LUBAC is required for RIG-I sensing of RNA viruses

Abstract

The ability of cells to mount an interferon response to virus infections depends on intracellular nucleic acid sensing pattern recognition receptors (PRRs). RIG-I is an intracellular PRR that binds short double stranded viral RNAs to trigger MAVS-dependent signalling. The RIG-I/MAVS signalling complex requires the coordinated activity of multiple kinases and E3 ubiquitin ligases to activate the transcription factors that drive type I and type III interferon production from infected cells. The linear ubiquitin chain assembly complex (LUBAC) regulates the activity of multiple receptor signalling pathways in both ligase-dependent and -independent ways. Here, we show that the three proteins that constitute LUBAC have separate functions in regulating RIG-I signalling. Both HOIP, the E3 ligase capable of generating M1-ubiquitin chains, and LUBAC accessory protein HOIL-1 are required for viral RNA sensing by RIG-I. The third LUBAC component, SHARPIN, is not required for RIG-I signalling. These data cement the role of LUBAC as a positive regulator of RIG-I signalling and as an important component of antiviral innate immune responses.

HOIL1 mediates MDA5 activation through ubiquitination of LGP2

“…Together, these data indicate that HOIP E3 ligase activity plays a minor role in promoting RLR signaling and IFN induction, that may be obscured during viral infection, suggesting possible viral antagonism of RLR signaling or linear ubiquitination, or additional role(s) for HOIP in viral infection beyond innate RNA-sensing. These results corroborate findings from a recent study that reported a partial requirement for HOIP protein and HOIP E3 ligase activity for RIG-I signaling in A549 cells, reportedly through an interaction with and regulation of NEMO and TBK1 (37).…”

“…These findings suggest that the LUBAC regulates a component(s) common to MDA5 and RIG-I signaling, likely at the level of MAVS, or downstream of MAVS, where the signaling by both RLRs converges. This is in line with findings by Liu et al (36) and Teague et al (37), which support a model in which linear ubiquitin chains generated by HOIP facilitate the recruitment and retainment of NEMO to the MAVS signaling complex, which in turn recruits and activates kinases that activate IRF3 and NF-kB. Liu et al also reported possible redundancy between the LUBAC and TRAF2 in promoting MAVS signaling, which may partly explain why the LUBAC is not absolutely required for RLR-mediated IFN induction.…”

“…Nevertheless, multiple groups have reported a partial defect in RIG-I-mediated IFN induction with loss of HOIP, or HOIP's E3 catalytic activity (36,37), similar to what we present in this study. We extend these findings by identifying a similar role for HOIP ligase activity in MDA5 signaling.…”

“…In this study, we resolved the role of HOIL1 in MDA5 antiviral signaling by delineating the relative contributions of HOIL1 E3 ubiquitin ligase activity and LUBAC activity. The role of the LUBAC in RIG-I signaling has been previously studied, with conflicting conclusions likely stemming from differences in HOIP expression levels, incomplete deletion of LUBAC components, and cell type-dependent effects (20, 33–37). Nevertheless, multiple groups have reported a partial defect in RIG-I-mediated IFN induction with loss of HOIP, or HOIP’s E3 catalytic activity (36, 37), similar to what we present in this study.…”

“…Linear ubiquitination regulates multiple immune-related pathways, such as canonical NF-kB activation, NLRP3 inflammasome formation, and extrinsic cell death pathways (21,22,(25)(26)(27)(28)(29)(30)(31)(32). Several studies have also shown that the LUBAC negatively RIG-I signaling, while other studies have reported an opposing role (33)(34)(35)(36)(37).…”