A reporter system for screening CNV risk factors

University Of Michigan At Ann Arbor, Ann Arbor MI

Investigators

Abstract

Abstract Copy number variants (CNVs) are chromosomal deletions and duplications that play a major role in human genetic disease and cancer. Aberrant DNA replication and nonhomologous repair of DNA damage in mitotic cells are implicated in the formation of sporadic, non-recurrent CNVs, which account for most de novo, disease-associated CNVs in the germline and cancer. Despite their clear importance to human health and variation, very little is known about the genetic and environmental factors involved in CNV formation. Using a model human cell culture system coupled with unbiased, genome-wide analysis technologies, we have demonstrated that replication stress, such as that caused by aphidicolin, hydroxyurea, and low-dose ionizing radiation, creates de novo CNVs that span tens to thousands of kilobases in human and mouse cells, and found that these mutations arise independently of Xrcc4-mediated end joining. We recently completed a mammalian-cell study that synthesized our experimental observations into a predictive model of CNV formation under replication stress. Analysis of our large set of experimentally-induced CNVs established that exogenous replication stress induces CNV formation at both singleton and hotspot loci with no difference in size or breakpoint structures. These hotspots provide ideal targets for studies of genetic factors involved in their formation. The key determining factor in whether a genome region is a hotspot is the presence of a very large, actively transcribed gene. Deletion CNVs arise near the center of these transcribed and late replicons, with de novo CNVs characterized by microhomologous junctions placed randomly within these hotspots. This pattern mimics many non-recurrent CNVs in genomic disease, cancer, and non- cancerous somatic tissues, indicating that CNVs arising in late-replicating, large transcription units are important to human genomic instability, and model a broader class of disease-associated, non-recurrent CNVs. While our experimental approach is powerful, it is also very time- and resource-intensive, not lending itself easily to high-throughput analyses. In this R21 proposal, we address this technical challenge through design of a cell reporter assay, taking advantage of our predictive model of CNV formation to develop a rapid assay for CNV formation. The ability to predict where CNVs will frequently occur allows us to place a reporter within a hotspot, where CNV formation can be monitored as a proxy for genome-wide non-recurrent CNVs without the need for time-consuming clonal isolation and whole genome analyses. We will test a panel of cell cycle checkpoint, replication, and DNA repair genes for their role in CNV formation following replication stress, generating novel insight into the genetics of CNV mutagenesis. The proposed studies will generate a valuable experimental tool that can be used to rapidly assess the effects of large numbers of genes and environmental agents on CNV formation, allowing us to determine the underlying mechanisms involved in this very important class of mutation with broad significance to normal human variation, numerous genetic disorders, and cancer.