Dynamics of a natural transposable element invasion

Transposable elements (TE) are parasitic DNA sequences in genomes. They spread solely to their own benefit even if this spread generates disadvantages to the host, like harmful mutations. In the egoistic quest for survival, TEs have even been known to break the species boundary where they were transmitted to a novel, previously uninfected, host species probably due to parasites like mites. Upon arrival in a previously uninfected population TEs may multiply rapidly such that the genome size of the invaded species may double within a few generations (e.g. 50). As this would lead to the extinction of invaded populations the invasion of a TE has to be counteracted but it is not clear how this is achieved. We recently captured a population of the fly Drosophila simulans at an early stage of a P-element, possibly the most widely studied TE, invasion. We propose to exploit this unprecedented opportunity, and to study a natural TE invasion in experimentally evolving D. simulans populations. This will allow to identify the extent to which multiple mechanism, like negative selection against TEs or small RNAs, contribute to counteract the P-element invasion. We will also test whether the environment influences a TE invasion, where for example populations in hot regions may be more readily invaded than populations in cold regions. Finally we want to develop a model that captures the dynamics of the TE invasion, which may at some point in the future allow to predict, and maybe even to control, the progression of a TE invasion.

In the project "Dynamics of natural transposable element invasions" we investigated the dynamics of transposable elements (TEs) from multiple diverse angles. This resulted in important contributions in several areas. First, we developed a novel software, DeviaTE, for analyzing the composition of transposable elements in diverse species. This tool provides an intuitive graphical overview of the abundance and composition of TEs in model and non-model organisms. It will thus be useful to a wide range of researchers. To gain a better and if possible also a quantitative understanding of TE invasions we performed computer simulations with our novel tool Invade. These simulations showed that piRNA clusters, i.e. repeat-rich regions controlling the activity of TEs, are crucial genomic regions, preventing the extinction of species from an accumulation of the deleterious effects of TE insertions over a wide range of parameters. The simulations also showed that clusters ought to evolve fast and harbor plenty of polymorphic TE insertions. To test these hypotheses it is necessary to compare the composition of piRNA clusters within and between species. However, piRNA clusters are repeat-rich and notoriously difficult to assemble. Therefore we first established several novel quality metrics and strategies to assemble these regions using long-read data. Based on these guides we assembled and analyzed clusters from several species. As a further complication it is extremely difficult to perform multiple alignments of highly repetitive regions, and thus to analyze the evolution of repeat-rich regions. Therefore, we developed a novel tool, Manna, which performs multiple alignments of repeat annotations. This tool finally opens up repeat-rich regions for evolutionary analysis. As predicted by our simulations clusters have plenty of polymorphic TE insertions from many TE families. Using Manna, we also showed that the composition of piRNA clusters is extremely fast-evolving. Solely 8% of the cluster content is conserved among closely related species, which raises the important question of which forces drive this rapid evolution of piRNA clusters. In a different work, we sequenced the "living fossils" of Drosophila research, i.e. Drosophila melanogaster lines collected over the course of the last century from different geographic locations. This revealed that within a mere 50 years 4 different TE families invaded worldwide Drosophila populations: Tirant invaded first around 1938, followed by the I-element, hobo and finally, the P-element. This is a remarkably high density of TE invasions within a short period of time. Lastly, we developed a novel strategy that allows tracing the geographic spread of TEs like the P-element in natural populations using extant population samples.