Open Access

Liliana Elisabeth Layer

• The human Long Interspersed Nuclear Element-1 (LINE-1, L1) is a member of the group of autonomous non-LTR retrotransposons found in almost every eukaryotic genome. L1 elements generate copies of themselves by reverse transcription of an RNA intermediate and integrate into the host genome by a process called Target Primed Reverse Transcription (TPRT). They are responsible for the generation of approximately 35% of the human genome, cover about 17% of the genome and represent the only group of active autonomous transposable elements in humans. L1 activity bears several risks for the integrity of the human genome, since the L1-encoded protein machinery generates DNA double-strand breaks (DSBs) and is capable of conducting numerous genome-destabilizing effects, e.g. causing deletions at insertion sites, disrupting or rearranging coding sequences and deregulating transcription of functional host genes. On the other side, L1 elements have had and still exert a great impact on human genome structure and evolution by increasing the genome size and rearranging and modulating gene expression. Furthermore, due to its endogenous and generally non-pathogenic nature, L1 is a promising candidate as vector for gene delivery in somatic gene therapy. The structure of the flanking regions between de novo L1 integrants and the genomic sequence suggests an involvement of cellular DSB repair pathways in L1 mobilization. To elucidate the role of DSB repair proteins in L1 retrotransposition, I disabled DSB repair factors ATM, ATR, DNA-PK, p53 and Ku70 by knock down (KD) using short hairpin RNA (shRNA) expression constructs. To inhibit the function of DSB repair factors PARP and Rad51, I used dominant negative (DN) PARP and Rad51 mutants. Applying a well established L1-retrotransposition reporter assay in HeLa cells, de novo retrotransposition events were launched in order to test L1 for its retrotransposition activity in the context of altered DSB repair conditions. I could show that L1 retrotransposition frequency after ATM KD had increased by 3-fold, while ATR and p53 KD reduced L1 retrotransposition by approximately one third. Unfortunately, the cytotoxic effects of the DNA-PK and Ku70 shRNA expression constructs were too strong to determine potential effects of DNA-PK and Ku70 KD on L1 retrotransposition. Inhibition of PARP function by expression of the DN mutant and overexpression of wild type PARP were found to increase L1 retrotransposition by 1.8 and 1.5, respectively, while Rad51 DN had no detectable effect. Interestingly, overexpression of wild type Rad51 seemed to roughly double L1 retrotransposition frequencies. Since in my experiments KD or expression of DN mutants was time-delayed to the onset of L1 retrotransposition after transfection into the cells, I developed a temporally controllable, tetracyclin transactivator (tTA)-dependent L1 retrotransposition reporter assay which will be of great value for future L1 retrotransposition studies that rely on temporally controllable retrotransposition. Due to a previously published hypothesis of L1 playing a role in brain development by contributing to somatic mosaicism in neuronal precursor cells, I generated a transgenic mouse (LORFUS) using the tTA-dependent L1 construct to further test this hypothesis. LORFUS harbors a bidirectional cassette driving simultaneous expression of a GFP-tagged L1 retrotransposition reporter and beta-galactosidase. It was bred to another transgenic mouse line expressing tTA in the forebrain. The double transgenic offspring was used to characterize L1 expression and retrotransposition patterns in the brain at postnatal day 15 (P15). General transgene expression indicated by beta-galactosidase activity was found in hippocampus, cortex and striatum, while retrotransposition events revealed by GFP expression were found in hippocampus, cortex, striatum, olfactory bulb and brainstem. These results suggested L1 retrotransposition in the granule layer of the dentate gyrus earlier than P15 and migration of cells carrying these events along the rostral migratory stream into the olfactory bulb. To facilitate the use of L1 as gene delivery tool in gene therapy or genetic engineering, I furthermore intended to manipulate the L1 target site recognition to allow the site-specific integration into defined genomic locations. To this end, I performed crystal structure-guided mutational analysis exchanging single amino acid residues within the endonuclease (EN) domain of L1 to identify residues influencing target site recognition. However, individual point mutations did not change the nicking pattern of L1-EN, but resulted in a reduction of endonucleolytic activity reflected by a reduced retrotransposition frequency. This suggests that additional factors other than the DNA nicking specificity of L1-EN contribute to the targeted integration of non-LTR retrotransposons in the host genomes.