Structural insights into RNA recognition by RIG-I

Intracellular RIG-I-like receptors (RLRs, including RIG-I, MDA-5, and LGP2) recognize viral RNAs as pathogen-associated molecular patterns (PAMPs) and initiate an antiviral immune response. To understand the molecular basis of this process, we determined the crystal structure of RIG-I in complex with double-stranded RNA (dsRNA). The dsRNA is sheathed within a network of protein domains that include a conserved "helicase" domain (regions HEL1 and HEL2), a specialized insertion domain (HEL2i), and a C-terminal regulatory domain (CTD). A V-shaped pincer connects HEL2 and the CTD by gripping an α-helical shaft that extends from HEL1. In this way, the pincer coordinates functions of all the domains and couples RNA binding with ATP hydrolysis. RIG-I falls within the Dicer-RIG-I clade of the superfamily 2 helicases, and this structure reveals complex interplay between motor domains, accessory mechanical domains, and RNA that has implications for understanding the nanomechanical function of this protein family and other ATPases more broadly.     

RIG-I and dsRNA-Induced IFNβ Activation

Except for viruses that initiate RNA synthesis with a protein primer (e.g., picornaviruses), most RNA viruses initiate RNA synthesis with an NTP, and at least some of their viral _pppRNAs remain unblocked during the infection. Consistent with this, most viruses require RIG-I to mount an innate immune response, whereas picornaviruses require mda-5. We have examined a SeV infection whose ability to induce interferon depends on the generation of capped dsRNA (without free 5′ tri-phosphate ends), and found that this infection as well requires RIG-I and not mda-5. We also provide evidence that RIG-I interacts with poly-I/C in vivo, and that heteropolymeric dsRNA and poly-I/C interact directly with RIG-I in vitro, but in different ways; i.e., poly-I/C has the unique ability to stimulate the helicase ATPase of RIG-I variants which lack the C-terminal regulatory domain.     

Identification of Loss of Function Mutations in Human Genes Encoding RIG-I and MDA5: IMPLICATIONS FOR RESISTANCE TO TYPE I DIABETES*

Retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated gene 5 (MDA5) are essential for detecting viral RNA and triggering antiviral responses, including production of type I interferon. We analyzed the phenotype of non-synonymous mutants of human RIG-I and MDA5 reported in databases by functional complementation in cell cultures. Of seven missense mutations of RIG-I, S183I, which occurs within the second caspase recruitment domain repeat, inactivated this domain and conferred a dominant inhibitory function. Of 10 mutants of MDA5, two exhibited loss of function. A nonsense mutation, E627*, resulted in deletion of the C-terminal region and double-stranded RNA (dsRNA) binding activity. Another loss of function mutation, I923V, which occurs within the C-terminal domain, did not affect dsRNA binding activity, suggesting a novel and essential role for this residue in the signaling. Remarkably, these mutations are implicated in resistance to type I diabetes. However, the A946T mutation of MDA5, which has been implicated in type I diabetes by previous genetic analyses, affected neither dsRNA binding nor IFN gene activation. These results provide new insights into the structure-function relationship of RIG-I-like receptors as well as into human RIG-I-like receptor polymorphisms, antiviral innate immunity, and autoimmune diseases.Innate and adaptive immune systems constitute the defense against infections by pathogens. Immediately after an infection occurs, various cells in the body sense the virus and initiate antiviral responses in which type I IFN² plays a critical role, both in viral inhibition and in the subsequent adaptive immune response (1). The production of IFN is initiated when sensor molecules such as Toll-like receptors (TLRs) and RLRs detect virus-associated molecules. TLRs detect pathogen-associated molecular patterns (PAMPs) at the cell surface or in the endosome in immune cells such as dendritic cells and macrophages (2). RLRs sense viral RNA in the cytoplasm of most cell types and induce antiviral responses, including the activation of IFN genes (3). RLRs include RIG-I, MDA5, and laboratory of genetics and physiology 2 (LGP2).It is proposed that RLRs sense and activate antiviral signals through the coordination of their functional domains (4). The N-terminal region of RIG-I and MDA5 is characterized by two repeats of CARD and functions as an activation domain (3). This domain is responsible for the transduction of signals downstream to IFN-β promoter stimulator 1 (IPS-1) (also known as MAVS, VISA, and Cardif). The primary sequence of the CTD, consisting of ∼140 amino acids, is conserved among RLRs. The CTD of RIG-I functions as a viral RNA-sensing domain as revealed by biochemical and structural analyses (5, 6). Both dsRNA and 5′-ppp-ssRNA, which are generated in the cytoplasm of virus-infected cells, are recognized by a basic cleft structure of RIG-I CTD. In addition to its RNA recognition function, the CTD of RIG-I and LGP2 functions as a repression domain through interaction with the activation domain. The repression domain is responsible for keeping RIG-I inactive in non-stimulated cells (3, 7). The helicase domain, with DEXD/H box-containing RNA helicase motifs, is the largest domain found in RLRs. Once dsRNA or 5′-ppp-ssRNA is recognized by the CTD, the helicase domain causes structural changes to release the activation domain. ATP binding and/or its hydrolysis is essential for the conformational change because Walker''s ATP-binding site within the helicase domain is essential for signaling by RIG-I and MDA5.Analyses of knock-out mice have revealed that RIG-I and MDA5 recognize distinct RNA viruses (8, 9). Picornaviruses are detected by MDA5, but many other viruses such as influenza A, Sendai, vesicular stomatitis, and Japanese encephalitis are detected by RIG-I. The difference is based on the distinct non-self RNA patterns generated by viruses, as demonstrated by the finding that RIG-I is selectively activated by dsRNA or 5′-ppp ssRNA, whereas MDA5 is activated by long dsRNA (10–12).Single nucleotide polymorphisms (SNPs) of the human RIG-I and MDA5 genes including several non-synonymous SNPs (nsSNPs), which potentially alter the function of the proteins encoded, are reported in databases. In this report, we investigated the functions of nsSNPs of RIG-I and MDA5 by functional complementation using respective knock-out cells. We identified loss of function mutations of RIG-I and MDA5. Notably, two MDA5 mutations, E627* and I923V, recently reported to have a strong association with resistance to T1D (13), were severely inactive. The results suggest a novel molecular mechanism for the activation of RLRs and will contribute to our understanding of the functional effects of RLR polymorphisms and the critical relationship between RLR nsSNPs and diseases.     

Knock-in of the duck retinoic acid-inducible gene I (RIG-I) into the Mx gene in DF-1 cells enables both stable and immune response-dependent RIG-I expression

Waterfowls, such as ducks, are natural hosts of avian influenza virus (AIV) and can genetically limit the pathogenicity. On the other hand, some AIV strains cause severe pathogenicity in chickens. It is suggested that differences in the pathogenicity of AIV infection between waterfowls and chickens are related to the expression of retinoic acid-inducible gene I (RIG-I), a pattern recognition receptor that chickens evolutionally lack. Here, we knocked-in the duck RIG-I bearing the T2A peptide sequence at the 3′ region of the Mx, an interferon-stimulated gene (ISG), in chicken embryo fibroblast cells (DF-1) using the precise integration into target chromosome (PITCh) system to control the duck RIG-I expression in chickens. The expression patterns of the duck RIG-I were then analyzed using qPCR. The knocked-in DF-1 cells expressed RIG-I via the stimulation of IFN-β and poly(I:C) in a dose-dependent manner. Moreover, poly(I:C) stimulation in the knocked-in DF-1 cells upregulated RIG-I-like receptor (RLR) family signaling pathway-related genes IFN-β, OASL, and IRF7. The IFN-β-dependent expression of RIG-I and upregulation of IFN-β in the poly(I:C) stimulation demonstrated a positive-feedback loop via RIG-I, usually evident in ducks. Overall, this novel strategy established RIG-I-dependent immune response in chickens without overexpression of RIG-I and disruption of the host genes.     

