High-resolution spatial mapping of cell state and lineage dynamics in vivo with PEtracer

Abstract

INTRODUCTION: Organismal lineage maps, such as that of Caenorhabditis elegans, have provided critical insights into the dynamic interplay of cell state, ontogeny, and spatial cues that coordinate cell fate determination. However, creating analogous, comprehensive maps in mammals has remained out of reach. The ability to jointly track cell state and lineage with high spatiotemporal resolution would offer powerful insights into complex biological processes in development and disease. RATIONALE: We sought to develop an engineered lineage tracing system that enables joint cell state and lineage profiling in vivo across relevant spatial and temporal scales—from single cells to whole tissues, and ranging from a single cell division to the life of an organism. Imaging-based cell profiling methods, such as multiplexed error-robust fluorescence in situ hybridization (MERFISH), are well suited for multimodal cell phenotyping at tissue scale with subcellular resolution. Thus, the design and engineering of our lineage tracing system were guided by compatibility with imaging readouts. Specifically, using FISH-based imaging to assay lineage information requires a predefined set of lineage marks that are distinguishable by hybridization probes. We selected prime editing as the basis for an evolving lineage recorder capable of continuously installing immutable, predefined marks with sufficient diversity to enable robust lineage reconstruction. RESULTS: Guided by extensive theoretical modeling and experimental considerations, we built and optimized a prime editing (PE)–based lineage tracing system (PEtracer), which installs one of eight five-nucleotide marks at dozens of genomically integrated, barcoded lineage tracing cassettes. We tuned editing kinetics to record lineage information over diverse time frames ranging from days to months. Using cells containing predefined lineage information, we confirmed high readout efficiency and accuracy by both sequencing and imaging. Furthermore, rigorous benchmarking experiments demonstrated near-perfect phylogenetic reconstruction accuracy, establishing a strong foundation for in vivo studies. We then applied PEtracer to a transplantable, syngeneic mouse model of breast cancer lung metastasis to study the dynamics of cancer cell extravasation, colonization, and outgrowth. By combining high-resolution, MERFISH-based spatial transcriptomic profiling with rich phylogenies for tens of thousands of cells in clonally seeded tumors, we provide an unprecedented view of cancer growth in space and time as well as the organization of immune and stromal cells within these tumors. Joint analysis of cell state, lineage, and spatial microenvironmental factors across multiple sections of sampled tumors revealed malignant cell modules that link cell fitness and plasticity to both cell-intrinsic and -extrinsic drivers. Specifically, we identified a niche adjacent to the normal lung associated with a heritable, high-fitness, and epithelial-like state. CONCLUSION: PEtracer is an evolving lineage tracing system for high-resolution cell state and lineage profiling across space and time in vivo. Our system integrates cutting-edge approaches in genome engineering, lineage tracing, and advanced imaging methods and can be leveraged to study models of cancer initiation, progression, metastasis, and drug resistance as well as fundamental questions in development, stem cell biology, and homeostasis. Moreover, PEtracer is positioned to generate spatially integrated multimodal datasets to train burgeoning artificial intelligence foundation models of dynamic cellular behaviors in health and disease.