Browsing by Subject "long interspersed nuclear element"

Long interspersed nuclear element 1 (LINE-1 or L1) belongs to a class of retrotransposons. In other words, it is a DNA element that can copy and paste itself around the genome. There are approximately 500,000 copies present in humans, but only around 5,000 are expected to remain transcriptionally competent. The activity of L1s is generally strongly repressed in normal human tissues, but in many cancers, these elements are reactivated. Both L1 transposition and transcription can have significant effects on cellular function, making it an interesting topic of research from a pathological point-of-view. By studying and understanding more about this transposon, it could be possible to find novel screening methods or even therapeutics for different cancers. One of these cancer types is high-grade serous ovarian carcinoma (HGSOG), which is known for exhibiting L1 upregulation. However, the quantification of L1 transcription has been proven to be very challenging, mostly due to alignment issues caused by the repetitive nature of the element. In addition, a large proportion of L1s reside within genes, meaning that L1 sequence -containing transcripts frequently do not originate from the L1’s own promoter. This thesis aimed to tackle these challenges; I quantified L1 expression at the single-locus level in 11 pre- and post-chemo HGSOC sample pairs, as well as in 5 samples from healthy women, based on single-cell RNA-sequencing. In addition to comparing L1 activity in different sample and cell types, I researched whether L1 activity was associated with any changes in gene expression. The poly(A) site of an L1 is relatively weak, meaning that L1 transcription frequently extends over it. Based on this fact, the utilized approach was to quantify L1 expression based on reads mapping to the 1 kilobase downstream window of each L1 locus, thus minimizing the alignment issues of repetitive elements. Thereafter, the features of the detected loci were carefully assessed to separate false-positive L1s from those with evidence supporting genuine activity, such as tumor sample enriched expression, lack of correlation to host gene, and detection with bulk RNA-sequencing. The activity of the latter loci was then further analyzed to search for differences in L1 expression between pre- and post-chemo samples. In addition, the association between L1-activity and gene expression was examined based on regression models both at the individual gene and molecular signature gene set-level. It was found that L1 expression data is filled with factitiously active loci, highlighting the importance of careful analysis and wet lab validations when studying transposon activity. However, regardless of the issues arising from a sparse and unreliable dataset, I showed that L1 activity was negatively associated with the expression of MYC target genes. MYC has been previously shown to be a transcriptional repressor of the L1, indicating that the obtained results are legitimate. Even though the results obtained from this study appear to be biologically justifiable, they would require further validation to ensure their authenticity. In addition, for the future it would be essential to enhance the sensitivity of the utilized workflow to minimize the sparsity of the data, so that statistical analyses performed would become more reliable. Nevertheless, it was shown that assessing L1 expression at the single-cell level using RNA-sequencing is executable.