RNA synthesis by the brome mosaic virus RNA-dependent RNA polymerase in human cells reveals requirements for de novo initiation and protein-protein interaction

Brome mosaic virus (BMV) is a model positive-strand RNA virus whose replication has been studied in a number of surrogate hosts. In transiently transfected human cells, the BMV polymerase 2a activated signaling by the innate immune receptor RIG-I, which recognizes de novo-initiated non-self-RNAs. Active-site mutations in 2a abolished RIG-I activation, and coexpression of the BMV 1a protein stimulated 2a activity. Mutations previously shown to abolish 1a and 2a interaction prevented the 1a-dependent enhancement of 2a activity. New insights into 1a-2a interaction include the findings that helicase active site of 1a is required to enhance 2a polymerase activity and that negatively charged amino acid residues between positions 110 and 120 of 2a contribute to interaction with the 1a helicase-like domain but not to the intrinsic polymerase activity. Confocal fluorescence microscopy revealed that the BMV 1a and 2a colocalized to perinuclear region in human cells. However, no perinuclear spherule-like structures were detected in human cells by immunoelectron microscopy. Sequencing of the RNAs coimmunoprecipitated with RIG-I revealed that the 2a-synthesized short RNAs are derived from the message used to translate 2a. That is, 2a exhibits a strong cis preference for BMV RNA2. Strikingly, the 2a RNA products had initiation sequences (5'-GUAAA-3') identical to those from the 5' sequence of the BMV genomic RNA2 and RNA3. These results show that the BMV 2a polymerase does not require other BMV proteins to initiate RNA synthesis but that the 1a helicase domain, and likely helicase activity, can affect RNA synthesis by 2a.     

The intrinsically disordered CARDs‐Helicase linker in RIG‐I is a molecular gate for RNA proofreading

The innate immune receptor RIG‐I provides a first line of defense against viral infections. Viral RNAs are recognized by RIG‐I''s C‐terminal domain (CTD), but the RNA must engage the helicase domain to release the signaling CARD (Caspase Activation and Recruitment Domain) domains from their autoinhibitory CARD2:Hel2i interactions. Because the helicase itself lacks RNA specificity, mechanisms to proofread RNAs entering the helicase domain must exist. Although such mechanisms would be crucial in preventing aberrant immune responses by non‐specific RNAs, they remain largely uncharacterized to date. This study reveals a previously unknown proofreading mechanism through which RIG‐I ensures that the helicase engages RNAs explicitly recognized by the CTD. A crucial part of this mechanism involves the intrinsically disordered CARDs‐Helicase Linker (CHL), which connects the CARDs to the helicase subdomain Hel1. CHL uses its negatively charged regions to antagonize incoming RNAs electrostatically. In addition to this RNA gating function, CHL is essential for stabilization of the CARD2:Hel2i interface. Overall, we uncover that the CHL and CARD2:Hel2i interface work together to establish a tunable gating mechanism that allows CTD‐chosen RNAs to bind the helicase domain, while at the same time blocking non‐specific RNAs. These findings also indicate that CHL could represent a novel target for RIG‐I‐based therapeutics.     

Uridine Composition of the Poly-U/UC Tract of HCV RNA Defines Non-Self Recognition by RIG-I

