Somatic mobile element detection in pediatric cancer

Abstract

Mobile elements (MEs) form a large part of our repetitive genome, but they distinguish themselves by their ability to move from one place to another. In humans the class I elements, LINE1, Alu and SVA, are still capable of moving via reverse transcriptase activity, also called ‘retrotransposition’. However, only LINE1 elements are autonomous, as only a small subset of LINE1 express the required proteins which facilitate retrotransposition of all three types of ME. Normally, retrotransposition is repressed, but can be lifted during embryonal development or in cancer. The resulting ME insertions (MEIs) can affect gene function or cause structural variants (SVs). Research into somatically occurring MEIs and tools to detect them are limited: Whilst MEIs has been observed to occur somatically in adult cancer, no extensive investigation has been performed in pediatric cancer yet. Therefore, the goal of our study was to assess the frequency of somatic MEIs in our pediatric solid tumor cohort and the reliability of tools to detect them. To this end, we first benchmarked a set of tools using simulated somatic MEIs before running the best combination of tools on a cohort of paired WGS pediatric cancer patients. Somatic MEIs were seen to occur rarely in a low number of patients in a randomized pattern over the tumor genome. Nevertheless, we observed some insertions that caused translocations or affected cancer genes. Patients with osteosarcoma were particularly affected by MEIs compared to patients with one of the other tumor types. Therefore, while most insertions are passenger mutations, some might have the potential to drive tumor development. We thus show that somatic MEI are difficult to detect but are active in pediatric cancers at a low rate.