Abstract:
Retrotransposons are mobile genetic elements, which replicate via a "copy-and-paste" mechanism, thereby creating a new copy of their DNA in the genome. LINE-1 (L1) is the most abundant retrotransposon in the human genome, directly accounting for 17% of the genomic DNA. L1 has shaped the human genome in many ways during evolution and is still active nowadays. It is considered as the major source of human interindividual genetic variation, and has been implicated in several human diseases as well. While data on the impact and significance of the L1 element has been accumulating in recent years, the understanding of the underlying molecular mechanism of retrotransposition has been lagging behind.
L1 encodes two proteins called L1ORF1p and L1ORF2p. Both proteins are essential for retrotransposition. L1ORF2p contains a nicking endonuclease and a reverse transcriptase domain. The role of the L1ORF1p in retrotransposition was much more elusive in the beginning of this thesis.To understand the molecular mechanics and phylogeny of L1ORF1p, we decided to obtain a high resolution structure of the protein.
First, we used bioinformatics to identify a non-canonical RRM (RNA recognition motif) domain within human L1ORF1p, as well as in the ORF1p proteins from many phylogenetically unrelated non-LTR retrotransposons. In addition to the central RRM domain of the L1ORF1p, we experimentally determined the domain boundaries of the coiled coil and of a C-terminal domain (CTD), and we showed that coiled coil is necessary and sufficient for a trimerization of human L1ORF1p.
Then, we used X-ray crystallography and NMR to determine the structures of individual domains, as well as of the trimeric L1ORF1p. From these studies L1ORF1p emerges as a highly sophisticated RNA binding protein that shows an unprecedented flexibility in the arrangement of its individual domains. These structures together with in vitro and in vivo experiments also suggest how single-stranded nucleic acids are bound by the trimer.
Phylogenetic analyses of a mammalian L1ORF1p suggest an ancient origin of the RRM domain and support the modular evolution of non-LTR retrotransposons. Finally, the trimeric structure of L1ORF1p is interesting, when compared to viral fibres and membrane fusion proteins, which form similar trimers. This provides a new twist in the discussion about the origin of non-LTR retrotransposons.