Topolink™: Advancing Cancer Genomics with a High-Resolution, High-Throughput Approach for Low Sequencing and Highly Sensitive Detection of Structural Variants

Abstract

Despite numerous technological advances, including the widespread adoption of massively parallel genome sequencing, many clinically relevant cancer driver mutations go undetected. Additionally, even with the most comprehensive cancer profiling using a combination of whole genome and whole transcriptome sequencing, a driver mutation goes undetected in approximately 20% of cancers, making targeted treatment of these patients challenging. The reason for this gap in understanding is presumably two-fold: 1) current technologies do not have the required sensitivity for detection of the causative variants; or 2) causative variants are epigenetic or regulatory in nature, meaning the driving alteration does not result in a change to the core DNA sequence. An explanation for 1) is that structural variants (SVs) are particularly challenging to detect using shotgun-based approaches, since these depends on the presence of specific chimeric molecules within the library that directly span or bridge the breakpoint. Long-read technologies circumvent this limitation but have significant drawbacks with respect to cost and strict sample requirements. Another possibility is that standard sequencing - which results in hundreds or thousands of variants of unknown significance (VUS) - may in fact contain causative variants, but our understanding of function of these variants is limited. For 2), it is increasingly appreciated that epigenetic and chromatin topological features are fundamental in the gene regulation and disease. To address this gap in our understanding of cancer, we utilized TopoLink™ proximity ligation library protocol that yields high-quality, high-resolution, unbiased HiC libraries and that can be performed in under 8 hours. To our knowledge, TopoLink™ is the first and only assay of its kind. Proximity ligation offers enhanced sensitivity for detection of SVs, and the restriction enzyme-free digestion method ensures the uniform coverage needed for accurate detection of single nucleotide variations (SNVs) and copy number variants (CNVs). To our knowledge, TopoLink™ is the first and only assay combine the speed, throughput, and unbiased primary base coverage of WGS with the improved detection of large SVs in Hi-C data. Detection of SVs provides a critical basis for personalized therapies in hematological cancers. To test the ability of TopoLink™ to detect clinically relevant SVs, we used the BCR-ABL1 positive CML cell line K562 to determine the limit of detection of interchromosomal translocations. In addition, we calculated the sensitivity of SV detection relative to current industry-leading sequencing technologies using the breast carcinoma cell line HCC1187 gold standard. We noted enhanced sensitivity for SV detection for TopoLink™ relative to both long-read technologies and WGS. Using TopoLink™, sensitivity was greater than 95% at a genome coverage of less than 1x, whereas both long-read and WGS sensitivity drops below 95% sensitivity at approximately 10x genome coverage. Similarly, using hybridization capture of BCR-ABL1 in a TopoLink™ library of K562 cells reduced the required sequencing burden by 10-fold relative to standard shotgun approaches. Importantly, we demonstrate that TopoLink™ libraries maintain topological features consistent with biologically relevant topologically-associated domains (TADs) and chromatin loops, thus enabling insight into novel epigenetic cancer drivers. Therefore, we demonstrate TopoLink™ proximity ligation libraries as a complementary technique that offers enhanced sensitivity of clinically relevant structural variants, with the added benefit of improving discovery of novel epigenetic mechanisms. Finally, the reduced sequencing costs needed to detect clinically relevant SVs allows for improved diagnostics and personalized medicine in a clinical or translational research setting.