Novel PCR-based assay to detect and characterize LINE-1 retrotransposition in human cancers

Transposable elements (TEs) comprise almost half of the human genome because of their mobile activity during genome evolution. Although most families of TE are currently inactive in the human genome, one young TE family called Long Interspersed Nuclear Element (LINE)-1 is found to be active in germ cells, during embryonic development, and in diseases such as cancer, especially epithelial cancer types. LINE-1 is an RNA transposon (retrotransposon) which means it moves from one genomic location to another via an RNA intermediate by a process called retrotransposition. By following this replicative cycle, LINE-1 has already made more than 500,000 near-identical copies of itself and hence occupied 17% of the human genome. However, most of these copies have accumulated inactivating mutations and/or truncations and are retrotransposition-incompetent. The diploid genome in an individual is estimated to have only 80-100 potentially active LINE-1s. These potentially active LINE-1s are epigenetically silenced in normal cells. Widespread epigenetic modification during cancer alters this silencing mechanism and activates these LINE-1s which have the potential to generate mutagenic insertions in the genome. Detection of de novo LINE-1 insertions among more than 500,000 pre-existing LINE-1 copies remains challenging. To address this unmet need, we developed a novel PCR-based method that detects de novo insertions originating from one of the most active LINE-1 loci in the human genome (located within the TTC28 gene, hereafter referred to as TTC28 LINE-1). Compared to whole-genome sequencing (WGS), this approach is substantially more sensitive, identifying nearly three times more de novo TTC28 LINE-1 insertions in two colorectal cancer samples. Moreover, by taking advantage of long-read single-molecule sequencing, we were able to characterize de novo LINE-1 insertions in their entirety at nucleotide-level resolution. Using uterine leiomyoma as the disease model, we show that LDI-PCR is also suited for detecting DNA rearrangements in rearrangement-prone genomic regions with high sensitivity. Next, we traced the activity of TTC28 LINE-1 in a panel of ten high-grade serous ovarian carcinoma (HGSOC) cell lines, five of which were proficient for homologous recombination (HR), an error-free DNA double-strand break repair pathway, and the other five were HR-deficient. Although TTC28 LINE-1 mRNA was expressed in all HGSOC cell lines, HR-proficient HGSOC cell lines showed a higher frequency of de novo TTC28 LINE-1 insertions in comparison to HR-deficient cell lines. We speculate that the simultaneous loss of HR-mediated DNA repair and gain of LINE-1 activity (which generates DNA double-strand breaks as part of the insertional process) could be detrimental to HR-deficient cell lines. HR-proficient cell lines, in contrast, might tolerate LINE-1 activity and thus have accrued more de novo LINE-1 insertions. Altogether, this thesis provides a highly sensitive method to detect and characterize de novo LINE-1 retrotransposition events. By employing this approach, this thesis also demonstrates that LINE-1 retrotransposition is more frequent in HGSOC cell lines that are proficient in repairing DNA double-strand breaks by the HR pathway.