Research ArticleIMMUNOLOGY

RNA binding activates RIG-I by releasing an autorepressed signaling domain

See allHide authors and affiliations

Science Advances  02 Oct 2019:
Vol. 5, no. 10, eaax3641
DOI: 10.1126/sciadv.aax3641
  • Fig. 1 Design of a FRET reporter for RIG-I activation.

    (A) Donor and acceptor fluorophores were placed on the 2CARD and hel2i domains of RIG-I, respectively. Models of RIG-I in the autoinhibited state [Protein Data Bank (PDB) ID: 4A2W] and the activated state (PDB ID: 5F9H) predict that the distance between fluorophores changes significantly upon activation. (B) FlAsH donor fluorophore was attached by inserting a tetracysteine motif (CCPGCC) immediately C-terminal to the 2CARD domain. Alexa Fluor 594 acceptor fluorophore was attached via copper-free click chemistry to the unnatural amino acid azido-phenylalanine (AzF), which was incorporated by amber codon suppression. Structures were modeled using PyMOL (49).

  • Fig. 2 Dual-labeled RIG-I reports 2CARD ejection by change in FRET.

    (A) A schematic representation of the process used to dual-label RIG-I with donor and acceptor fluorophores. Critical features of this protocol are highlighted. (B) One micromolar TCEP is required to reduce nonspecific labeling of RIG-I by DIBO alkyne Alexa Fluor 594. Labeled protein was analyzed by SDS–polyacrylamide gel electrophoresis (PAGE), followed by fluorescent imaging (excitation/emission: 532 nm/LP575). In the absence of TCEP, negative control protein lacking AzF (−AzF) is nonspecifically labeled. (C) βME is required for efficient FlAsH labeling (top), but large amounts of βME cause nonspecific fluorescence by free FlAsH (bottom). One micromolar βME was identified as an ideal concentration that is sufficient for labeling without causing background fluorescence. Fluorescent emission spectra were measured after FlAsH excitation at 479 nm. (D) The optimized protocol produces specifically dual-labeled RIG-I. Tagged and untagged proteins were both used in the dual-labeling protocol and analyzed by fluorescent imaging of an SDS-PAGE gel (Alexa Fluor excitation/emission: 532 nm/LP575; FlAsH excitation/emission: 473 nm/530DF20). Only tagged protein was covalently modified, demonstrating the specificity of labeling. (E) Dual-labeled protein exhibits FRET. Fluorescent emission spectra were measured after FlAsH excitation at 479 nm, demonstrating acceptor emission at ~615 nm, significantly above bleed-through fluorescence from direct excitation/emission of individual fluorophores. (F) RNA causes a decrease in FRET corresponding to an ejection of the 2CARD domain. Treatment with benzonase digests the RNA and resets the 2CARD domain. Error bars correspond to the SEM across replicate samples.

  • Fig. 3 RNA binding ejects the 2CARD domain.

    Optimal RIG-I ligands containing 5′-triphosphorylated blunt duplex stems (SLR14 and SLR30) eject the 2CARD domain, whereas RNAs that do not bind (U55) do not eject the 2CARD domain. A duplex longer than 8 bp is required for efficient 2CARD ejection (SLR8). The 5′ triphosphate is not strictly required for 2CARD ejection (OH-SLR14), but internal duplexes do not eject the 2CARD domain {low polyinosinic:polycytidylic acid [poly(I:C)]}. Two molar equivalents of RNA were added to 50 nM protein to ensure saturation. Poly(I:C) was added at an equivalent concentration to SLR10 [low poly(I:C) = 0.79 ng/μl] or at 100 times the concentration [high poly(I:C) = 79 ng/μl)]. Error bars correspond to the SEM across replicate samples.

  • Fig. 4 ATP analogs have little effect on the conformation of the 2CARD domain.

    RIG-I was mixed with RNA, followed by the indicated ATP analog. A one-way analysis of variance (ANOVA) identified an effect of ATP analog on 2CARD ejection (F5,12 = 9.362, P = 0.0008) and a Dunnett’s post hoc test identified adenosine diphosphate (ADP)–AlFx as the only sample significantly different from the +SLR14 sample that lacked ATP analog (*P = 0.0143). Error bars correspond to the SEM across replicate samples.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/10/eaax3641/DC1

    Fig. S1. Tags were placed in locations designed to facilitate labeling and avoid perturbation of RIG-I function.

    Fig. S2. Controls demonstrate that dual labeling of RIG-I does not perturb its function, and FRET reports the distance between hel2i and 2CARD.

    Fig. S3. pppNS ejects 2CARD at high concentrations.

    Fig. S4. ADP-AlFx subtly but reproducibly increases 2CARD ejection.

    Table S1. RNA sequences used in this study.

    References (50, 51)

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Tags were placed in locations designed to facilitate labeling and avoid perturbation of RIG-I function.
    • Fig. S2. Controls demonstrate that dual labeling of RIG-I does not perturb its function, and FRET reports the distance between hel2i and 2CARD.
    • Fig. S3. pppNS ejects 2CARD at high concentrations.
    • Fig. S4. ADP-AlFx subtly but reproducibly increases 2CARD ejection.
    • Table S1. RNA sequences used in this study.
    • References (50, 51)

    Download PDF

    Files in this Data Supplement:

Navigate This Article