CAREER: Discovering the Spatial Modes of Tissue Perturbation with Volumetric DNA Microscopy

Living tissues build and repair themselves through the coordinated actions of millions of individual cells, but the rules connecting a cell's history, its gene activity, and its location in the tissue remain hard to observe. These rules matter well beyond basic biology: the same organization that guides healthy development breaks down in cancer and must be deliberately rebuilt in regenerative medicine. Today's tools force a hard choice. They can survey which genes are active across many cells, or preserve a tissue's intact three-dimensional structure, but they rarely capture a cell's ancestry, its molecular state, and its physical neighborhood together in a single whole specimen. This CAREER project takes a different route: instead of using a microscope, it programs DNA inside intact tissue to encode how the tissue's molecules are arranged in space, so that an ordinary DNA sequencing machine can reconstruct that organization in three dimensions. Paired with new computational tools, the approach reveals how instructions inherited inside cells and signals from their surroundings together shape a tissue's form. The work strengthens national biotechnology for studying complex tissues, delivers open software and a public three-dimensional data resource for the research community, and trains students across molecular engineering, genomics, imaging, computation, and biology to turn large biological measurements into testable models of health and disease. The project's central objective is to determine when a cell's lineage -- its place in the tissue's family tree of cell divisions -- acts as a predictive blueprint for the tissue's three-dimensional shape, and when signals from neighboring cells dominate instead. The research first tracks the family history of cells in laboratory-grown model tissues, including neural-tube organoids grown from a single cell, while switching individual genes on and off with CRISPR gene editing, then measures how closely each cell's response matches its relatives'. It next applies volumetric DNA microscopy, the DNA-programming method introduced above, to map lineage, gene activity, and three-dimensional position onto the same cells in intact specimens. Computational analysis compresses these responses into a small number of dominant spatial patterns and tests whether they explain organization across the whole tissue. Predicted patterns are checked against high-resolution imaging that counts individual RNA molecules, and detailed models -- from a simple baseline, to each cell family's own gene activity, to models that add the influence of neighbors -- are compared against the natural variation between repeated experiments. The outcome is a measurement-and-modeling framework that separates inherited cell programs from local neighborhood effects as a tissue takes shape, advancing NSF priorities in Biotechnology and Artificial Intelligence through analysis methods that generalize to other complex, network-structured biological data. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.