Ligand-induced receptor oligomerization is a ubiquitous mechanism for signal activation in the immune system. Understanding how these receptors oligomerize during ligand recognition and activate downstream signaling pathways is fundamental to understanding their functions and is pre-requisite to therapeutic application of these receptors. In this proposal, we describe our strategies to investigate the oligomerization mechanism, oligomer architecture and signal activation mechanism of a conserved innate immune receptor, retinoic acid inducible gene-1 (RIG-I). RIG-I and its paralog, MDA5, represent a major receptor family that recognizes viral RNAs in the cytoplasm of a broad range of cell types. RIG-I and MDA5 both contain a tandem caspase activation recruitment domain (2CARD) for signal activation and a helicase domain and a C-terminal domain for RNA binding and RNA- dependent ATP hydrolysis. 2CARDs of RIG-I and MDA5 activate their downstream adaptor molecule, MAVS, by promoting its monomer-to-filament transition. MAVS filaments, in turn, recruit further downstream signaling molecules to activate the IFN?/? signaling pathways. We have previously shown that MDA5 cooperatively forms a filament along the length of dsRNA, and that its formation is important for high affinity interaction with dsRNA, oligomerization of 2CARD and dsRNA length dependent regulation of signaling activity. Unlike MDA5, oligomerization of RIG-I has been unclear, and has been thought to strictly require a co-factor, K63-linked polyubiquitin chains. Recently, we found that RIG-I assembles into a filament during ATP hydrolysis, and that the filament can directly activate MAVS in the absence of polyubiquitin chains, suggesting a novel mechanism for RNA recognition and signal activation by RIG-I. We here propose to determine the precise mechanisms by which RIG-I assembles into a filament (Aim 1) and stimulates MAVS filament formation independent of or together with polyubiquitin chains (Aim 2), and how RIG-I filaments form and function in the context of viral infection (Aim 3). This proposal builds upon our previous research on the MDA5 filament (Wu et al, Cell, 2013; Peisley et al, PNAS, 2012 & PNAS, 2011; Rice et al, Nat. Genetics, In press), our discovery of the RIG-I filament (Peisley et al, Mol. Cell, 2013), a very recent crystal structure of RIG-I 2CARDtetramer bound by K63-Ubn (Peisley et al, Nature, Epub) and finally the atomic structures of the MAVS filament as well as the RIG-I 2CARD:MAVS CARD complex (manuscript in preparation). We believe that the proposed research will provide a comprehensive picture of functions of RIG-I and help us dissect commonalities and divergences between RIG-I and MDA5 in viral RNA detection and signal activation mechanisms. Furthermore, detailed mechanistic understanding of the RIG-I filament could potentially offer insights into novel therapeutic strategies to modulate the activity of RIG-I in antiviral and anticancer therapies.