Ubiquitin-dependent and -independent roles of E3 ligase RIPLET in innate immunity

Abstract

The conventional view posits that E3 ligases function primarily through conjugating ubiquitin (Ub) to their substrate molecules. We report here that RIPLET, an essential E3 ligase in antiviral immunity, promotes the antiviral signaling activity of the viral RNA receptor RIG-I through both Ub-dependent and -independent manners. RIPLET uses its dimeric structure and a bivalent binding mode to preferentially recognize and ubiquitinate RIG-I pre-oligomerized on dsRNA. In addition, RIPLET can cross-bridge RIG-I filaments on longer dsRNAs, inducing aggregate-like RIG-I assemblies. The consequent receptor clustering synergizes with the Ub-dependent mechanism to amplify RIG-I-mediated antiviral signaling in an RNA-length dependent manner.These observations show the unexpected role of an E3 ligase as a co-receptor that directly participates in receptor oligomerization and ligand discrimination. It also highlights a previously unrecognized mechanism by which the innate immune system measures foreign nucleic acid length, a common criterion for self vs. non-self nucleic acid discrimination. local concentration of 2CARD (Peisley et al., 2013). Second, in addition to RIG-I filament formation, K63-linked polyubiquitin (K63-Ub n ) was also shown to promote RIG-I signaling (Jiang et al., 2012). Structural studies further showed that K63-Ub n chains bind the periphery of the core 2CARD tetramer, bridging the adjacent subunits and stabilizing its assembly . The 2CARD tetramer then acts as a helical template to nucleate the MAVS filament formation for downstream signal activation (Wu and Hur, 2015).Despite the detailed understanding of the action of K63-Ub n on RIG-I, much remains debated as to how and when Ub is placed on RIG-I, which E3 ligase is involved and how it interplays with RIG-I filament formation. Previous studies reported that TRIM25 and RIPLET (i.e. RNF135) are two essential E3 ligases important for RIG-I signaling (Gack et al., 2007;Gao et al., 2009;Oshiumi et al., 2009;Oshiumi et al., 2010). A more recent study proposed that RIPLET and TRIM25 sequentially act on RIG-I upon viral RNA engagement . That is, RIPLET first acts on the C-terminal portion of RIG-I, releasing 2CARD, which is then modified by TRIM25. However, given the previous finding that RNA binding is sufficient to release 2CARD (Kowalinski et al., 2011), it was unclear whether RIPLET indeed acts to release 2CARD and if so, how. At the same time, accumulating evidence suggested that TRIM25 has multiple, RIG-I-independent antiviral functions (Choudhury et al., 2014;Li et al., 2017;Manokaran et al., 2015;Meyerson et al., 2017;Zheng et al., 2017), raising the question whether the observed effect of TRIM25 on RIG-I represents a direct or an indirect effect.We here report a combination of cellular and biochemical data showing that RIG-I activation is dependent on RIPLET, not TRIM25, and that RIPLET suffices to ubiquitinate and activate RIG-I. In addition, RIPLET recognizes the filamentous form of RIG-I using two distinct binding modes, the interplays of which o...

RIPLET, and not TRIM25, is required for endogenous RIG‐I‐dependent antiviral responses

“…The mechanism of Riplet action appears to be considerably more complex than previously thought. Cadena and colleagues have suggested that a Riplet dimer acts as a coreceptor for RIG‐I, with two Riplet‐PRYSPRY domains cross‐bridging and stabilizing two RIG‐I molecules oligomerized on dsRNA . Consistent with our findings, and those of Shi and colleagues, Cadena et al .…”

“…Cadena and colleagues have suggested that a Riplet dimer acts as a coreceptor for RIG‐I, with two Riplet‐PRYSPRY domains cross‐bridging and stabilizing two RIG‐I molecules oligomerized on dsRNA . Consistent with our findings, and those of Shi and colleagues, Cadena et al . also show that deletion of Trim25 has no impact on RIG‐I activity in response to either RNA ligands or infection with SeV.…”

“…To determine whether Riplet regulation of RIG‐I responses was dependent on the type of RNA ligand, RIG‐I‐, Riplet‐ and TRIM25‐deficient A549 cells were transfected with 5′‐ppp dsRNA (a specific RIG‐I agonist), high‐ and low‐molecular‐weight polyinosinic:polycytidylic acid (a dsRNA analog that can activate RIG‐I/MDA5 and TLR3) and a short hairpin 5′‐ppp dsRNA (M7; reported to have enhanced activity for RIG‐I) . Consistent with the original description, M7 was highly potent, achieving induction of IFN at 1000‐fold lower amounts than 5′‐ppp RNA (Figure a).…”

To TRIM or not to TRIM: the balance of host–virus interactions mediated by the ubiquitin system

“…The notion of different E3 ligases playing critical roles in the activation of RIG-I has recently been challenged with some groups identifying Riplet, and not TRIM25, as the required E3 ligase [126][127][128]. The challenge was raised when Riplet -/-, but not TRIM25 -/-, cell lines failed to induce innate immune signalling upon stimulation [126], although these experiments did not address different time points or other stimulation conditions that may trigger RIG-I activation, or the possibility of functional redundancy. Furthermore, it is unclear why different reports using the same TRIM25 -/-MEF cell lines have yielded different results.…”

“…Additionally, the authors examined a single TRIM25 -/-HEK293T clone which would not eliminate the possibility of off-target effects as a result of the guide RNA used, especially because two TRIM25 -/-HEK293 clones from another study showed a reduced IFN-I response during SeV infection [114]. Riplet -/mice and cell lines are known to have significant losses in both IFN-I and pro-inflammatory responses to infection [123,126,128]. Riplet is also known to bind and ubiquitinate several Lys residues in the CTD and linker region of RIG-I while TRIM25 only targets Lys residues in the 2CARD domain ( Fig.…”