Viral infection of mammalian cells triggers the innate immune response through non-self recognition of pathogen associated molecular patterns (PAMPs) in viral nucleic acid. Accurate PAMP discrimination is essential to avoid self recognition that can generate autoimmunity, and therefore should be facilitated by the presence of multiple motifs in a PAMP that mark it as non-self. Hepatitis C virus (HCV) RNA is recognized as non-self by RIG-I through the presence of a 5′-triphosphate (5′-ppp) on the viral RNA in association with a 3′ poly-U/UC tract. Here we define the HCV PAMP and the criteria for RIG-I non-self discrimination of HCV by examining the RNA structure-function attributes that impart PAMP function to the poly-U/UC tract. We found that the 34 nucleotide poly-uridine “core” of this sequence tract was essential for RIG-I activation, and that interspersed ribocytosine nucleotides between poly-U sequences in the RNA were required to achieve optimal RIG-I signal induction. 5′-ppp poly-U/UC RNA variants that stimulated strong RIG-I activation efficiently bound purified RIG-I protein in vitro, and RNA interaction with both the repressor domain and helicase domain of RIG-I was required to activate signaling. When appended to 5′-ppp RNA that lacks PAMP activity, the poly-U/UC U-core sequence conferred non-self recognition of the RNA and innate immune signaling by RIG-I. Importantly, HCV poly-U/UC RNA variants that strongly activated RIG-I signaling triggered potent anti-HCV responses in vitro and hepatic innate immune responses in vivo using a mouse model of PAMP signaling. These studies define a multi-motif PAMP signature of non-self recognition by RIG-I that incorporates a 5′-ppp with poly-uridine sequence composition and length. This HCV PAMP motif drives potent RIG-I signaling to induce the innate immune response to infection. Our studies define a basis of non-self discrimination by RIG-I and offer insights into the antiviral therapeutic potential of targeted RIG-I signaling activation.     

Dicer-related helicase 3 forms an obligate dimer for recognizing 22G-RNA

Dicer is a specialized nuclease that produces RNA molecules of specific lengths for use in gene silencing pathways. Dicer relies on the correct measurement of RNA target duplexes to generate products of specific lengths. It is thought that Dicer uses its multidomain architecture to calibrate RNA product length. However, this measurement model is derived from structural information from a protozoan Dicer, and does not account for the helicase domain present in higher organisms. The Caenorhabditis elegans Dicer-related helicase 3 (DRH-3) is an ortholog of the Dicer and RIG-I family of double-strand RNA activated ATPases essential for secondary siRNA production. We find that DRH-3 specifies 22 bp RNAs by dimerization of the helicase domain, a process mediated by ATPase activity and the N-terminal domain. This mechanism for RNA length discrimination by a Dicer family protein suggests an alternative model for RNA length measurement by Dicer, with implications for recognition of siRNA and miRNA targets.     

The RIG-I ATPase core has evolved a functional requirement for allosteric stabilization by the Pincer domain

Retinoic acid-inducible gene I (RIG-I) is a pattern recognition receptor expressed in metazoan cells that is responsible for eliciting the production of type I interferons and pro-inflammatory cytokines upon detection of intracellular, non-self RNA. Structural studies of RIG-I have identified a novel Pincer domain composed of two alpha helices that physically tethers the C-terminal domain to the SF2 helicase core. We find that the Pincer plays an important role in mediating the enzymatic and signaling activities of RIG-I. We identify a series of mutations that additively decouple the Pincer motif from the ATPase core and show that this decoupling results in impaired signaling. Through enzymological and biophysical analysis, we further show that the Pincer domain controls coupled enzymatic activity of the protein through allosteric control of the ATPase core. Further, we show that select regions of the HEL1 domain have evolved to potentiate interactions with the Pincer domain, resulting in an adapted ATPase cleft that is now responsive to adjacent domains that selectively bind viral RNA.     

The Essential,Nonredundant Roles of RIG-I and MDA5 in Detecting and Controlling West Nile Virus Infection

Virus recognition and response by the innate immune system are critical components of host defense against infection. Activation of cell-intrinsic immunity and optimal priming of adaptive immunity against West Nile virus (WNV), an emerging vector-borne virus, depend on recognition by RIG-I and MDA5, two cytosolic pattern recognition receptors (PRRs) of the RIG-I-like receptor (RLR) protein family that recognize viral RNA and activate defense programs that suppress infection. We evaluated the individual functions of RIG-I and MDA5 both in vitro and in vivo in pathogen recognition and control of WNV. Lack of RIG-I or MDA5 alone results in decreased innate immune signaling and virus control in primary cells in vitro and increased mortality in mice. We also generated RIG-I^−/− × MDA5^−/− double-knockout mice and found that a lack of both RLRs results in a complete absence of innate immune gene induction in target cells of WNV infection and a severe pathogenesis during infection in vivo, similar to findings for animals lacking MAVS, the central adaptor molecule for RLR signaling. We also found that RNA products from WNV-infected cells but not incoming virion RNA display at least two distinct pathogen-associated molecular patterns (PAMPs) containing 5′ triphosphate and double-stranded RNA that are temporally distributed and sensed by RIG-I and MDA5 during infection. Thus, RIG-I and MDA5 are essential PRRs that recognize distinct PAMPs that accumulate during WNV replication. Collectively, these experiments highlight the necessity and function of multiple related, cytoplasmic host sensors in orchestrating an effective immune response against an acute viral infection.     

