During this project, we mostly achieved our objectives for WP1 and WP3 and we have published part of the results for these two work packages. The work pertaining to WP2 is still ongoing but will also eventually lead to at least three publications. Details of the main results obtained per work package are listed below.
-WP1. The biogenesis of miRNAs requires the sequential action of two ribonucleases, Drosha and Dicer, to give rise to an RNA duplex. This process is highly regulated by proteins that can bind directly to the pri-miRNA, or to Drosha or Dicer to modulate their efficiency. As a model for this work, we are using KSHV, which encodes a cluster of twelve miRNAs. Although all KSHV miRNAs derive from the same transcript, their accumulation is regulated at the post-transcriptional level. We first determined the importance of sequence determinants and/or the secondary structure of KSHV pri-miRNA transcript in its cleavage efficiency by Drosha. This part has been concluded and we could identify two important determinants that are essential for the correct accumulation of all miRNAs within the cluster. These results have been deposited as a preprint on the bioRxiv repository and will be published soon. The second part of this work package consists in a proteomics analysis of putative proteins binding individual KSHV pre-miRNAs. We have designed an approach to pull-down proteins binding to synthetic pre-miRNAs. We obtained data from more than eighty mass spectrometry runs that we are in the process of validating.
-WP2. Here, we first set to identify cellular miRNAs that can regulate virus infection. To this end, we have performed a genome wide screen to look at the effect of each human miRNA on Sindbis virus (SINV) infection. We found a handful of miRNAs that positively regulate this virus. Very interestingly, among these was the neuron-specific miR-124. We also showed that miR-124 can act on the related Chikungunya virus. We found that miR-124 acts by directly binding to the viral genome and that blocking it negatively impacts viral replication in neuronal cells. These results have been published in Journal of Virology. In another part of this WP, we studied ribonucleases that can degrade small RNAs. More specifically, we studied the involvement of the DIS3L2 nuclease in viral infection. This led to the unexpected discovery that DIS3L2 KO cells are more resistant to SINV infection, while cells expressing a catalytically inactivated DIS3L2 show an increased infection. We are in the process of finalizing this part of the project. Finally, we took an alternative approach to identify other players involved in small RNA decay. We are using an in vivo biotinylation approach to find partners of the mouse Argonaute 2 protein during infection by MCMV. This part of the project could not be completed and will require additional work.
-WP3. To study the interactome of the Dicer protein, we generated stable cell lines expressing a tagged version of the protein. We immunoprecipitated the tagged Dicer in non-infected and virus-infected cells. We obtained very interesting results, which show that known dsRNA sensors interact with Dicer only in infected cells. This indicates that Dicer could play a role in sensing viral RNA, but rather than actively degrade it, can help to recruit other antiviral factors. We found that one such factor is PKR and that the domain of Dicer required for its interaction is the helicase domain. We published these findings in the journal PLOS Pathogens. In parallel, we also tackled the question of the importance of RNAi from a completely unbiased point of view. We set on to globally identify proteins involved in the sensing of dsRNA and in the cellular response to this molecule. To this end, we developed two unrelated but complementary approaches. The first one consists in pulling-down dsRNA in mock and SINV-infected cells using an antibody that specifically recognizes dsRNA molecules and to then analyze proteins using mass spectrometry. This allowed us to identify known dsRNA binding proteins (dsRBP) such as ADAR1, PKR and Dicer, as well as unexpected ones including components of nuclear paraspeckles. We validated the importance of one of those in the cellular response to SINV. We have posted these results in the form of a pre-print on bioRxiv and submitted the manuscript for publication. The other approach we undertook was based on a genome-wide CRISPR-Cas9 survival screen aiming at identifying genes involved in the cell death induced by dsRNA transfection. Although we could not identify factors that are essential for signaling downstream of dsRNA detection, this approach allowed us to identify components that are critical for the cellular entry of dsRNA. These results were published in the journal mSphere.
