Programmable nucleic acid cytometry for unraveling heterogeneity in tumors and therapies

SUMMARY Over the two decades, antibody-based approaches for isolating cell populations have been a fixture of molecular biology, including upstream of single-cell and bulk genomics, where specific populations are isolated and profiled via sequencing. Despite the widespread use of antibody-based methods to sort cells, notable limitations exist. First, many populations in relevant cancer settings can be discriminated only by genomic mutations, transcript expression, or intracellular protein abundance. Second, limited, if any, nuclear proteins known to discriminate populations of nuclei, which limits options following dissociation from tissues and tumors. Finally, even if surface proteins define a population, high-quality antibodies may not be available. Alternatively, nucleic acid cytometry, defined by fluorescence in cells and nuclei with specific RNA or DNA, addresses these limitations of protein-based methods. Nucleic acid cytometry i) targets a more diverse set of cellular molecules; ii) enables compatibility with nuclei; and iii) is robust for targets with robustness to off- targets without burdensome antibody discovery. The rational, programmable framework defining nucleic acid cytometry is an attractive method for isolating arbitrary cell populations in tissues and tumors but requires further development for wide adoption and routine use. Here, we propose to continue our advanced development and validation of nucleic acid cytometry for profiling rare cell populations from tumors and tissues. We will build on our recent work describing the Programmable Enrichment via RNA FlowFISH by sequencing (PERFF-seq), which couples Fluorescence In Situ Hybridization (FISH) to high-throughput single-cell RNA sequencing (scRNA-seq). We have demonstrated that PERFF-seq yields high-quality single-cell RNA-seq data for rare cells defined by one or more expressed markers, including from populations isolated from frozen and formalin-fixed paraffin-embedded (FFPE) tissue samples. Though our work suggested applicability to unravel heterogeneity in cancer-associated populations, additional work is needed for broader adoption and validation across heterogeneous nucleic acid targets. First, we will expand the scope of targetable transcripts that will resolve cell populations to include non- human transcripts, including oncogenic viruses and synthetic DNA used in cell therapies. Second, we will develop a novel probe co-hybridization strategy that may reduce the required input number of cells into the assay by nearly an order of magnitude to unlock routine profiling of FFPE tissue. Finally, we will develop a new approach to identify and sort cells based on one or more somatic mutations, which may be critical in profiling early oncogenic events and/or minimal residual disease. In sum, this proposal will build upon recent developments enabling nucleic acid cytometry and broaden its utility to reveal underlying cellular heterogeneity of rare populations in tumor environments and therapies.