Evidence for a Novel Mechanism of Influenza Virus-Induced Type I Interferon Expression by a Defective RNA-Encoded Protein

Influenza A virus (IAV) defective RNAs are generated as byproducts of error-prone viral RNA replication. They are commonly derived from the larger segments of the viral genome and harbor deletions of various sizes resulting in the generation of replication incompatible viral particles. Furthermore, small subgenomic RNAs are known to be strong inducers of pattern recognition receptor RIG-I-dependent type I interferon (IFN) responses. The present study identifies a novel IAV-induced defective RNA derived from the PB2 segment of A/Thailand/1(KAN-1)/2004 (H5N1). It encodes a 10 kDa protein (PB2_∆) sharing the N-terminal amino acid sequence of the parental PB2 protein followed by frame shift after internal deletion. PB2_∆ induces the expression of IFNβ and IFN-stimulated genes by direct interaction with the cellular adapter protein MAVS, thereby reducing viral replication of IFN-sensitive viruses such as IAV or vesicular stomatitis virus. This induction of IFN is completely independent of the defective RNA itself that usually serves as pathogen-associated pattern and thus does not require the cytoplasmic sensor RIG-I. These data suggest that not only defective RNAs, but also some defective RNA-encoded proteins can act immunostimulatory. In this particular case, the KAN-1-induced defective RNA-encoded protein PB2_∆ enhances the overwhelming immune response characteristic for highly pathogenic H5N1 viruses, leading to a more severe phenotype in vivo.     

Structure and function of LGP2, a DEX(D/H) helicase that regulates the innate immunity response

RNA recognition receptors are important for detection of and response to viral infections. RIG-I and MDA5 are cytoplasmic DEX(D/H) helicase proteins that can induce signaling in response to RNA ligands, including those from viral infections. LGP2, a homolog of RIG-I and MDA5 without the caspase recruitment domain required for signaling, plays an important role in modulating signaling by MDA5 and RIG-I, presumably through heterocomplex formation and/or by serving as a sink for RNAs. Here we demonstrate that LGP2 can be coexpressed with RIG-I to inhibit activation of the NF-kappaB reporter expression and that LGP2 protein produced in insect cells can bind both single- and double-stranded RNA (dsRNA), with higher affinity and cooperativity for dsRNA. Electron microscopy and image reconstruction were used to determine the shape of the LGP2 monomer in the absence of dsRNA and of the dimer complexed to a 27-bp dsRNA. LGP2 has striking structural similarity to the helicase domain of the superfamily 2 DNA helicase, Hef.     

The Ighmbp2 helicase structure reveals the molecular basis for disease-causing mutations in DMSA1

Mutations in immunoglobulin µ-binding protein 2 (Ighmbp2) cause distal spinal muscular atrophy type 1 (DSMA1), an autosomal recessive disease that is clinically characterized by distal limb weakness and respiratory distress. However, despite extensive studies, the mechanism of disease-causing mutations remains elusive. Here we report the crystal structures of the Ighmbp2 helicase core with and without bound RNA. The structures show that the overall fold of Ighmbp2 is very similar to that of Upf1, a key helicase involved in nonsense-mediated mRNA decay. Similar to Upf1, domains 1B and 1C of Ighmbp2 undergo large conformational changes in response to RNA binding, rotating 30° and 10°, respectively. The RNA binding and ATPase activities of Ighmbp2 are further enhanced by the R3H domain, located just downstream of the helicase core. Mapping of the pathogenic mutations of DSMA1 onto the helicase core structure provides a molecular basis for understanding the disease-causing consequences of Ighmbp2 mutations.