Sensing of viral nucleic acids by RIG-I: from translocation to translation

Abstract

The innate immune system is a first layer of defense against infection by pathogens. It responds to pathogens by activating host defense mechanisms via interferon and inflammatory cytokine expression. Pathogen associated molecular patterns (PAMPs) are sensed by specific pattern recognition receptors. Among those, the ATP dependent helicase related RIG-I like receptors RIG-I, MDA5 and LGP2 sense the presence of viral RNA in the cytoplasm of host cells. While the precise PAMPs and functions of MDA5 or LGP2 are still unclear, RIG-I senses predominantly viral RNA containing a 5'-triphosphate along with dsRNA regions. Here we review our current knowledge of how these PAMPs are sensed and integrated by RIG-I, and how RIG-I's innate immune function can be used in translational medical approaches.

Figure 1. Domain structure of and signalling by the RLR family members

(A) RIG-I, MDA5 and LGP2 share a C-terminal RD domain and an SF2-type ATPase domain. In addition RIG-I and MDA5 have two N-termial CARD domains required for signalling via IPS1 that are absent in LGP2. (B) Signalling in the RLR pathway is initiated by binding of RNA to the respective receptor (RIG-I or MDA5). LGP2 may assist ligand recognition or inhibit signalling under certain circumstances. Binding of one of the receptor proteins to the mitochondrial adapter IPS1 leads to the activation of transcription factors of the IRF and NF-kB families. This ultimately leads to the production of interferons and inflammatory cytokines.

Figure 2. PAMPs recognised by RIG-I are found in viral genomic RNAs

Recent studies suggest that DI-genomic RNA may serve as a principal ligand for RIG-I in the situation of viral infection of a cell. In the case of Sendai Virus DI-genomes arise by copy back mechanisms forming a self-complementary structure (left). Influenza virus genome segments are inherently self-complementary at the ends (panhandle) but may also give rise to internally deleted DI-genomic segments (right).

Figure 3. Structures and mechanisms of RLR activation

(A) Regulatory/Repressor domains (ribbon models with highlighted secondary structure) of RIG-I, MDA5 and LGP2 possess the same fold, including a zinc (magenta sphere) binding site. Nucleic acids bind to a central core RNA binding site, while specificity for different types of RNA ends is provided by a specificity loop. (B) RDs emerge as dsRNA end binding domains. RIG-I RD preferentially binds to dsRNA ends containing a triphosphate moiety (color coded spheres) while LGP2 appears to recognise unphosphorylated ends. (C) Possible model for PAMP integration and signalling by RIG-I. Binding of pppRNA to RD and the SF2 domain could switch the enzyme into an active dimer. Translocation of the SF2 domain on dsRNA may induce a conformational change in RIG-I with exposed CARDs, leading to ubiquitylation and interaction with the IPS1 adaptor.