Error-suppressed whole genome sequencing for genotoxicant-induced structural variant detection

点击翻译按钮获取中文摘要

Project Summary Structural variants (SVs) are an important class of human genetic alteration that creates large changes in genomic content in single mutational events. Copy number variants (CNVs) change the representation of thousands to millions of DNA base pairs, potentially including multiple genes, whereas copy number neutral inversions and translocations lead to gene fusions and dysregulation of epigenetic control mechanisms. SVs and CNVs are critical drivers of cancer, a primary mechanism of constitutional germline genetic disease, and are increasingly appreciated as an ongoing form of somatic mosaicism affecting the function of tissues. SVs are subject to the same classes of risk factors as single-nucleotide variants (SNVs). These factors, which lead to inter-individual variation in lifetime mutation burden and therefore disease risk, include inherited predispositions based on DNA repair capacity and environmental exposures to mutagenic genotoxicants. Genotoxicants that induce disruption or breakage of chromosomes are known as clastogens. Importantly, assays commonly used to identify clastogens or double-strand break (DSB)-inducing genotoxicants are non-specific and fail to reveal whether specific agents lead to new, stable SV junctions in cells and, if so, what the nature and distribution of those junctions is. There is a large need for assays that can efficiently reveal SV mutagens via definitive, positive readouts of breakpoint junctions, the molecular endpoint of interest to toxicologists. Such an assay would be highly impactful in the genome sciences as a way of surveying populations of cells for mosaic SV mutations, including in studies of hazard identification and evaluation, as well as in basic science studies of cancer evolution and heterogeneity, tissue somatic mosaicism, and DSB repair mechanisms. SNVs have further demonstrated the great value of error-minimized, high-throughput sequencing for mutation detection and genotoxicant characterization. Led by Duplex Sequencing, such approaches reveal baseline and induced SNV frequencies with exquisite sensitivity by maximizing signal-to-noise ratios. However, prior work on error-minimized sequencing has been largely limited to SNVs. Our data confirm that the approaches do not address the distinct error mechanisms that lead to false SV detections. Our rationale is that an error-minimized sequencing method designed to address the processes that give rise to SV artifacts will be as impactful for SVs as Duplex Sequencing has been for SNVs. This R21 project will devise, develop, and validate such a method and begin to apply it to important genetic toxicology paradigms, with the following aims: (1) Develop a robust svWGS protocol for error-suppressed, non-targeted SV junction sequencing; and (2) Validate svWGS using well-controlled, high-value case examples with translational potential. Our long-term objective is to apply svWGS in detailed characterizations of emerging candidate human clastogens.