Abstract 1360: Understanding genetic variation in cancer using targeted nanopore long read sequencing

Abstract Structural variations (SV), a hallmark of genomic instability in cancer can either activate oncogenes or inactivate tumor suppressor genes. SVs tend to be recurrent and have been associated with several cancer types. They are pervasive in nature however, next-generation sequencing (NGS) is mostly blind to SVs. They lack sensitivity and exhibit very high false positive rates (up to 89%) in SV detection. Long-read, single-molecule sequencing platforms like Pacific Biosciences and Oxford Nanopore Technologies (ONT) can address larger variations as they typically generate read lengths of tens of thousands of bases and have helped identify thousands of genomic features pertinent to cancer that were previously missed by short-read sequencing. However, the throughput and coverage offered by whole genome long read sequencing makes it infeasible to conduct large-scale genomic studies. Targeted sequencing significantly improves accuracy and coverage by offering the depth necessary to detect rare alleles in a heterogenous population of cells. It will facilitate large population SV analysis, which will help define the landscape of such variants in the population and help identify regions of therapeutic or diagnostic interest. However, a lack of efficient long-read compatible targeting techniques makes it difficult to study specific regions of interest on existing long-read platforms. To address this, we are evaluating CRISPR/Cas-based systems of targeted long-read sequencing to enrich for specific regions of the cancer genome (panel of 12 genes, including BRCA1 and BRCA2) in two breast cell lines – MCF 10A and SK-BR-3. Using the ONT Cas9-mediated PCR-free enrichment system we targeted 12 genes of interest in two breast cell lines – MCF 10A and SK-BR-3. Initially, we used guides that targeted a 200kb region around the BRCA1 gene in SK-BR-3 and generated a 198kb read spanning the entire BRCA1 region. However, the number of reads on target were not sufficient to accurately call SVs. To address the issue of low on-target reads, we included an Affinity-based Cas-9-Mediated Enrichment (ACME) step, that uses a HisTag-based isolation and pulldown of background/non-target reads. By targeting all 11 genes together in a single reaction, our ACME worked extremely well, since each reaction now had over 10 times as many guides as opposed to targeting a single gene per reaction. This gave us more cut sites and Cas9-bound non-targets, increasing the enrichment efficiency of the HisTag pulldown, giving us a 75-fold enrichment of the BRCA2 region and close to 100x coverage of the entire 95kb target. We observed an increase in enrichment and coverage of the other genes on the panel as well, with enrichment as high as 4000-fold for some genes. Our goal is to develop a long-read breast cancer panel to facilitate large population SV analysis, which will help define the landscape of such variants in the population and identify regions of therapeutic or diagnostic interest. Citation Format: Shruti V. Iyer, Sara Goodwin, Melissa Kramer, W. Richard McCombie. Understanding genetic variation in cancer using targeted nanopore long read sequencing [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1360.