Understanding spontaneous mitotic crossover by single-cell multi-omics

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT Homologous recombination (HR) is important for repairing DNA double strand breaks (DSB), and thus is an essential process in embryonic development, meiosis and suppressing tumorigenesis. HR can also be a doubleedged sword: unrepaired breaks lead to cell death; errors by HR, particularly the crossover type, can cause extensive genome rearrangements. While the biochemical process of HR in the repair of an induced DSB has been elucidated by various methods, two critical gaps remain regarding spontaneous HR: 1) the lesions driving spontaneous mitotic HR (e.g. in addition to DSBs, template switching in replication initiated by single-strand nicks can also promote HR); and 2) HR partner choice that determines whether HR is error-free or not. The two gaps are inter-related as HR partner choice could depend on the types of initiating lesions. What determines whether HR is error-free or not is a fundamental question of what governs genome integrity. A maior technical roadblock is the lack of scalable, genome-wide tools for studying rare spontaneous HR events in many mutants. To scale YJ2. genome-wide HR mapping efforts, I developed sci-L3, which enables linear amplification of single-cell genomes that scales to 1 M cells, enabling generating hundreds of single-cell global HR maps per mutant for thousands of mutants. Moreover, genetic assays require detecting "scars" in the genome as traces of repair (e.g. in cancer mutational signature studies), which miss 99% of error-free HR between identical sister chromatids. There is thus a critical need for unbiased global assays that detect both mutational and error-free HR. I recently addressed this gap by developing sci-L3-Strand-seq as the first scalable mapping tool for HR between identical sister chromatids. Our central vision is to determine how spontaneous mitotic crossovers cause genome rearrangements by scalable single-cell assays. In Area1, we will use sci-L3 to explore the full mutant space in hybrid yeast diploids. By generating HR maps in all the single mutants in a pooled manner (160 single-cell HR maps/mutant for 6,000 mutants), we can simultaneously test and generate thousands of hypotheses regarding different lesions and pathways that drive different types of genome instability events genome-wide. In Area2, we focus on a deciding factor for whether HR is error-free or not: HR partner choice of allelic sister chromatid, allelic homolog and non-allelic repeats. With sci-L3-Strand-seq, we propose to map all the seven classes of crossover outcomes in two systems: mammalian cell lines and mouse embryos. In cell lines, we will investigate genomewide distributions of both error-free and mutational HR outcomes in the wild-type as well as hundreds of perturbations of HR-related genes to determine factors affecting HR partner choice including (epi)genomic contexts, 3D genome organization and HR gene knockdown. We will also develop in vivo sci-L3-Strand-seq/RNA co-assay to dissect cell-type variation in HR partner choice in mouse embryos. In sum, this proposal pursues cost-effective, massively parallel and genome-wide mapping of mitotic crossover outcomes as functions of genetic perturbations and cell types by developing and applying advanced single-cell multi-omics tools.