One major objective of my ERC project is to understand how the expression of retroelements is regulated. I suspected that chromatin accessibility induced by environmental trigger, such as viral infections, is a major determining factor for retroelement de-repression.
We set up Oxford Nanopore long read transcriptomics in our lab as proposed in WP1. Here, we implemented different library prep protocols, including a protocol called CELLOseq, which enables single cell long read quantification of authentic LINE1 transcripts. In addition to that, we implemented 10x 5´single cell transcriptomics, which during the library preparation uses a template switching oligo at the 5´end, thus enabling to identify true mRNA start sites. These two technologies are geared towards the unambiguous quantification of authentic LINE1 transcripts. We are now working
We performed a series of bulk transcriptomics with different technologies to study the relationship of immune-response transcriptional networks with transcription of specific retroelement (RTE) loci with available bioinformatic pipelines that we adjusted to increase their accuracy in locus specific mapping. Correlation analysis demonstrated that RTE in close proximity to differentially expressed genes were much more likely to become de-repressed than RTE that are located further away from a differentially expressed gene. In fact, we observed an inverse correlation between the expression level of an RTE and it´s distance to a significantly de-repressed response gene. This correlation was observed using different acute stimuli like global DNA damage by gama irradiation, LPS or Interferon alpha, as well as during simple spontaneous transcriptional reprogramming of cells cultured for extensive period of time, which we call “pseudo aging”. The effect was true for human and for mouse cells. These findings underpin the very fundamental hypothesis of the grant, that a significant proportion of retroelement transcriptional activity is in fact dictated by the chromatin landscape, an thus subject to environmental triggers.
We generated new dox-inducible L1-reporter constructs, which enabled us quantify the retrotransposition frequency following the induction of proximal DNA damage. Using this new system, we observed that the frequency of L1 retrotransposition was increased upon induction of double-strand breaks proximal to the integration site of the synthetic L1 reporter element. Since the L1-reporter transgenic cell culture was polyclonal regarding the integration site of the construct, we believe that this is a general observation across the whole genome.
We established a 2´3´cGAMP biosensor system in our lab, which will directly allow us to read out cGAS activation in the context of retrotransposition. This approach was not part of the original plan, but as some time has passed since writing of the proposal and technology keeps evolving, we will use the cGAMP-biosensor system to address questions about the immunogenicity of different LINE1 replication intermediates. We introduced the FRET-based cGAMP biosensor into 293 cells and expressed different mutants of cGAS (wild type, enzymatic dead, histone-binding mutant = constitutively active) and we saw cGAMP sensing that correlated with the expected levels of cGAS activity. Next, we treated the reporter cells with 5-Azacytidin to induce global de-repression of retroelements in the presence and absence of nucleoside reverse transcriptase inhibitors (NRTI) Truvada and Emtricitabine. We then induced the expression of the three cGAS versions. NRTI-treatment reduced cGAMP sensing in cells that expressed the constitutively active cGAS mutant, but not in cells expressing WT cGAS. We are now following up these observations by using more retroelement-specific triggers as compared to global epigenetic stress.
Collectively, we have established most of the toolboxes that we proposed and gathered data in support of my first hypothesis, that retroelement expression is subject to common environmental triggers.
