tissues, (b) simultaneous low-pass whole-genome sequencing (lpWGS) of the same cell pool, (c) integrated single-cell and spatial single-cell transcriptomics, (d) analytical approaches to integrate and model multi-modal single-cell data in space, time and context of interactions among cells, (e) tools to elucidate cell state stability and transitions, (f) combinations of genome-editing perturbations with single-cell read outs that can be linked to drug screens via gene expression profiling, and (g) network-based Master Regulator analyses to elucidate mechanistic determinants of transcriptional cell state. This will be extended by experimental innovations, that (h) accurately model tumor progression in vivo and recapitulate entire human ecosystems, (i) enable labeling of metastatic niches coupled with single-cell genomics, and (j) provide a platform to test pharmacological modulation of cancer cell intrinsic and tumor-microenvironmental features predicted from human studies and modeling. The presented innovative framework will be broadly application to other cancer contexts.