Gene Fusions & 3rd-Generation Sequencing
A Brief Overview

A common feature in many different cancer types is the presence of gene fusions that are created through incorrect repair of DNA damage. When two chromosomes suffer double strand breaks, the ends of the chromosome may be incorrectly rejoined, resulting in the exons of one gene being connected to those of another unrelated gene. Despite its accidental origin, the resulting fusion gene can still be transcribed, spliced, and translated into a protein sequence. While many of the resulting gene fusions will simply result in the generation of non-functional proteins, in some other cases the fusions may actually create a new oncogenic protein. Despite the rarity of these types of events, because they provide a growth/survival advantage over normal cells, they can rapidly generate a tumor which may then acquire additional secondary mutations that provide additional selective advantages.

Given their importance in cancer, there has been a long-standing clinical interest in being able to identify such genetic drivers in tumors, in order to stratify patients with regard to risk and to guide treatment. The growth of NGS technologies has powered cancer genomics for the last decade, and yet despite its impressive power to precisely measure gene expression and identify single base pair mutations, gene fusion detection remains a technical challenge. This is because the strongest evidence for gene fusions comes from sequencing reads that overlap the splice junctions of a transcript, and because these represent only a fraction of all the reads, statistically significant numbers of reads can be difficult to obtain. Other approaches for de novo assembly may also be used, but these are also technically challenging and computationally intensive. Interestingly, in the last few years, a 3rd generation DNA sequencing technology from Oxford Nanopore Technologies (ONT) has been attempting to overcome this gap in clinical capabilities.

While standard NGS approaches rely on the use of fluorescent dyes incorporated during DNA synthesis to generate sequence information, ONT makes use of electrical changes at individual pores in a membrane. As DNA strands continuously pass through the pores, the bases within the pore cause voltage changes that can be modelled and converted back into sequence information in real time. The advantages of this approach include the ability to use RNA directly without having to reverse transcribe cDNA, along with the ability to detect modifications to the individual nucleotides. The other critical advantage for addressing the problem of fusion gene identification is that sequence read lengths using ONT can be >1Mbp in size. This means that an entire gene fusion transcript can be sequencing as a single read, removing the need to identify the rare reads that cover only the fusion point.

Given the capabilities of the ONT technology, its rapid application to clinical oncology is unsurprising. For instance, work at the Massachusetts General Hospital Department of Pathology in 2018 demonstrated that within 10 minutes of the start of sequencing, the BCR-ABL gene fusion could be detected using a leukemic cell line and tests on FFPE preserved patient samples also rapidly and accurately detected gene fusions. Since this point the ONT long-read technology has been used to detect gene fusions in a wide range of cancers and has also been used to verify large scale structural variations associated with a number of diseases. It should be noted however that while powerful, the error rate based on decode voltage changes currently remains higher than other NGS technologies such as Illumina. Despite this limitation, the ONT sequencing platform continues to improve and has already staked a strong claim for resolving a critical problem for clinical detection of gene fusions.