Preview

PDF, English Download (16MB) | Terms of use

Abstract

RIG-I-like receptors (RLR) are cytosolic pattern recognition receptors (PRR) which are pivotal for the detection of virus infection. RIG-I senses viral double-stranded RNAs (dsRNA) and initiates cellular antiviral defense responses, resulting in the expression of type I and III interferons (IFN). Secreted IFNs signal in an auto and paracrine manner and mediate the transcriptional induction of distinct IFN-stimulated genes (ISG), which collectively establish an antiviral state of the cell. Whereas the topology of this pathway has been described thoroughly, the dynamics, particularly of the RIG-I-mediated IFN induction, are much less understood. In this study, I employed electroporation-based transfection with virus-like 5’ppp-dsRNA to synchronously activate the RIG-I signaling pathway in human lung adenocarcinoma (A549) cells and thus characterize the dynamics of cell-intrinsic innate immune responses. For this purpose, I focused on time-resolved western blotting of key pathway components, live-cell imaging of transcription factor relocalization, and quantitative RT-PCR of various target genes. Although previous studies reported intriguing cell-intrinsic stochasticity in the activation of the RLR and IFN signaling pathways, simultaneous 5’ppp-dsRNA stimulation of A549 cells resulted in highly deterministic and synchronous RIG-I signaling. By employing an IFN-blind A549 cell line, harboring functional knockouts of the receptors required for type I, II, and III IFN signaling, I analyzed the differences between primary RIG-I-mediated signaling and the subsequent signaling phase downstream of the IFN receptors. Interestingly, IFN signaling through the JAK/STAT cascade was not required to induce the production of IFN. Utilizing the generated kinetic data of antiviral signaling as foundation and collaborating with computational scientists, we developed and calibrated a comprehensive mathematical model of the cell-intrinsic antiviral response system. This model is able to predict the kinetics of signaling events induced upon dsRNA recognition by RIG-I as well as feedback and signal amplification through IFN and JAK/STAT signaling. Furthermore, I examined the impact of viral antagonists on signaling dynamics by employing viral proteins interfering with the host antiviral response system at defined steps: the dengue virus (DENV) protein NS5 interferes with IFN signaling, whereas the NS3/4A protease of hepatitis C virus (HCV), the Npro protease of classical swine fever virus (CSFV), as well as NS1 of influenza A virus (IAV) all target RLR signaling and thereby inhibit the induction of IFN. Additionally, the ORF6 protein of SARS-CoV-2 exhibits a multi-level strategy to impede defense responses by targeting both IFN induction and IFN signaling. Strikingly, the impact of these viral antagonists on antiviral signaling dynamics could be properly simulated by the established mathematical model. Consequently, our model permits in silico simulation of viral interference with the antiviral response system and provides a powerful tool to study the impact of yet unknown viral antagonists or other factors perturbing antiviral innate immune responses. Lastly, since previous work in our lab demonstrated an unexpectedly high overlap of RLR and IFN signaling-induced ISG expression, I used whole-transcriptome expression profiling to examine and compare the underlying roles of the transcription factors IRF1 and IRF3 in IFN-independent RIG-I-mediated signaling. Interestingly, dsRNA-induced expression of certain genes was both IRF3 and IFN-independent. The dsRNA-induced expression of some of those genes was partially NF-κB-dependent, whereas others seemed to be dependent on IRF1 or other transcription factors. In conclusion, this study provides insights into RIG-I-mediated but IFN-independent signaling upon virus-like dsRNA stimulation and might, in conjunction with the established mathematical model, facilitate deciphering the complexity of the virus-host interface.