L1 is dynamic during neurogenesis

L1 retrotransposition is a significant source of endogenous mutagenesis in the human genome. This process is driven by a handful of highly active source L1s in each individual. L1 mobilization can occur in natural pluripotent cells, such as the stem cells present in the early embryo, and during the artificial generation of human induced pluripotent stem cells (hiPSCs) via cellular reprogramming. hiPSCs have been proposed for use in regenerative medicine, rendering an understanding of L1 retrotransposition and mutagenesis during hiPSC induction and cultivation essential.Although endogenous L1 insertions are known to arise in hiPSC lines [1], more information is required regarding L1 mobilisation and insertional patterns in order to elucidate the regulation and impact of L1 activity in this context. For example, transductions can be useful to find active, retrotransposition-competent L1s (RC-L1s) responsible for de novo retrotransposition events, and infer relationships between progenitor and offspring L1 copies. It has also been reported that endogenous retrotransposition in hiPSCs may generate an elevated fraction of full-length insertions [1, 2]. Crucially, L1 insertions bearing transductions have as yet not been identified in hiPSCs or embryonic stem cells (ESCs). Additionally, L1 promoters are methylated by the host genome to reduce their potential to initiate L1 mRNA transcription [3]. The genome-wide methylation state of L1 families, and specific L1 loci, has been analysed in ESCs and hiPSCs, and in neural stem cells (NSCs) [4-10]. However, the methylation state of individual L1 loci, including de novo L1 insertions and their progenitor elements, has to date not been reported.In the research described in this thesis, I identified a full-length de novo L1 insertion in a cultured hiPSC line. This L1 insertion was not present in the matched parental human dermal fibroblast (HDF) line and, as a result, was annotated as a reprogramming-associated event. Notably, the L1 carried transduced genomic sequences at its 5' and 3' termini. Using these unique transduced sequences, we identified the associated source (donor) L1 element, which had previously been reported to retrotranspose in vitro [11]. This donor L1 was part of an extended group of closely related L1s identified via their shared 3' transductions – an L1 transduction family [12]. Interestingly, the ancestral L1 copy, or lineage progenitor L1, of this family was previously reported as being inactive in vitro [13]. However, we found that the de novo L1 insertion, its immediate donor L1, and some other members of the transduction family, retrotransposed efficiently in vitro, indicating that these L1s comprise a lineage still capable of mobilisation, including during reprogramming.I next analyzed DNA methylation amongst the L1 transduction family prior to fibroblast reprogramming, in hiPSCs, and during neuronal differentiation. Notably, members of the family were demethylated during reprogramming and were then progressively methylated during neuronal differentiation. The de novo L1 insertion was hypomethylated compared to its donor L1, and the donor L1 was in turn less methylated than the older family members.This thesis therefore identifies an extended L1 transduction family that is mobile during reprogramming, likely as a result of specific relaxation of DNA methylation via an unknown mechanism. Using an in vitro system, my experimental data allows us to predict dynamic L1 methylation patterns during neurodifferentiation in vivo. Finally, my work further highlights L1 mutagenesis as a potential obstacle to the use of hiPSCs in biomedical applications.