Epigenetic control of the mobility of a human retrotransposon

Resumen

More than half of the human genome is made of transposable element (TE) derived sequences, and active TEs continue to impact our genome. In humans, Long INterspersed Elements class 1 (LINE-1 or L1) retrotransposons are active and autonomous TEs that move in genomes using a copy-and-paste mechanism. Over evolution, LINE-1s have amplified to astonishing numbers in the human genome, comprising >17% of our genome (i.e., >500000 L1 copies per genome). Despite their abundance, only a small subset of 80-100 LINE-1s continue to generate interindividual variability in humans, and are known as Retrotransposition-Competent L1s (RC-L1s). Human RC-L1s are 6-kb in length and, from 5´to 3´, contain: a 5´UnTranslated Region (UTR, of ∼900bp) with internal promoter activity, two non-overlapping Open Reading Frames (ORFs, termed L1-ORF1p and L1-ORF2p), and end in a short 3´UTR containing a polyadenylation signal. RC-L1s are expressed and mobilize during early human embryogenesis and, at much lower rate, in germ cells. Indeed, L1 activity in the germline genome ensure their evolutionary success over time. Interestingly, LINE-1s are expressed and can mobilize in brain and cancer cells. Although L1 activity in the soma might not have any impact on L1 evolution, provocative hypotheses suggest that L1 activity in somatic cells might play a role on human biology. New LINE-1 insertions can impact the genome by a myriad of mechanisms, and therefore their activity during early embryogenesis can, sporadically, result in new genetic disorders in newborns. Thus, cells have evolved numerous mechanisms to regulate and restrict L1 mobilization. Several L1 regulation mechanisms, acting at the transcriptional and/or post-transcriptional levels have been characterized in the past. In fact, in this Thesis I studied how L1 retroelements are regulated, as we know relatively little about mechanisms and pathways that regulate L1 expression and retrotransposition. Several factors related to the metabolism of nucleic acids, such as TREX1, SAMHD1 and ADAR-1, have been shown to regulate LINE-1 retrotransposition. Remarkably, inactivating mutations in these genes have also been associated with the development of Aicardi Goutières Syndrome (AGS). AGS is a rare disorder characterized for a strong immune response to endogenous nucleic acids. However, inactivating mutations in any of the three subunits of RNASEH2 (A, B, or C) is the most common mutation in AGS patients. Prior to this Thesis, the potential regulatory role of human RNase H2 on L1 retrotransposons was unexplored, although, by analogy with other AGS genes, it was suggested that RNase H2 might inhibit L1 retrotransposition. Thus, in this Thesis I dissected the role of human RNase H2 on LINE-1 regulation. Taking advantage of novel genome editing strategies (CRISPR/Cas9), I generated several cell lines lacking RNase H2 activity; then, I used these cells to determine whether the mobilization of a panel of active TEs is affected in KnockOut (KO) RNase H2 cells. Using engineered TE mobilization assays and biochemical methods, I demonstrated that RNase H2 is paradoxically required for human L1 retrotransposition, although it is dispensable for other mammalian retrotransposons that encode their own RNase H domain/activity (Mus D LTR-retrotransposons) or for DNA-Transposons (which mobilize using a fundamentally different cut-and-paste mechanism). I also demonstrated that RNASEH2A inactivating mutations characterized in AGS patients would prevent L1 retrotransposition, further suggesting that not all AGS patients would be characterized for high L1 activity in their genomes. In the second part of my Thesis I also studied L1 regulation processes, using innovative and advanced proteomic analyses of L1-interactors. Previous research from our lab demonstrated that new L1 insertions in human Pluripotent Cells (PCs) are epigenetically silenced during/shortly after L1 retrotransposition, in a sequence independent manner and involving histone deacetylation of newly inserted L1s. Because further research demonstrated that L1-silencing is attenuated in isogenic Differentiated Cells (DCs), we propose that L1-silencing might be a novel L1 regulatory mechanism of PCs. Our working hypothesis suggests that this pluripotent specific mechanism might normally operate in pluripotent embryonic cells, a cellular niche where de novo L1 insertions accumulate in humans, reducing the overall load of retrotransposition accumulated in the heritable human genome over evolution. Despite this knowledge, we don't know how L1-silencing works at a mechanistic level, nor how de novo L1 insertions are recognized in PCs. Thus, I decided to use proteomic methods to dissect the mechanism of L1-silencing in PCs. Here, I characterized the interactome of endogenous LINE-1s using the embryonic carcinoma cell line PA-1 as a pluripotent model, comparing isogenic PCs and DCs. To do that, I affinity captured endogenous L1 retrotransposition intermediates (i.e., L1 RiboNucleoprotein Particles) naturally expressed in PCs and DCs. Subsequently, I analyzed the composition of the L1-ORF1p interactome using mass spectrometry. Remarkably, we found that the L1-interactome is highly dynamic during cellular differentiation, and we found that most are novel L1 interactors. Furthermore, by comparing the interactomes of pluripotent and differentiated cells, I generated a list of candidate genes that could participate in L1-silencing. Future research would analyze the involvement of these factors in L1-silencing, using PCs and DCs and lost and gain of function approaches, respectively. In summary, the results included in this Thesis have augmented our mechanistic knowledge of how L1 retrotransposition is regulated in pluripotent human cells, a cellular niche where L1 is expressed and retrotranspose. A deeper understanding of L1-regulatory mechanisms would ultimately reveal how L1s can drive human genome evolution and would allow to define their causative role in human disorders like AGS.